Imaging Pearls ❯ November 2024
|
-- OR -- |
|
3D and Workflow
- “Cinematic rendering, a recently described 3D rendering technique, uses a global illumination model that considers direct and indirect lighting to create images with photorealistic quality. Cinematic rendering can accentuate subtle texture changes and improve tumor conspicuity relative to traditional 2D images, 3D volume rendering, or maximum intensity projection images. Cinematic rendering may be able to enhance the visualization of spatial relationships among the tumor and adjacent vasculature, differentiating true tumor infiltration from simple proximity to vessels. This can potentially improve the assessment of resectability and assist in determining optimal vascular reconstruction options. Cinematic rendering vascular maps illustrate the major arteries and veins with exquisite detail and can highlight the presence of variant vascular anatomy that may increase the risk of complications, such as hemorrhage, ischemia, anastomotic leakage, or pseudoaneurysm formation.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - ”Cinematic rendering vascular maps illustrate the major arteries and veins with exquisite detail and can highlight the presence of variant vascular anatomy that may increase the risk of complications, such as hemorrhage, ischemia, anastomotic leakage, or pseudoaneurysm formation. At our institution, cinematic rendering has been routinely incorporated into the multidisciplinary PDAC clinic since 2018, and it has played an important role in tumor staging as well as patient management. Moreover, cinematic rendering data can be imported into augmented reality headsets to provide an immersive experience for the surgeon for operative planning.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Cinematic rendering, a recently described 3D rendering technique, uses a global illumination model that considers direct and indirect lighting to create images with photorealistic quality. Cinematic rendering can accentuate subtle texture changes and improve tumor conspicuity relative to traditional 2D images, 3D volume rendering, or maximum intensity projection images. Cinematic rendering may be able to enhance the visualization of spatial relationships among the tumor and adjacent vasculature, differentiating true tumor infiltration from simple proximity to vessels. This can potentially improve the assessment of resectability and assist in determining optimal vascular reconstruction options. Cinematic rendering vascular maps illustrate the major arteries and veins with exquisite detail and can highlight the presence of variant vascular anatomy that may increase the risk of complications, such as hemorrhage, ischemia, anastomotic leakage, or pseudoaneurysm formation.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - ”Cinematic rendering vascular maps illustrate the major arteries and veins with exquisite detail and can highlight the presence of variant vascular anatomy that may increase the risk of complications, such as hemorrhage, ischemia, anastomotic leakage, or pseudoaneurysm formation. At our institution, cinematic rendering has been routinely incorporated into the multidisciplinary PDAC clinic since 2018, and it has played an important role in tumor staging as well as patient management. Moreover, cinematic rendering data can be imported into augmented reality headsets to provide an immersive experience for the surgeon for operative planning.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063
Deep Learning
- “Based on available data, it is estimated that roughly 30% of radiologists in the United States are currently utilizing AI within their practice. Additionally, only 30% of current radiology practices reportedly employ AI tools. Despite the relatively modest current adoption rate, there is a growing expectation for AI to become significantly involved in radiology practice in the future. The slow implementation of AI in clinical radiology practice is attributed to the complexity of integration and the necessity for further validation of its efficacy.”
Artificial Intelligence in Radiology: What Is Its True Role at Present, and Where Isthe Evidence?
Alireza Mohseni, Elena Ghotbi, Foad Kazemi
Radiol Clin N Am 62 (2024) 935–947 - “Radiomics, focused on extracting quantitative features from medical images, contributes to personalized medicine by capturing the heterogeneity of phenotypes. Despite being distinct from AI, radiomics is closely linked to it, and the integration of radiomics with DL features enhances predictive capacities in radiology. Radiomics outperforms traditional models in predicting clinical outcomes, offering supplementary guidance in clinical decision- making. Its integration into nomograms demonstrates excellent discrimination in predicting various pathologic and clinical features.”
Artificial Intelligence in Radiology: What Is Its True Role at Present, and Where Isthe Evidence?
Alireza Mohseni, Elena Ghotbi, Foad Kazemi
Radiol Clin N Am 62 (2024) 935–947 - “Another noteworthy software is the vessel occlusion detection software, which has garnered attention for its evaluation in acute stroke scenarios, showcasing the transformative potential of AI in radiology. This software specializes in automated detection of intracranial vessel occlusions on computed tomography angiography (CTA), aiding in the rapid diagnosis and treatment of acute ischemic stroke. Its cost effectiveness and accuracy make it a valuable tool in stroke management protocols, highlighting the diverse applications of AI in radiology beyond traditional imaging interpretation.”
Artificial Intelligence in Radiology: What Is Its True Role at Present, and Where Isthe Evidence?
Alireza Mohseni, Elena Ghotbi, Foad Kazemi
Radiol Clin N Am 62 (2024) 935–947 - “The future of AI in radiology is also influenced by the changing regulatory environment, marked by ongoing endeavors to tackle regulatory and ethical concerns linked to AI applications. The analysis by Pesapane and colleagues underscores the importance of standardizing AI software specifications, classifications, and evaluations to ensure the effective integration of AI applications in clinical practice, reflecting the dynamic nature of the regulatory landscape and its effects on the future of radiology. Additionally, the collaborative statement by European and North American multisociety on the ethics of artificial intelligence in radiology highlights the necessity for ethical governance and transparency in AI applications, shedding light on the evolving regulatory environment and its implications for the future of radiology.”
Artificial Intelligence in Radiology: What Is Its True Role at Present, and Where Isthe Evidence?
Alireza Mohseni, Elena Ghotbi, Foad Kazemi
Radiol Clin N Am 62 (2024) 935–947 - Pearls:
Using AI for CAD and diagnosis in radiology has been shown to improve diagnostic accuracy for conditions including pulmonary nodules, breast abnormalities, and tuberculosis when used in combination with radiologist interpretation.
Radiomics utilize quantitative image features to develop personalized prediction models of prognosis and treatment response across diverse organs and imaging modalities.
Implementation of AI workflows in radiology requires extensive validation using standardized datasets and close integration with existing clinical systems.
Artificial Intelligence in Radiology: What Is Its True Role at Present, and Where Is the Evidence?
Alireza Mohseni, Elena Ghotbi, Foad Kazemi
Radiol Clin N Am 62 (2024) 935–947 - Relying solely on AI diagnoses without radiologist supervision could lead to critical imaging findings being overlooked or misinterpreted. Insufficient clinical validation and lack of transparency around AI algorithms hampers real-world adoption in patient care settings due to ethical and accuracy concerns. AI applications trained on small or biased datasets struggle to generalize well to diverse patient populations seen in clinical practice.
Artificial Intelligence in Radiology: What Is Its True Role at Present, and Where Is the Evidence?
Alireza Mohseni, Elena Ghotbi, Foad Kazemi
Radiol Clin N Am 62 (2024) 935–947 
Artificial Intelligence in Radiology: What Is Its True Role at Present, and Where Is the Evidence?
Alireza Mohseni, Elena Ghotbi, Foad Kazemi
Radiol Clin N Am 62 (2024) 935–947
- “Radiology plays an important role in the initial diagnosis and staging of patients with pancreatic ductal adenocarcinoma (PDAC). CT is the preferred modality over MRI due to wider availability, greater consistency in image quality, and lower cost. MRI and PET/CT are usually reserved as problem-solving tools in select patients. The National Comprehensive Cancer Network (NCCN) guidelines define resectability criteria based on tumor involvement of the arteries and veins and triage patients into resectable, borderline resectable, locally advanced, and metastatic categories. Patients with resectable disease are eligible for upfront surgical resection, while patients with high-stage disease are treated with neoadjuvant chemotherapy and/or radiation therapy with hopes of downstaging the disease. The accuracy of staging critically depends on the imaging technique and the experience of the radiologists. Several challenges in accurate preoperative staging include prediction of lymph node metastases, detection of subtle liver and peritoneal metastases, and disease restaging following neoadjuvant therapy.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Artificial intelligence (AI) has the potential to function as ‘second readers’ to improve upon the radiologists’ detection of small early-stage tumors, which can shift more patients toward surgical resection of potentially curable cancer. AI may also provide imaging biomarkers that can predict disease recurrence and patient survival after pancreatic resection and assist in the selection of patients most likely to benefit from surgery, thus improving patient outcomes.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “PDACs classically present as hypoenhancing masses with associated pancreatic duct dilatation and glandular atrophy of the body and tail. Pancreatic head tumors can cause common bile duct dilatation in addition to pancreatic duct dilatation, also known as the ‘double duct sign’. Up to 20% of PDACs enhance to the same degree as the background pancreas, and this isoattenuating pattern is more commonly found with smaller ( ≤20 mm) tumors. These small isoattenuating tumors can be difficult to detect on CT; therefore, radiologists often rely on secondary signs of the pancreatic duct or common bile duct dilatation for tumor detection. MRI and PET/CT have reported sensitivities of 79.2 and 73.7% in the detection of isoattenuating tumors, respectively, and may aid in detecting suspected pancreatic tumors that are occult on CT. Endoscopic ultrasound is crucial in confirming tissue diagnosis of suspected pancreatic malignancy. It is also an important second-line modality in detecting suspected pancreatic tumors that are occult on CT or MRI.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063- “Cinematic rendering, a recently described 3D rendering technique, uses a global illumination model that considers direct and indirect lighting to create images with photorealistic quality. Cinematic rendering can accentuate subtle texture changes and improve tumor conspicuity relative to traditional 2D images, 3D volume rendering, or maximum intensity projection images. Cinematic rendering may be able to enhance the visualization of spatial relationships among the tumor and adjacent vasculature, differentiating true tumor infiltration from simple proximity to vessels. This can potentially improve the assessment of resectability and assist in determining optimal vascular reconstruction options. Cinematic rendering vascular maps illustrate the major arteries and veins with exquisite detail and can highlight the presence of variant vascular anatomy that may increase the risk of complications, such as hemorrhage, ischemia, anastomotic leakage, or pseudoaneurysm formation.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - ”Cinematic rendering vascular maps illustrate the major arteries and veins with exquisite detail and can highlight the presence of variant vascular anatomy that may increase the risk of complications, such as hemorrhage, ischemia, anastomotic leakage, or pseudoaneurysm formation. At our institution, cinematic rendering has been routinely incorporated into the multidisciplinary PDAC clinic since 2018, and it has played an important role in tumor staging as well as patient management. Moreover, cinematic rendering data can be imported into augmented reality headsets to provide an immersive experience for the surgeon for operative planning.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Artificial intelligence (AI) is poised to revolutionize medicine, and radiology is a natural gateway due to the inherent digital nature of radiology data. AI can be broadly defined as using computers to perform tasks typically associated with human intelligence. Machine learning, a branch of AI, enables the extraction of meaningful patterns from examples rather than through explicit programming. Deep learning (DL), a subfield of machine learning first developed in the 1950s, utilizes networks of interconnected nodes that process input data and adjust the network weights to minimize prediction errors. Recent developments in powerful parallel computing hardware, the availability of large training data, and improved network architectures have notably enhanced the performance of deep learning, which has significant potential for clinical translation.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Radiomics converts imaging data into high-dimensional features that can be used to characterize spatial heterogeneity inherent in disease processes. The features of radiomics can be classified into signal intensity, shape, and texture. Signal intensity (first-order) features are derived from histograms of individual voxel signal intensities, providing measures of central tendency and shape of the distribution. Shape features are extracted from the three-dimensional surface of the region of interest. Texture features are calculated in three dimensions, considering the correlation of signal intensities of adjacent voxels. In addition, feature extraction may be performed after applying a secondary filter, such as a wavelet or Gaussian filter.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “AI can theoretically function as ‘second readers’ to improve radiologists’ sensitivity in the detection of small tumors, which potentially can be cured with surgical resection. A preliminary study by Liu et al. showed promising results suggesting that DL could accurately differentiate CT scans of patients with PDAC from CT scans of healthy controls. More recently, Chen et al. developed a DL tool that differentiated CT scans of patients with PDAC vs. healthy controls with 89.9% sensitivity, 95.9% specificity, and 93.4% accuracy in the local test set. They validated this DL tool on a Taiwanese nationwide external validation set and achieved 89.7% sensitivity, 92.8% specificity, and 91.4% accuracy. Also, Park et al. developed a different DL tool that achieved high sensitivity comparable to radiologists in the detection of not only pancreatic solid masses (98–100%) but also cystic masses 1.0 cm or larger (sensitivity 92–93%), bringing us closer to a universal pancreatic neoplasm detector.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Other studies have used radiomics to facilitate the detection of PDAC, demonstrating that radiomics signatures from PDAC were distinct from the background pancreas. More impressively, radiomics signatures could identify subtle differences in prediagnostic CT scans obtained with a median of 386 days before PDAC diagnosis, with 95.5% sensitivity, 90.3% specificity, and 92% accuracy. If these promising results are validated in future studies, radiologists will be able to diagnose patients significantly earlier at lower disease stages. In this scenario, a higher proportion of newly diagnosed patients will be eligible for curative surgical resection, which will have a significant positive impact on patient outcomes.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Researchers have also used radiomics features to predict the development of liver metastases after PDAC resection in multiple studies. Zambirinis et al. analyzed 254 radiomics features from the liver from preoperative CTs in 688 patients with resected PDAC and the radiomics model identified patients at risk for early (< 6 months) liver metastases with an AUC of 0.71. Huang et al. extracted 3906 radiomics features from the pancreatic tumor from preoperative MRIs in 204 patients with resected PDAC, and the radiomics model achieved 75.0% sensitivity, 82.2% specificity, and an AUC of 0.815 in predicting the development of liver metastases. We speculate that radiologic features from both the primary tumor and the liver parenchyma are important in predicting future liver metastases. Future studies should incorporate features from both the tumor and the liver, in combination with clinical features, to optimize the prediction of liver metastases.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Preliminary studies using emerging technologies such as advanced visualization and AI have revealed the potential of these tools to improve the initial diagnosis and staging of patients with PDAC. However, there remain several limitations. Most of these studies have been single-center retrospective studies, and their promising results should be validated in future multicenter prospective studies. Secondly, one of the major criticisms of AI is its ‘blackbox’ nature, making it difficult for clinicians to decipher the rationale behind AI predictions. Explainable or ‘glassbox’ AI is an active area of research that aims to render AI models more easily understandable and may help improve their clinical acceptance. Thirdly, these tools should be integrated seamlessly into the workflow to ensure widespread clinical implementation.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063
- The Declaration of Helsinki (DoH) is a statement of principles intended to guide the ethical conduct of health and medical research on a global scale, developed and endorsed by the World Medical Association first in 1964. Inspired by the demand that medical research with human participants be conducted to the highest ethical standards in the wake of Nazi war crimes, the DoH looks beyond the prevention of egregious harms to the promotion of ethical science more generally. The DoH has been amended to reflect the dynamic nature of both ethics and science on 9 occasions since 1964 (including notes of clarification), with a 10th amendment published in 2024. In the 11 years between the prior amendment in 2013 and the most recent 2024 amendment, medical science has also evolved in important ways. In this Viewpoint, I highlight some important updates to the Do Hand comment on its relevance in the context of the growing influence of artificial intelligence (AI) in health and medical science.
The Revised Declaration of Helsinki—Considerations for the Future of Artificial Intelligence in Health andMedical Research
James A. Shaw, PT, PhD
JAMA October 19, 2024. doi:10.1001/jama.2024.22074 - “The renewed declaration has made important changes, but the contexts of health and medical science are also changing rapidly. Despite the hype and lack of clarity that often accompany the discussions of AI as a class of technologies, AI applications have also shown promise for a range of health care tasks.3 Furthermore, AI technologies hold potential to enhance the conduct of research in a variety of ways, ranging from accelerating drug discovery to enhancing the efficiency of clinical trials. These documented advances, alongwith massive financial investments, have led toawidespread interest in using AI techniques across health and medical science. Three issues should inform the implementation of and future revisions to the DoH, and other forms of research ethics guidance, on AI in health and medical science”
The Revised Declaration of Helsinki—Considerations for the Future of Artificial Intelligence in Health andMedical Research
James A. Shaw, PT, PhD
JAMA October 19, 2024. doi:10.1001/jama.2024.22074 - “There are a number of divergent positions on the matter of protecting privacy vs sharing data in the name of solidarity, each needing to navigate complex social and legal expectations in specific jurisdictions. Although the global dialogue in which the DoH is embedded is hugely important on these matters, and the Declaration of Taipei that specifically addresses the issue of ethics and big data is an essential contribution, these statements have yet to catalyze a shared vision on the governance of health related data that supports public goods.”
The Revised Declaration of Helsinki—Considerations for the Future of Artificial Intelligence in Health andMedical Research
James A. Shaw, PT, PhD
JAMA October 19, 2024. doi:10.1001/jama.2024.22074 - Second, the implementation of the DoH must acknowledge variability inAI literacy both within and between jurisdictions. A survey published in 2024 on knowledge and perceptions of AI in 21 countries found discrepancies between respondents’ relatively high confidence in their knowledge of AI and their relatively low knowledge of specific outputs of AI, such as deepfakes.Furthermore, it is widely acknowledged that limited expertise in AI is a challenge for institutional review boards and research ethics committees around the world, interfering with their abilities to make informed decisions about research involving AI. The implementation of the DoH, and the Declaration of Taipei on which it relies for guidance on research with large datasets, will depend in profound ways on whether and how the public and professionals adequately understand AI and its uses in health care.
The Revised Declaration of Helsinki—Considerations for the Future of Artificial Intelligence in Health andMedical Research
James A. Shaw, PT, PhD
JAMA October 19, 2024. doi:10.1001/jama.2024.22074 - “Third, the implementation of the DoH will need to acknowledge the lack of clarity around present and future harms of health related AI. Setting aside dialogue regarding hypothetical doomsday scenarios and existential threats, the actual present-day harms that may accrue, especially to communities already structurally marginalized as a consequence of these technologies, remain poorly understood. Beyond some well-known examples in health care, researchers are beginning to identify “hidden” harms of AI development and deployment that should also be considered in discussions regarding the ethics of health and medical research involving AI. For example, the DoH does not explicitly encourage researchers to avoid conflicts of interest, but in a field in which industry is immensely powerful and well-resourced, such conflicts are salient to the ethics of medical research.”
The Revised Declaration of Helsinki—Considerations for the Future of Artificial Intelligence in Health andMedical Research
James A. Shaw, PT, PhD
JAMA October 19, 2024. doi:10.1001/jama.2024.22074 - “When implemented well, the DoH has potential to affect health and medical science involving AI in positive ways. However, declarations are only as good as the capacity and commitment of people and institutions who use them. The capacity to meaningfully adopt the principles of the DoH in the era of AI requires creativity, foresight, and the resources necessary to keep abreast of progress in the field and the rapidly evolving knowledge of ethical issues that accompany these advances.”
The Revised Declaration of Helsinki—Considerations for the Future of Artificial Intelligence in Health and Medical Research
James A. Shaw, PT, PhD
JAMA October 19, 2024. doi:10.1001/jama.2024.22074 
- “ Rajpurkar emphasizes the importance of understanding the “data generation process,” including the artifacts and biases baked into data, which is illustrated by a specific example where an AI model exploited metadata rather than clinically relevant features. Dr. Rajpurkar addresses the urgent need for more open and accessible medical data with his initiative on Medical AI Data for All (MAIDA). We also examine the changing role of clinicians in an AI-augmented health care system, and discuss a collaborative approach where human expertise guides AI development and implementation. Looking ahead, we envision a future where AI systems generate comprehensive medical reports and engage in natural language interactions, while emphasizing the need for ongoing focus on safety, efficacy, and equitable access. ”
Pixels and Pitfalls: Building Robust Artificial Intelligence for Medical Imaging
Pranav Rajpurkar , Andrew L. Beam , Arjun K. Manrai
NEJM AI 2024; 1 (10) - “Collaboration between clinicians and AI researchers remains crucial. Clinicians play a key role in problem definition, ensuring that AI systems address real clinical needs and account for the full complexity of medical tasks. Interdisciplinary teams of young researchers and senior clinical collaborators have had substantial impact.”
Pixels and Pitfalls: Building Robust Artificial Intelligence for Medical Imaging
Pranav Rajpurkar , Andrew L. Beam , Arjun K. Manrai
NEJM AI 2024; 1 (10) - “The siloed nature of medical data and privacy concerns are major obstacles, potentially holding us back by a decade or more. We urgently need a paradigm shift toward open, accessible medical data for research, balancing privacy concerns with the immense potential for improving health care through AI. Initiatives such as MAIDA aim to create a global, standardized dataset for evaluating AI models with the goal of accelerating model development, similar to that which has been done in the general machine learning community.”
Pixels and Pitfalls: Building Robust Artificial Intelligence for Medical Imaging
Pranav Rajpurkar , Andrew L. Beam , Arjun K. Manrai
NEJM AI 2024; 1 (10) - “As AI capabilities grow, the role of clinicians will inevitably evolve. Rather than fearing replacement, health care professionals should embrace AI as a powerful tool, focusing on problem definition and the nuanced aspects of patient care that machines cannot replicate. The future likely lies in seamless human–AI collaboration, with clinicians guiding the development, implementation, and use of these technologies.”
Pixels and Pitfalls: Building Robust Artificial Intelligence for Medical Imaging
Pranav Rajpurkar , Andrew L. Beam , Arjun K. Manrai
NEJM AI 2024; 1 (10) - “ While artificial general intelligence (AGI) and artificial superintelligence (ASI) remain speculative, the possibility of capabilities that meet or surpass expert levels in areas such as treatment planning, clinical reasoning, and cognitive empathy merits consideration and debate in anticipation of the day when these capabilities become viable. Current ethical debates about AI often center on immediate concerns like data bias, but the progression toward AGI and ASI presents a profound and novel challenge: these systems might develop ethical frameworks that fundamentally differ from human-derived ethics. Planning for them must anticipate these changes, ensuring that their ethical paradigms uphold human values. Given the transformative potential of AGI and ASI, a multidisciplinary dialogue among medical professionals, policy makers, and technology experts is essential to prepare for these advancements.”
If Machines Exceed Us: Health Care at an Inflection Point
Eyal Klang
NEJM AI 2024; 1 (10
- “Histopathology image evaluation is indispensable for cancer diagnoses and subtype classification. Standard artificial intelligence methods for histopathology image analyses have focused on optimizing specialized models for each diagnostic task. Although such methods have achieved some success, they often have limited generalizability to images generated by different digitization protocols or samples collected from different populations3. Here, to address this challenge, we devised the Clinical Histopathology Imaging Evaluation Foundation (CHIEF) model, a general purpose weakly supervised machine learning framework to extract pathology imaging features for systematic cancer evaluation. CHIEF leverages two complementary pretraining methods to extract diverse pathology representations: unsupervised pretraining for tile-level feature identification and weakly supervised pretraining for whole-slide pattern recognition.”
A pathology foundation model for cancer diagnosis and prognosis prediction.
Wang X, Zhao J, Marostica E, et al.
Nature. 2024 Sep 4. doi: 10.1038/s41586-024-07894-z. Epub ahead of print. PMID: 39232164. - “We developed CHIEF using 60,530 whole-slide images spanning 19 anatomical sites. Through pretraining on 44 terabytes of high resolution pathology imaging datasets, CHIEF extracted microscopic representations useful for cancer cell detection, tumour origin identification, molecular profile characterization and prognostic prediction. We successfully validated CHIEF using whole-slide images from 32 independent slide sets collected from 24 hospitals and cohorts internationally. Overall, CHIEF outperformed the state-of-the-art deep learning methods by up to 36.1%, showing its ability to address domain shifts observed in samples from diverse populations and processed by different slide preparation methods. CHIEF provides a generalizable foundation for efficient digital pathology evaluation for patients with cancer.”
A pathology foundation model for cancer diagnosis and prognosis prediction.
Wang X, Zhao J, Marostica E, et al.
Nature. 2024 Sep 4. doi: 10.1038/s41586-024-07894-z. Epub ahead of print. PMID: 39232164. - ”We established the CHIEF model, a general-purpose machine learning framework for weakly supervised histopathological image analyses. Unlike commonly used self-supervised feature extractors, CHIEF leveraged two types of pretraining procedure: unsupervised pretraining on 15 million unlabelled tile images and weakly supervised pretraining on more than 60,000 WSIs. Tile-level unsupervised pretraining established a general feature extractor for haematoxylin–eosin-stained histopathological images collected from heterogeneous publicly available databases, which captured diverse manifestations of microscopic cellular morphologies. Subsequent WSI-level weakly supervised pretraining constructed a general-purpose model by characterizing thesimilarities and differences between cancer types.”
A pathology foundation model for cancer diagnosis and prognosis prediction.
Wang X, Zhao J, Marostica E, et al.
Nature. 2024 Sep 4. doi: 10.1038/s41586-024-07894-z. Epub ahead of print. PMID: 39232164. - CHIEF consistently attained superior performance in a variety of cancer identification tasks using either biopsy or surgical resection slides CHIEF achieved a macro-average area under the receiver operating characteristic curve (AUROC) of 0.9397 across 15 datasets representing 11 cancer types, which is approximately 10% higher than that attained by DSMIL (a macro-average AUROC of 0.8409), 12% higher than that of ABMIL (a macro-average AUROC of 0.8233) and 14% higher than that of CLAM (a macro-average AUROC of 0.8016). In all five biopsy datasets collected from independent cohorts, CHIEF possessed AUROCs of greater than 0.96 across several cancer types, including oesophagus (CUCH-Eso), stomach (CUCH-Sto), colon (CUCH-Colon) and prostate (Diagset-B and CUCH-Pros). On independent validation with seven surgical resection slide sets spanning five cancer types (that is, colon (Dataset-PT), breast (DROID-Breast), endometrium (SMCH-Endo and CPTAC-uterine corpus endometrial carcinoma (UCEC)), lung (CPTAC-lung squamous cell carcinoma (LUSC)) and cervix (SMCH-Cervix and TissueNet)), CHIEF attained AUROCs greater than 0.90
A pathology foundation model for cancer diagnosis and prognosis prediction.
Wang X, Zhao J, Marostica E, et al.
Nature. 2024 Sep 4. doi: 10.1038/s41586-024-07894-z. Epub ahead of print. PMID: 39232164. 
- “The CHIEF framework successfully characterized tumour origins, predicted clinically important genomic profiles, and stratified patients into longer-term survival and shorter-term survival groups. Furthermore, our approach established a general pathology feature extractor capable of a wide range of prediction tasks even with small sample sizes. Our results showed that CHIEF is highly adaptable to diverse pathology samplesobtained from several centres, digitized by various scanners, and obtained from different clinical procedures (that is, biopsy and surgicalresection). This new framework substantially enhanced model generalizability, a critical barrier to the clinical penetrance of conventional computational pathology models.”
A pathology foundation model for cancer diagnosis and prognosis prediction.
Wang X, Zhao J, Marostica E, et al.
Nature. 2024 Sep 4. doi: 10.1038/s41586-024-07894-z. Epub ahead of print. PMID: 39232164. - “In conclusion, CHIEF is a foundation model useful for a wide range of pathology evaluation tasks across several cancer types. We have demonstrated the generalizability of this foundation model across several clinical applications using samples collected from 24 hospitals and patient cohorts worldwide. CHIEF required minimal image annotations and extracted detailed quantitative features from WSIs, which enabled systematic analyses of the relationships among morphological patterns, molecular aberrations and important clinical outcomes. Accurate, robust and rapid pathology sample assessment provided by CHIEF will contribute to the development of personalized cancer management.”
A pathology foundation model for cancer diagnosis and prognosis prediction.
Wang X, Zhao J, Marostica E, et al.
Nature. 2024 Sep 4. doi: 10.1038/s41586-024-07894-z. Epub ahead of print. PMID: 39232164.
- Objectives Large language models like GPT-4 have demonstrated potential for diagnosis in radiology. Previous studies investigating this potential primarily utilized quizzes from academic journals. This study aimed to assess the diagnostic capabilities of GPT-4-based Chat Generative Pre-trained Transformer (ChatGPT) using actual clinical radiology reports of brain tumors and compare its performance with that of neuroradiologists and general radiologists.
Methods We collected brain MRI reports written in Japanese from preoperative brain tumor patients at two institutions from January 2017 to December 2021. The MRI reports were translated into English by radiologists. GPT-4 and five radiologists were presented with the same textual findings from the reports and asked to suggest differential and final diagnoses. The pathological diagnosis of the excised tumor served as the ground truth. McNemar’s test and Fisher’s exact test were used for statistical analysis.
Comparative analysis of GPT-4-based ChatGPT’s diagnostic performance with radiologists using real-world radiology reports of brain tumors
Yasuhito Mitsuyama et al.
Eur Radiol. 2024 Aug 28. doi: 10.1007/s00330-024-11032-8. Online ahead of print. - Results In a study analyzing 150 radiological reports, GPT-4 achieved a final diagnostic accuracy of 73%, while radiologists’ accuracy ranged from 65 to 79%. GPT-4’s final diagnostic accuracy using reports from neuroradiologists was higher at 80%, compared to 60% using those from general radiologists. In the realm of differential diagnoses, GPT- 4’s accuracy was 94%, while radiologists’ fell between 73 and 89%. Notably, for these differential diagnoses, GPT-4’s accuracy remained consistent whether reports were from neuroradiologists or general radiologists.
Conclusion GPT-4 exhibited good diagnostic capability, comparable to neuroradiologists in differentiating brain tumors from MRI reports. GPT-4 can be a second opinion for neuroradiologists on final diagnoses and a guidance tool for general radiologists and residents.
Clinical relevance statement This study evaluated GPT-4-based ChatGPT’s diagnostic capabilities using real-world clinical MRI reports from brain tumor cases, revealing that its accuracy in interpreting brain tumors from MRI findings is competitive with radiologists.
Comparative analysis of GPT-4-based ChatGPT’s diagnostic performance with radiologists using real-world radiology reports of brain tumors
Yasuhito Mitsuyama et al.
Eur Radiol. 2024 Aug 28. doi: 10.1007/s00330-024-11032-8. Online ahead of print. 
- This study is the first attempt to evaluate GPT-4’s ability to interpret actual clinical radiology reports, rather than from settings like image diagnosis quizzes. The majority of previous research suggested the utility of GPT-4 in diagnostics, but these relied heavily on hypothetical environments such as quizzes from academic journals or examination questions. This approach can lead to a cognitive bias since the individuals formulating the imaging findings or exam questions also possess the answers. In these simulated scenarios, there’s also a propensity to leave out minor findings. Such minor findings, while often deemed insignificant in an experimental setup, are frequently encountered in real-world clinical practice and can have implications for diagnosis. In contrast, our study deviates from this previous methodology by using actual clinical findings, generated in a state of diagnostic uncertainty. This approach facilitates a more robust and practical evaluation of GPT-4’s accuracy, keeping in mind its potential applications in real-world clinical settings.
Comparative analysis of GPT-4-based ChatGPT’s diagnostic performance with radiologists using real-world radiology reports of brain tumors
Yasuhito Mitsuyama et al.
Eur Radiol. 2024 Aug 28. doi: 10.1007/s00330-024-11032-8. Online ahead of print. - “There are several limitations. This study only used the wording of actual clinical radiology reports and did not evaluate the effect of including other information such as patient history and the image itself, meaning the radiologists’ performance might not match their real-world diagnostic abilities. Furthermore, recent advancements in large language models have enabled the input of not only text but also images. Evaluating the performance of large language models that combine both radiology report texts and images could provide deeper insights into their potential usefulness in radiology diagnostics.”
Comparative analysis of GPT-4-based ChatGPT’s diagnostic performance with radiologists using real-world radiology reports of brain tumors
Yasuhito Mitsuyama et al.
Eur Radiol. 2024 Aug 28. doi: 10.1007/s00330-024-11032-8. Online ahead of print. - Aims and objectives: This study evaluates the accuracy of two AI language models, ChatGPT 4.0 and Google Gemini (as of August 2024), in answering a set of 79 text-based pediatric radiology questions from “Pediatric Imaging: A Core Review.” Accurate interpretation of text and images is critical in radiology, making AI tools valuable in medical education.
Methods: The study involved 79 questions selected from a pediatric radiology question set, focusing solely on text-based questions. ChatGPT 4.0 and Google Gemini answered these questions, and their responses were evaluated using a binary scoring system. Statistical analyses, including chi-square tests and relative risk(RR) calculations, were performed to compare the overall and subsection accuracy of the models.
Results: ChatGPT 4.0 demonstrated superior accuracy, correctly answering 83.5% (66/79) of the questions, compared to Google Gemini's 68.4% (54/79), with a statistically significant difference (p=0.0255, RR=1.221). No statistically significant differences were found between the models within individual subsections, with pvalues ranging from 0.136 to 1.
Conclusion: ChatGPT 4.0 outperformed Google Gemini in overall accuracy for text-based pediatric radiology questions, highlighting its potential utility in medical education. However, the lack of significant differences within subsections and the exclusion of image-based questions underscore the need for further research with larger sample sizes and multimodal inputs to fully assess AI models' capabilities in radiology.
Comparative Accuracy of ChatGPT 4.0 and Google Gemini in Answering Pediatric Radiology Text-Based Questions.
Abdul Sami M, Abdul Samad M, Parekh K, et al.
(October 05, 2024) Cureus 16(10): e70897. DOI 10.7759/cureus.70897 - “The limitations of this study include the exclusion of image-based questions, utilizing a single subspecialized question set within pediatric radiology, and the low sample size for certain sections, all of which rendered it difficult to determine statistically significant differences in accuracy where they might exist. Moreover, this study did not provide a longitudinal assessment of the AI models' performance, which is relevant given the rapid pace of improvement in these models. The absence of evaluation of the models’ performances might evolve over time limiting our understanding of their potential and reliability in clinical settings. Additionally, the lack of comparison to a human radiologist's performance further limits the ability to contextualize these AI models. The implications of these limitations include an incomplete assessment of the models' true performance, particularly given that this study was conducted with a controlled question set rather than real-life patient scenarios involving actual patient data.”
Comparative Accuracy of ChatGPT 4.0 and Google Gemini in Answering Pediatric Radiology Text-Based Questions.
Abdul Sami M, Abdul Samad M, Parekh K, et al.
(October 05, 2024) Cureus 16(10): e70897. DOI 10.7759/cureus.70897 - “The educational implications of this study are significant in fields such as general or specialized medical education. AI models like ChatGPT and Google Gemini could be utilized in training students with both text and image-based questions. For example, ChatGPT 4.0’s current performance suggests that it could be used to better understand case studies or accurately summarize educational content. The potential of AI in medical education extends beyond radiology, with applications in other specialties where analyzing text is critical. Integrating AI models could further support medical residents or physicians inbreaking down complex topics or creating personalized learning experiences.”
Comparative Accuracy of ChatGPT 4.0 and Google Gemini in Answering Pediatric Radiology Text-Based Questions.
Abdul Sami M, Abdul Samad M, Parekh K, et al.
(October 05, 2024) Cureus 16(10): e70897. DOI 10.7759/cureus.70897 - “The study demonstrates a statistically significant difference in accuracy between ChatGPT 4.0 and Google Gemini when answering standardized radiology-related questions, with ChatGPT 4.0 achieving an accuracy rate of 83.5% compared to Google Gemini's 68.4%, suggesting that ChatGPT 4.0 may be more reliable for certain text-based tasks in medical education. However, the observed variability across different pediatric radiology subspecialties and the exclusion of image-based questions indicate that both AI models have distinct strengths and weaknesses that should be carefully considered. The findings emphasize the potential role of AI in enhancing medical education and diagnostic capabilities, particularly in radiology, while also underscoring the need for responsible integration of these technologies to complement human expertise.”
Comparative Accuracy of ChatGPT 4.0 and Google Gemini in Answering Pediatric Radiology Text-Based Questions.
Abdul Sami M, Abdul Samad M, Parekh K, et al.
(October 05, 2024) Cureus 16(10): e70897. DOI 10.7759/cureus.70897
- Objectives: This research aimed to assess the value of radiomics combined with multiple machine learning algorithms in the diagnosis of pancreatic ductal adenocarcinoma (PDAC) lymph node (LN) metastasis, which is expected to provide clinical treatment strategies. Conclusions: Combining radiomics and machine learning algorithms demonstrated the potential for identifying the LN metastasis of PDAC. As a non-invasive and efficient preoperative prediction tool, it can be beneficial for decision-making in clinical practice.
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317 - Methods: A total of 128 patients with pathologically confirmed PDAC and who underwent surgical resection were randomized into training (n=93) and validation (n=35) groups. This study incorporated a total of 13 distinct machine learning algorithms and explored 85 unique combinations of these algorithms. The area under the curve (AUC) of each model was computed. The model with the highest mean AUC was selected as the best model which was selected to determine the radiomics score (Radscore). The clinical factors were examined by the univariate and multivariate analysis, which allowed for the identification of factors suitable for clinical modeling. The multivariate logistic regression was used to create a combined model using Radscore and clinical variables. The diagnostic performance was assessed by receiver operating characteristic curves, calibration curves, and decision curve analysis (DCA).
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317 - Radiomics provides a large amount of medical image information that can reveal hidden features of diseases that are invisible to the naked eye. Previous research has demonstrated the efficacy of integrating radiomics and machine learning (ML) techniques in many applications. Several researchers have constructed multiphasic contrast-enhanced CT (CECT) radiomics models to evaluate the preoperative LN status of PDAC. These studies indicated that radiomics models have significant potential in predicting pancreatic cancer with lymph node metastasis (LNM). However, the researchers predominantly chose modeling algorithms based on their preferences and limitations in knowledge. It is imperative that evidence be provided to select appropriate models for solving clinical problems. To the best of our knowledge, no study has explored the use of radiomics combined with multiple ML algorithms in thepreoperative identification of the LN status in PDAC
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317 - “Although CT is the most commonly used method to evaluate pancreatic cancer resectability, Imai et al. indicated that loworder data such as LN diameter and volume measured from CT image cannot reflect obvious differences between patients with pancreatic cancer with and without LN metastasis. This may explain why visual assessment of CT scans has a low efficiency for detecting LN metastases among patients with pancreatic cancer. Radiomics, also referred to as a “whole-tumor virtual biopsy technique,” enables the extraction of numerous image features from the entire tumor, reflecting its heterogeneity and characteristics. A biopsy examines only a portion of the tumor tissue, which cannot comprehensively assess the intratumor heterogeneity before surgery. In contrast, radiomics can non-invasively reflect comprehensive information about the entire tumor. This approach has promise for enhancing diagnostic capabilities and facilitating the development of personalized treatment approaches.”
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317 - “In summary, of the 233 radiomics models examined, the model built by applying AP+VP radiomics features and a combination of Lasso-Logistic algorithm had the most favorable performance in both the training and validation cohorts. Our investigation showed that integrating the AP+VP-Radscore with clinical parameters yielded the best performance. This combined model has the potential to serve as an accurate and non-invasive instrument for forecasting LN metastasis of PDAC, hence facilitating clinical decision-making.”
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317
Kidney
- Mixed Epithelial and Mesenchymal Neoplasms: Facts
- Consists of mixed epithelial and stromal tumors (MEST) and cystic nephromas
- MEST was previously called
--- Leiomyomatous renal hamartoma
--- Multilocular cyst with ovarian stroma
--- Cystic hamartoma of the renal pelvis
--- Adult mesoblastic nephroma - MEST (Mixed epithelial and stromal tumor of the kidney)
Mixed epithelial and stromal tumor (MEST) is a rare benign renal neoplasm composed of epithelial and stromal component and shows variable solid and cystic areas. Along with adult cystic nephroma, it is included in the family of mixed epithelial and stromal tumor in the 2022 WHO classification of renal neoplasms. Most tumors are benign; however, they may recur if incompletely excised. Rare examples of malignant transformation of the epithelial or stromal component have been reported. - MEST (Mixed epithelial and stromal tumor of the kidney)
- The majority of diagnosed tumors occurred in women at menopausal age and females affected ten times more than males
- age of diagnosis varied from 17 to 78 years old and most commonly being in 5th decade.
- clinical presentation of MEST may include flank pain, blood in urine, urinary tract infections and abdominal mass - MEST (Mixed epithelial and stromal tumor of the kidney)
- MEST usually presents as unilateral and solitary kidney lesions. Usually the tumor is well-demarcated, triphasic CT scan commonly reveals complex cystic lesions, described as III or IV in the Bosniak classification, with solid components and contrast enhancement
- The majority of MESTs are benign lesions, showing no recurrence nor distant metastases. - MEST (Mixed epithelial and stromal tumor of the kidney)
Mixed epithelial and stromal tumor (MEST) of the kidney is a rare, typically benign lesion that occurs predominantly in perimenopausal women. At computed tomography (CT), it typically manifests as a multiloculated cystic renal mass with a variable proportion of solid and cystic components and containing internal septa that demonstrate heterogeneous and delayed contrast material enhancement. MEST may mimic a variety of benign and malignant renal lesions, such as adult cystic nephroma, complex renal cyst, and cystic renal cell carcinoma. The preoperative diagnosis of MEST can be problematic, and most cases are treated surgically. - By definition, ACN is a multilocular cystic lesion with no solid area at gross examination and with cystic septa less than 5 mm in thickness at microscopy, whereas MEST is a cystic or partially cystic mass with solid areas at gross examination and cystic septa greater than or equal to 5 mm (15). However, it is difficult to differentiate between MEST and ACN on the basis of imaging findings alone. Because both ACN and MEST are benign cystic renal lesions, the precise preoperative radiologic diagnosis may not be critically important, as long as the lesion in question can be differentiated from malignant lesions such as RCC and transitional cell carcinoma.
Mixed Epithelial and Stromal Tumor of theKidney: Radiologic-Pathologic Correlation
Linda C. Chu • Ralph H. Hruban • Karen M. Horton • Elliot K. Fishman
RadioGraphics 2010; 30:1541–1551 - Given its variable appearance, MEST may mimic an array of cystic renal lesions, including adult cystic nephroma (ACN), cystic renal cell carcinoma (RCC), complex cyst, multicystic dysplastic kidney (MDK), an obstructed duplicated renal collecting system, and renal abscess.
Mixed Epithelial and Stromal Tumor of the Kidney: Radiologic-Pathologic Correlation
Linda C. Chu • Ralph H. Hruban • Karen M. Horton • Elliot K. Fishman
RadioGraphics 2010; 30:1541–1551 - “Because most MESTs represent Bosniak category III or IV lesions, a cystic RCC is an important consideration in the differential diagnosis. Cystic change occurs in up to 15% of RCCs. The spectrum of cystic RCC includes multilocular cystic RCC, RCC arising from a preexisting benign cyst, and cystic degeneration of a previously solid RCC. Compared with MEST, cystic RCC tends to have thicker, irregularly enhancing septa and enhancing nodular or solid com-ponents. Complex renal cyst is also part of the spectrum of cystic renal lesions. Complex renal cysts contain thin, nonenhancing internal septa with thin mural calcifications and no mural nodularity.”
Mixed Epithelial and Stromal Tumor of the Kidney: Radiologic-Pathologic Correlation
Linda C. Chu • Ralph H. Hruban • Karen M. Horton • Elliot K. Fishman
RadioGraphics 2010; 30:1541–1551
- “Erdheim-Chester disease (ECD) is a rare multiorgan histiocytosis with diverse clinical manifestations. Initially described as “lipoid granulomatosis” by William Chester and Jakob Erdheim in 1930 , it was named after its founders by Henry Jaffe in 1972. ECD was previously considered inflammatory histiocytosis, but the discovery of activating mutations in the mitogen-activated protein kinase (MAPK) signaling pathway has established ECD as a histiocytic neoplasm, although inflammation contributes to organ damage . Hence, ECD was included in the “histiocytic and dendritic cell neoplasms” category in the 2016 World Health Organization classification of hematopoietic tumors. Similarly, ECD along with Langerhans cell histiocytosis (LCH) has been included in the Langerhans group in the 2016 revised Histiocyte Society classification of histiocytosis, as both entities share mutations in the MAPK pathway.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 - - The pattern of bilateral symmetric metadiaphyseal regions of osteosclerosis, especially around the knees, is virtually pathognomonic for ECD. These regions reveal intense uptake at scintigraphy and FDG PET/ CT, again a finding nearly exclusive to ECD.
- ECD in the retroperitoneum manifests as infiltrative soft-tissue masses in bilateral perirenal (hairy kidneys) and posterior pararenal spaces.
- Myocardial infiltration has a predilection for the right atrium and atrioventricular sulcus and manifests as a focal mural mass (pseudotumor).
- Chest radiographic findings may be normal. Pulmonary involvement is, however, seen at CT in 50% of patients and often shows smooth interlobular septal thickening. The histiocytic infiltrate is perilymphatic in distribution and causes smooth thickening of the interlobular septa, peribronchovascular interstitium, and visceral pleura.
- Intracranial lesions in ECD, including intra-axial, meningeal, perivascular, and pituitary infundibulum lesions, are rarely isolated, and patients who harbor these lesions almost always have osteosclerosis of the paranasal sinus walls and/or orbital disease.
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011- “Retroperitoneal and renal involvement in ECD is seen in 50%–60% of patients. Although patients with renal involvement are frequently asymptomatic, they may present with abdominal pain and dysuria . ECD in the retroperitoneum manifests as infiltrative soft-tissue masses in bilateral perirenal (hairy kidneys) and posterior pararenal spaces. On CT images, this soft tissue is isoattenuating relative to skeletal muscle and shows mild postcontrast enhancement. At MRI, it is iso- to hypointense relative to skeletal muscle on T1- and T2-weighted images and shows mild postcontrast enhancement. Because the normal renal parenchyma shows intense uptake of FDG, the sensitivity of FDG PET/CT for detection of retroperitoneal involvement, in particular perinephric ECD, is not as high as contrast-enhanced CT.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 - “Perinephric soft-tissue infiltrates, other than ECD, can also be seen in lymphoma, retroperitoneal fibrosis (RPF), RDD, amyloid, and extramedullary hematopoeisis (EMH) . ECD and lymphoma may involve the perinephric and retroaortic spaces and show uptake at FDG PET/ CT. Contiguous involvement of the kidneys or perinephric space from retroperitoneal masses or lymph nodes is a common pattern of renal lymphoma. Contrarily, ECD rarely affects the lymph nodes.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 - “Although ECD and RPF cause hydronephrosis, the former is epicentered around the renal hila and proximal ureters, whereas the latter is epicentered around aortic bifurcation and hence affects the distal ureters. As RPF extends more superiorly, it lies anterior and lateral to the aorta and rarely infiltrates the renal hila and perinephric spaces. Immunoglobulin G4 (IgG4)–related RPF may also show pancreatic involvement. Renal involvement in RDD most commonly manifests as bilateral hilar masses.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011- “ECD is a multiorgan inflammatory histiocytic neoplasm that most frequently affects the skeletal system and produces virtually pathognomonic osteosclerosis around the knees that shows increased radiotracer uptake on nuclear medicine studies. Other typical imaging findings comprise perinephric infiltrates (hairy kidneys), periaortic soft tissue (coated aorta), and a right atrial pseudotumor. However, radiologic findings are interpreted in conjunction with clinical and histologic features to establish the diagnosis. In addition to aiding in diagnosis, imaging also helps to determine the extent of disease, target sites for tissue sampling, assess treatment response,and predict the prognosis.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011
- IMPORTANCE Renal cell carcinoma (RCC) is a common malignancy, with an estimated 434 840 incident cases worldwide in 2022. In the US, it is the sixth most common cancer among males and ninth among females.
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010. - OBSERVATIONS Clear cell RCC is the most common histologic subtype (75%-80% of cases) and is characterized by inactivation of the von Hippel Lindau (VHL) tumor suppressor gene. Many patients (37%-61%) are diagnosed with RCC incidentally on an abdominal imaging study such as ultrasound or computed tomographic scan, and 70%of patients have stage I RCC at diagnosis. Although its incidence has increased approximately 1% per year from 2015 through 2019, the mortality rate of RCC has declined about 2% per year in the US from 2016through 2020. Patients with a solid renal mass or complex cystic renal mass should be referred to urology.
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “Treatment options for RCC confined to the kidney include surgical resection with partial or radical nephrectomy, ablative techniques (eg, cryoablation, radiofrequency ablation, radiation), or active surveillance for some patients (especially those with renal masses <2 cm). For patients with renal masses less than 4 cm in size (48%of patients), partial nephrectomy can result in a 5-year cancer-specific survival of more than 94%. For advanced or metastatic RCC, combinations of immune checkpoint inhibitors or the combination of immune checkpoint inhibitors with tyrosine kinase inhibitors are associated with tumor response of 42%to 71%, with a median overall survival of 46 to 56 months.”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - CONCLUSIONS AND RELEVANCE RCC is a common malignancy that is often diagnosed incidentally on an abdominal imaging study. Seventy percent of patients are diagnosed with stage I RCC and 11% of patients with stage IV. First-line treatments for early-stage RCC are partial or radical nephrectomy, which can result in 5-year cancer-specific survival of more than 94%, ablative techniques, or active surveillance. New treatment options for patients with metastatic RCC include immune checkpoint inhibitors and tyrosine kinase inhibitors.
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “Globally, there were 434,840 incident cases of RCC in 2022. RCC is most common in Europe, Oceania, and North America,2 potentially in part due to higher incidental detection rates of renal masses on abdominal imaging studies. It is the 15th most common cause of cancer-related death worldwide, with more than 179000 deaths reported in 2020. Despite the increasing incidence of about 1% per year from 2015 through 2019, the mortality rate from RCC has declined about 2% per year in the US from 2016 through 2020, partially due to improvements in therapy.”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “The 3 dominant histological subtypes are clear cell (75%- 80%), papillary (10%-15%), and chromophobe (5%).15 Clear cell RCC is named for its golden yellow clear cytoplasm.15 Loss of the von Hippel Lindau (VHL) tumor suppressor gene occurs in up to 90% of clear cell RCC tumors, with VHL inactivation leading to activation of hypoxia and angiogenesis pathways. VHL inactivation results inhighly vascular tumors with a high risk of bleeding. VHL inactivation usually occurs through a combination of a deleterious variant and loss of a portion of chromosome 3p, home to VHL, as well as several other common genetic variants.”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “Numerous hereditary renal cancer syndromes exist including VHL disease (2%of all cases), hereditary leiomyomatosis and RCC syndrome, hereditary papillary renal cell carcinoma syndrome, and Burt Hogg Dub. syndrome. Other hereditary cancer syndromes are associated with RCC, including hereditary paraganglioma-pheochromocytoma syndrome (secondary to germline alterations in succinate dehydrogenase [SDH] subunit genes SDHB, SDHC, SDHD).”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - The classic triad of flank pain, a palpable abdominal mass, and hematuria occurs in less than 10% of patients with newly diagnosed RCC.23 Because the retroperitoneal space can accommodate substantial tumor growth prior to symptom onset, only large RCCs are detected by palpation. Currently, the widespread use of abdominal imaging leads to incidental RCC detection in 37% to 61% of cases.With increased incidental detection, gross hematuria is currently reported in less than 25% of patients and occurs more often in advanced disease. Approximately 1.3% of patients with gross hematuria are diagnosed with RCC.
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - Paraneoplastic syndromes occur in 10% to 40%of patients with RCC and are not consistently associated with higher stage or grade across studies. Common paraneoplastic manifestations include fever (8%) hypercalcemia (1%-30%), anemia (22%-52%), thrombocytosis (8%-12%), erythrocytosis (2%-4%), and hypertension (3%-18%).Paraneoplastic erythrocytosis is associated with elevated erythropoietin levels, which is produced by RCC tumor cells upon VHL inactivation. Stauffer syndrome, a paraneoplastic syndrome first described in 1961, is characterized by elevated liver enzymes in approximately 3% of patients with RCC; hepatosplenomegaly without liver metastases can also occur with this syndrome. Paraneoplastic syndromes may resolve in up to 52% of patients after nephrectomy or systemic treatment of RCC, and persistence of paraneoplastic symptoms after nephrectomy may indicate residual disease.
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010- Metastatic disease (M1) occurs in approximately 10% of patients with newly diagnosed RCC and an additional 10% of patients with localized RCC will develop metastatic disease at a later time. The most common sites of RCC metastases are lung(70%),lymph node(45%), bone(32%), liver (18%), adrenal gland (10%),and brain (8%).RCC also metastasizes to atypical sites such as thyroid, pancreas, breast, skin, and muscle.
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “Up to 20% of small renal masses ( 4 cm) are misclassified radiographically as malignant and are pathologically benign.45 For patients who undergo surgery for small renal masses, partial nephrectomy is preferred and has a 5-year cancer-specific survival of 94% to 97%. Alternatively, some patients with small renal masses may undergo active surveillance, defined as initial monitoring of tumors with serial ultrasound or abdominal CT scans (3 months, 6 months; every 6 months for up to 3 years; followed by annual surveillance), with ablation or nephrectomy recommended with growth of the tumor. In studies of active surveillance, the average tumor growth rate is 0.56 cm per year. Pathological examination of the tumor for individuals who undergo delayed nephrectomy after active surveillance is no more likely to result in upstaging than those who undergo nephrectomy soon after diagnosis.”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “Ablative techniques such as cryotherapy or radiofrequency ablation are alternatives to surgical resection and are most commonly used for patients with comorbidities and high surgical risk who have tumors less than 3 cm. Ablation can be performed with or without (in cases with imaging characteristics highly suspicious of RCC) an antecedent renal mass biopsy. Risks of ablative techniques include bleeding (2%-4%), ureteral injury (2%), urine leak (0%-4%), and urinary tract infection (2%).52 Stereotactic body radiotherapy is an ablative therapy that is increasingly used for patients who are not candidates for surgery with stage I through III RCC. An individual patient data meta-analysis reported 5-year outcomes of 199 patients who underwent stereotactic ablative body radiotherapy for primary RCC.5 The local relapse rate was 5.5%.”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - ”Patients with RCC and VHL disease often have indolent, bilateral, and multifocal tumors.Typically, small RCCs in VHLdisease are monitored with serial ultrasound or MRI imaging every 3 to6months until they reach 3 cm, at which point they are treated with partial nephrectomy or ablative therapy, whenever possible.This conservative approach to delay nephrectomy in patients with VHL disease is undertaken to decrease the lifetime risk of chronic kidney disease given the high likelihood of future development of RCCs.”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “The dual immune checkpoint inhibitors combination of ipilimumab (anti-CTLA-4) and nivolumab (anti-PD-1) is approved by the US Food and Drug Administration (FDA) as first-line treatment for patients with intermediate- or poor-risk metastatic clear cell RCC. In these patients, this regimen had an objective esponse rate of 42% and improved overall survival compared with sunitinib in patients with intermediate- and poor-risk RCC (HR, 0.69; 95% CI, 0.59-0.81).73,74 The 5-year overall survival probability was 43% (compared with 31%with sunitinib).With ipilimumab plus nivolumab, the complete response rate was10%and 58%of patients with a response remained progression-free at 5 years.”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “RCC is a common malignancy that is often diagnosed incidentally on an abdominal imaging study. At diagnosis, RCC is stage I in 70% of patients and stage IV in 11% of patients. Preferred treatments for early-stage RCC are partial or radical nephrectomy, which can result in 5-year cancer-specific survival of more than 94%; ablative techniques; or active surveillance. New treatment options for patients with metastatic RCC include immune checkpoint inhibitors and tyrosine kinase inhibitors.”
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010 - “Renal cell carcinoma (RCC) is a heterogeneous group of neoplasms derived from the renal tubular epithelial cells. Chromophobe RCC (chRCC) is the third most common subtype of RCC, accounting for 5% of cases. chRCC may be detected as an incidental finding orless commonly may manifest with clinical symptoms. The mainstay of therapy for chRCC is surgical resection. chRCC has a better prognosis compared with the more common clear cell RCC. At gross pathologic analysis, chRCC is a solid well-defined mass with lobulated borders. Histologic findings vary by subtype but include large pale polygonal cells with abundant transparent cytoplasm, crinkled “raisinoid” nuclei with perinuclear halos, and prominent cell membranes.”
Chromophobe Renal Cell Carcinoma with Radiologic-Pathologic Correlation
Marko J, Craig R, Nguyen A et al.
Radiographics. 2021 SepOct;41(5):1408-1419 - “Pathologic analysis reveals only moderate vascularity. The most common imaging pattern is a predominantly solid renal mass with circumscribed margins and enhancement less than that of the renal cortex. The authors discuss chRCC with emphasis on correlative pathologic findings and illustrate the multimodality imaging appearances of chRCC by using cases from the Radiologic Pathology Archives of the American Institute for Radiologic Pathology.”
Chromophobe Renal Cell Carcinoma with Radiologic-Pathologic Correlation
Marko J, Craig R, Nguyen A et al.
Radiographics. 2021 SepOct;41(5):1408-1419 - “chRCC is often detected as an incidental finding or less commonly manifests with clinical symptoms. With increased use of cross-sectional abdominal imaging, there is a trend toward detection of asymptomatic renal masses, including chRCC. The reported clinical findings of chRCC include local symptoms such as abdominal or flank pain (34%–67% of cases), abdominal mass (27% of cases), and hematuria (17%–40% of cases) and systemic symptoms such as fever, cachexia, fatigue, and weight loss (20%–33% of cases). Most chRCCs are low-stage cancers at diagnosis, reflecting both the indolent nature of chRCC and the trend toward incidental detection at imaging. Lymph node and distant metastases are infrequent, seen in 2%–4% and 1%–4% of cases, respectively. The most common sites of metastases are the liver and lungs.”
Chromophobe Renal Cell Carcinoma with Radiologic-Pathologic Correlation
Marko J, Craig R, Nguyen A et al.
Radiographics. 2021 SepOct;41(5):1408-1419 - - The malignant renal epithelial tumor RCC includes multiple histologic subtypes, with clear cell RCC (ccRCC), papillary RCC (papRCC), and chromophobe RCC (chRCC) accounting for approximately 75%, 15%, and 5% of cases, respectively.
- chRCC is often detected as an incidental finding or less commonly manifests with clinical symptoms. With increased use of cross-sectional abdominal imaging, there is a trend toward detection of asymptomatic renal masses, including chRCC.
- chRCC is associated with hereditary Birt-Hogg-Dubé (BHD) syndrome, which is a rare autosomal dominant syndrome associated with germline mutations in the folliculin gene, FLCN.
- chRCC is a moderately vascular tumor. It enhances less than the renal cortex in all phases, as expected given the lack of prominent vascularity seen pathologically. In addition, peak enhancement occurs in the nephrographic phase. - Table 1: WHO Classification of Renal Cell Tumors
ccRCC
Multilocular cystic renal neoplasm of low malignant
potential
papRCC
Hereditary leiomyomatosis and RCC–associated
RCC
chRCC
Collecting duct carcinoma
Renal medullary carcinoma
MiT family translocation RCCs
Succinate dehydrogenase–deficient RCC
Mucinous tubular and spindle cell carcinoma
Tubulocystic RCC
Acquired cystic disease–associated RCC
Clear cell papRCC
RCC, unclassified
Papillary adenoma
Oncocytoma - chRCC is associated with hereditary Birt- Hogg-Dubé (BHD) syndrome , which is a rare autosomal dominant syndrome associated with germline mutations in the folliculin gene, FLCN. FLCN is a tumor suppressor gene located on the short arm of chromosome . BHD affects multiple organ systems, including the skin, lungs, and kidneys. In patients with BHD, one allele of the FLCN gene is mutated in the germline. When FLCN is biallelically inactivated in the kidney, there is loss of inhibition of the mammalian target of rapamycin pathway, which is one of the major pathways that regulates angiogenesis and tumor growth in RCC.
Chromophobe Renal Cell Carcinoma with Radiologic-Pathologic Correlation
Marko J, Craig R, Nguyen A et al.
Radiographics. 2021 SepOct;41(5):1408-1419 - “At CT, chRCC is typically a well-circumscribed mass with smooth or lobular contours . Most chRCCs are solid or mostly solid at CT . chRCCs may be homogeneous or heterogeneous, with the heterogeneous appearance more commonly reported. Calcification is seen in 14%–34% of cases . A central scar is present in 19%–34% of cases, with a spoke-wheel pattern of enhancement described in a minority of cases. On noncontrast CT images, chRCC tends to be isoattenuating to slightly hyperattenuating compared with the background renal parenchyma. chRCC shows moderate enhancement that peaks in the nephrographic phase or less commonly in the corticomedullary phase. Enhancement is less than that of the renal cortex during all phases. chRCC enhances less than ccRCC but more than papRCC.”
Chromophobe Renal Cell Carcinoma with Radiologic-Pathologic Correlation
Marko J, Craig R, Nguyen A et al.
Radiographics. 2021 SepOct;41(5):1408-1419 - “The differentiation of chRCC from the more common ccRCC relies largely on the reproducible differences in enhancement. At both CT and MRI, chRCC tends to enhance less than the renal cortex in all phases and shows peak enhancement in the nephrographic phase. ccRCC enhances earlier and more intensely, with peak enhancement greater than that of the cortex in the corticomedullary phase.”
Chromophobe Renal Cell Carcinoma with Radiologic-Pathologic Correlation
Marko J, Craig R, Nguyen A et al.
Radiographics. 2021 SepOct;41(5):1408-1419 - “chRCC is the third most common subtype of RCC. Although it may occur in association with BHD, it frequently develops sporadically and is often an incidental finding discovered at imaging. The pathologic features, including well-circumscribed margins, moderate vascularity, and lack of internal fat or a significant cystic component, predict the imaging findings of a solid well-circumscribed mass with enhancement less than that of the renal cortex in all phases. Knowledge of the clinically indolent behavior and expected imaging pattern of chRCC will aid the radiologist in the evaluation of a solid renal mass detected at imaging.”
Chromophobe Renal Cell Carcinoma with Radiologic-Pathologic Correlation
Marko J, Craig R, Nguyen A et al.
Radiographics. 2021 SepOct;41(5):1408-1419
- "The results of this study are noteworthy for several reasons: first, they show that even with the new PCD-CT technology and the promising study results of VNC and VNI imaging in cardiovascular applications, there is still a significant rate of false subtraction of contrast and calcium. Second, and surprisingly, VNI does not outperform conventional VNC in stone detection, despite the fact that this algorithm specifically targets the differentiation of calcium and iodine.”
Contrast Media Subtraction for Kidney Stone Detection: Not All Problems Are Solved with Photon- Counting Detector CT
Lukas Müller, MD, Phd, Tilman Emrich, MD
Acad Radiol 2024; 31:3657–3658 - “Overall, the study shows that VNC and VNI must continue to be considered in terms of their power, taking into account potential sources of error. The many promising approaches to VNC and VNI in PCD-CT must also be considered under these limitations and do not resolve the dilemma of omitting TNC scans for certain questions. VNC, but also VNI imaging requires the implementation of appropriate reconstruction parameters. If these parameters can be correctly assessed and applied, it is possible that VNI and VNC will save significant radiation dose in the future. However, the diagnostic "cost", which is still unclear for many questions, will always be decisive. Nevertheless, PCD-CT can become one of the catalysts for broad clinical application, we just have to learn how to apply this technique and its possibilities correctly.”
Contrast Media Subtraction for Kidney Stone Detection: Not All Problems Are
Solved with Photon- Counting Detector CT
Lukas Müller, MD, Phd, Tilman Emrich, MD
Acad Radiol 2024; 31:3657–3658 - Rationale and Objectives: To evaluate and compare the effectiveness of contrast media subtraction and kidney stone detection between a virtual non-iodine reconstruction algorithm (VNI; PureCalcium) and a virtual non-contrast (VNC) algorithm in excretory phase photon-counting detector computed tomography (PCD-CT), using a 3D printed kidney phantom under various tube voltages and radiation doses.
Conclusion: VNC demonstrated greater accuracy than VNI for contrast media subtraction and kidney stone visibility. Radiation dose and tube voltage had no significant impact. Nonetheless, both algorithms still exhibited frequent incomplete contrast media subtraction and partial kidney stone subtraction.
Photon-Counting Detector CT for Kidney Stone Detection in Excretory Phase CT—Comparison Between Virtual Non-contrast and Virtual Non-iodine Reconstructions in a 3D Printed Kidney Phantom
Philipe S. Breiding, et al.
Acad Radiol 2024; 31:3650–3656 - ”Our findings also revealed that smaller stones (< 5 mm) were more frequently subtracted. This is a finding that has also been observed using virtual unenhanced images from the excretory phase generated with energy-integrating detector dual-energy CT. Karlo et. al. found that detection of urinary stones < 4 mm on virtual unenhanced images was limited and the study by Takahashi et. al. demonstrated that only 29% of kidney stones with a diameter of 1–2 mm were detected on virtual unenhanced images. These findings further highlighting the influence of stone size on subtraction extent. ”
Photon-Counting Detector CT for Kidney Stone Detection in Excretory Phase CT—Comparison Between Virtual Non-contrast and Virtual Non-iodine Reconstructions in a 3D Printed Kidney Phantom
Philipe S. Breiding, et al.
Acad Radiol 2024; 31:3650–3656 - “In conclusion, VNC demonstrated greater accuracy than VNI for contrast media subtraction and kidney stone visibility. Nonetheless, both methods still exhibited frequent incomplete contrast media subtraction and partial kidney stone subtraction, indicating the need for further research to determine if VNC from PCD-CT can effectively replace true unenhanced images in excretory phase imaging of patients. ”
Photon-Counting Detector CT for Kidney Stone Detection in Excretory Phase CT—Comparison Between Virtual Non-contrast and Virtual Non-iodine Reconstructions in a 3D Printed Kidney Phantom
Philipe S. Breiding, et al.
Acad Radiol 2024; 31:3650–3656
Pancreas
- “Radiology plays an important role in the initial diagnosis and staging of patients with pancreatic ductal adenocarcinoma (PDAC). CT is the preferred modality over MRI due to wider availability, greater consistency in image quality, and lower cost. MRI and PET/CT are usually reserved as problem-solving tools in select patients. The National Comprehensive Cancer Network (NCCN) guidelines define resectability criteria based on tumor involvement of the arteries and veins and triage patients into resectable, borderline resectable, locally advanced, and metastatic categories. Patients with resectable disease are eligible for upfront surgical resection, while patients with high-stage disease are treated with neoadjuvant chemotherapy and/or radiation therapy with hopes of downstaging the disease. The accuracy of staging critically depends on the imaging technique and the experience of the radiologists. Several challenges in accurate preoperative staging include prediction of lymph node metastases, detection of subtle liver and peritoneal metastases, and disease restaging following neoadjuvant therapy.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Artificial intelligence (AI) has the potential to function as ‘second readers’ to improve upon the radiologists’ detection of small early-stage tumors, which can shift more patients toward surgical resection of potentially curable cancer. AI may also provide imaging biomarkers that can predict disease recurrence and patient survival after pancreatic resection and assist in the selection of patients most likely to benefit from surgery, thus improving patient outcomes.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Pancreatic ductal adenocarcinoma (PDAC) is the seventh leading cause of cancer mortality worldwide based on GLOBOCAN 2020 estimates, and over 466 000 patients with pancreatic cancer succumbed to the disease in 2020.. The age-standardized incidence of PDAC is fourfold to fivefold higher in countries with a high development index, with the greatest incidence in Europe, North America, Australia, and New Zealand. Therefore, PDAC is expected to surpass breast cancer as the third leading cause of cancer death in the United States and Europe. Despite therapeutic advances, the 5-year survival rate of patients with PDAC remains ~10% since most patients are diagnosed at an advanced stage of disease. Surgical resection remains the only curative therapy for patients with PDAC, and radiology plays a pivotal role in disease staging and patient management."
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “PDACs classically present as hypoenhancing masses with associated pancreatic duct dilatation and glandular atrophy of the body and tail. Pancreatic head tumors can cause common bile duct dilatation in addition to pancreatic duct dilatation, also known as the ‘double duct sign’. Up to 20% of PDACs enhance to the same degree as the background pancreas, and this isoattenuating pattern is more commonly found with smaller ( ≤20 mm) tumors. These small isoattenuating tumors can be difficult to detect on CT; therefore, radiologists often rely on secondary signs of the pancreatic duct or common bile duct dilatation for tumor detection. MRI and PET/CT have reported sensitivities of 79.2 and 73.7% in the detection of isoattenuating tumors, respectively, and may aid in detecting suspected pancreatic tumors that are occult on CT. Endoscopic ultrasound is crucial in confirming tissue diagnosis of suspected pancreatic malignancy. It is also an important second-line modality in detecting suspected pancreatic tumors that are occult on CT or MRI.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063- “Subtle liver or peritoneal metastases can be occult on preoperative CTs in up to 30% of cases, and diagnostic laparoscopy may be useful for patients at high risk of advanced disease. High-quality imaging is essential in the detection of subtle liver or peritoneal metastases. Small liver and peritoneal metastases may be obscured by image noise, thick image slices, or suboptimal contrast injection. Liver metastases from PDAC are typically hypoenhancing on the portal venous phase and can mimic the appearance of cysts or hemangiomas. On arterial phase images, these liver metastases may contain peripheral enhancing rims, and this targetoid appearance can significantly improve the diagnostic confidence of small liver metastases. MRI has improved tissue characterization compared to CT and is valuable for characterizing small indeterminate liver lesions.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Cinematic rendering, a recently described 3D rendering technique, uses a global illumination model that considers direct and indirect lighting to create images with photorealistic quality. Cinematic rendering can accentuate subtle texture changes and improve tumor conspicuity relative to traditional 2D images, 3D volume rendering, or maximum intensity projection images. Cinematic rendering may be able to enhance the visualization of spatial relationships among the tumor and adjacent vasculature, differentiating true tumor infiltration from simple proximity to vessels. This can potentially improve the assessment of resectability and assist in determining optimal vascular reconstruction options. Cinematic rendering vascular maps illustrate the major arteries and veins with exquisite detail and can highlight the presence of variant vascular anatomy that may increase the risk of complications, such as hemorrhage, ischemia, anastomotic leakage, or pseudoaneurysm formation.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - ”Cinematic rendering vascular maps illustrate the major arteries and veins with exquisite detail and can highlight the presence of variant vascular anatomy that may increase the risk of complications, such as hemorrhage, ischemia, anastomotic leakage, or pseudoaneurysm formation. At our institution, cinematic rendering has been routinely incorporated into the multidisciplinary PDAC clinic since 2018, and it has played an important role in tumor staging as well as patient management. Moreover, cinematic rendering data can be imported into augmented reality headsets to provide an immersive experience for the surgeon for operative planning.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Artificial intelligence (AI) is poised to revolutionize medicine, and radiology is a natural gateway due to the inherent digital nature of radiology data. AI can be broadly defined as using computers to perform tasks typically associated with human intelligence. Machine learning, a branch of AI, enables the extraction of meaningful patterns from examples rather than through explicit programming. Deep learning (DL), a subfield of machine learning first developed in the 1950s, utilizes networks of interconnected nodes that process input data and adjust the network weights to minimize prediction errors. Recent developments in powerful parallel computing hardware, the availability of large training data, and improved network architectures have notably enhanced the performance of deep learning, which has significant potential for clinical translation.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Radiomics converts imaging data into high-dimensional features that can be used to characterize spatial heterogeneity inherent in disease processes. The features of radiomics can be classified into signal intensity, shape, and texture. Signal intensity (first-order) features are derived from histograms of individual voxel signal intensities, providing measures of central tendency and shape of the distribution. Shape features are extracted from the three-dimensional surface of the region of interest. Texture features are calculated in three dimensions, considering the correlation of signal intensities of adjacent voxels. In addition, feature extraction may be performed after applying a secondary filter, such as a wavelet or Gaussian filter.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “AI can theoretically function as ‘second readers’ to improve radiologists’ sensitivity in the detection of small tumors, which potentially can be cured with surgical resection. A preliminary study by Liu et al. showed promising results suggesting that DL could accurately differentiate CT scans of patients with PDAC from CT scans of healthy controls. More recently, Chen et al. developed a DL tool that differentiated CT scans of patients with PDAC vs. healthy controls with 89.9% sensitivity, 95.9% specificity, and 93.4% accuracy in the local test set. They validated this DL tool on a Taiwanese nationwide external validation set and achieved 89.7% sensitivity, 92.8% specificity, and 91.4% accuracy. Also, Park et al. developed a different DL tool that achieved high sensitivity comparable to radiologists in the detection of not only pancreatic solid masses (98–100%) but also cystic masses 1.0 cm or larger (sensitivity 92–93%), bringing us closer to a universal pancreatic neoplasm detector.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Other studies have used radiomics to facilitate the detection of PDAC, demonstrating that radiomics signatures from PDAC were distinct from the background pancreas. More impressively, radiomics signatures could identify subtle differences in prediagnostic CT scans obtained with a median of 386 days before PDAC diagnosis, with 95.5% sensitivity, 90.3% specificity, and 92% accuracy. If these promising results are validated in future studies, radiologists will be able to diagnose patients significantly earlier at lower disease stages. In this scenario, a higher proportion of newly diagnosed patients will be eligible for curative surgical resection, which will have a significant positive impact on patient outcomes.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Researchers have also used radiomics features to predict the development of liver metastases after PDAC resection in multiple studies. Zambirinis et al. analyzed 254 radiomics features from the liver from preoperative CTs in 688 patients with resected PDAC and the radiomics model identified patients at risk for early (< 6 months) liver metastases with an AUC of 0.71. Huang et al. extracted 3906 radiomics features from the pancreatic tumor from preoperative MRIs in 204 patients with resected PDAC, and the radiomics model achieved 75.0% sensitivity, 82.2% specificity, and an AUC of 0.815 in predicting the development of liver metastases. We speculate that radiologic features from both the primary tumor and the liver parenchyma are important in predicting future liver metastases. Future studies should incorporate features from both the tumor and the liver, in combination with clinical features, to optimize the prediction of liver metastases.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063 - “Preliminary studies using emerging technologies such as advanced visualization and AI have revealed the potential of these tools to improve the initial diagnosis and staging of patients with PDAC. However, there remain several limitations. Most of these studies have been single-center retrospective studies, and their promising results should be validated in future multicenter prospective studies. Secondly, one of the major criticisms of AI is its ‘blackbox’ nature, making it difficult for clinicians to decipher the rationale behind AI predictions. Explainable or ‘glassbox’ AI is an active area of research that aims to render AI models more easily understandable and may help improve their clinical acceptance. Thirdly, these tools should be integrated seamlessly into the workflow to ensure widespread clinical implementation.”
Pancreatic ductal adenocarcinoma staging: a narrative review of radiologic techniques and advances
Linda C. Chu, Elliot K. Fishman
International Journal of Surgery (2024) 110:6052–6063
- Objectives: This research aimed to assess the value of radiomics combined with multiple machine learning algorithms in the diagnosis of pancreatic ductal adenocarcinoma (PDAC) lymph node (LN) metastasis, which is expected to provide clinical treatment strategies. Conclusions: Combining radiomics and machine learning algorithms demonstrated the potential for identifying the LN metastasis of PDAC. As a non-invasive and efficient preoperative prediction tool, it can be beneficial for decision-making in clinical practice.
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317 - Methods: A total of 128 patients with pathologically confirmed PDAC and who underwent surgical resection were randomized into training (n=93) and validation (n=35) groups. This study incorporated a total of 13 distinct machine learning algorithms and explored 85 unique combinations of these algorithms. The area under the curve (AUC) of each model was computed. The model with the highest mean AUC was selected as the best model which was selected to determine the radiomics score (Radscore). The clinical factors were examined by the univariate and multivariate analysis, which allowed for the identification of factors suitable for clinical modeling. The multivariate logistic regression was used to create a combined model using Radscore and clinical variables. The diagnostic performance was assessed by receiver operating characteristic curves, calibration curves, and decision curve analysis (DCA).
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317 - Radiomics provides a large amount of medical image information that can reveal hidden features of diseases that are invisible to the naked eye. Previous research has demonstrated the efficacy of integrating radiomics and machine learning (ML) techniques in many applications. Several researchers have constructed multiphasic contrast-enhanced CT (CECT) radiomics models to evaluate the preoperative LN status of PDAC. These studies indicated that radiomics models have significant potential in predicting pancreatic cancer with lymph node metastasis (LNM). However, the researchers predominantly chose modeling algorithms based on their preferences and limitations in knowledge. It is imperative that evidence be provided to select appropriate models for solving clinical problems. To the best of our knowledge, no study has explored the use of radiomics combined with multiple ML algorithms in thepreoperative identification of the LN status in PDAC
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317 - “Although CT is the most commonly used method to evaluate pancreatic cancer resectability, Imai et al. indicated that loworder data such as LN diameter and volume measured from CT image cannot reflect obvious differences between patients with pancreatic cancer with and without LN metastasis. This may explain why visual assessment of CT scans has a low efficiency for detecting LN metastases among patients with pancreatic cancer. Radiomics, also referred to as a “whole-tumor virtual biopsy technique,” enables the extraction of numerous image features from the entire tumor, reflecting its heterogeneity and characteristics. A biopsy examines only a portion of the tumor tissue, which cannot comprehensively assess the intratumor heterogeneity before surgery. In contrast, radiomics can non-invasively reflect comprehensive information about the entire tumor. This approach has promise for enhancing diagnostic capabilities and facilitating the development of personalized treatment approaches.”
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317 - “In summary, of the 233 radiomics models examined, the model built by applying AP+VP radiomics features and a combination of Lasso-Logistic algorithm had the most favorable performance in both the training and validation cohorts. Our investigation showed that integrating the AP+VP-Radscore with clinical parameters yielded the best performance. This combined model has the potential to serve as an accurate and non-invasive instrument for forecasting LN metastasis of PDAC, hence facilitating clinical decision-making.”
Contrast-enhanced CT radiomics combined with multiple machine learning algorithms for preoperative identification of lymph node metastasis in pancreatic ductal adenocarcinoma
Yue Huang et al.
Front. Oncol. 14:1342317. doi: 10.3389/fonc.2024.1342317
- “IPMNs can be challenging to diagnose accurately because several benign lesions can mimic their characteristics. Unfortunately, CT and MRI have mixed accuracy. In one study, the accuracy of CT or MRI for the overall characterization of a PCL is relatively high though with limited diagnostic accuracy for specific cystic pancreatic masses based on CT or MRI appearance. In the study, reviewers were correct in only 57% of cases in which they were confident in the leading diagnosis. Nevertheless, MRI has higher soft-tissue contrast and multiplanar capability result in superior differentiation of cystic neoplasms while others have found that MRI adds little to CT.”
Misdiagnosis of pancreatic intraductal papillary mucinous neoplasms and the challenge of mimicking lesions: imaging diagnosis and differentiation strategies
Mohammad Yasrab· Stephen J. Kwak · Parissa Khoshpouri· Elliot K. Fishman · Atif Zaheer
Abdominal Radiology (2024, in press) https://doi.org/10.1007/s00261-024-04551-x 
Misdiagnosis of pancreatic intraductal papillary mucinous neoplasms and the challenge of mimicking lesions: imaging diagnosis and differentiation strategies
Mohammad Yasrab· Stephen J. Kwak · Parissa Khoshpouri· Elliot K. Fishman · Atif Zaheer
Abdominal Radiology (2024, in press) https://doi.org/10.1007/s00261-024-04551-x- “The rising prevalence of pancreatic cystic lesions (PCLs), particularly intraductal papillary neoplasms (IPMNs), has been attributed to increased utilization of advanced imaging techniques. Incidental detection of PCLs is frequent in abdominal CT and MRI scans, with IPMNs representing a significant portion of these lesions. Surveillance of IPMNs is recommended due to their malignant potential; however, their overlapping imaging features with benign entities can lead to misdiagnosis, overtreatment, and overutilization of healthcare resources. This paper aims to highlight and differentiate lesions often mistaken for IPMNs, providing insight into their imaging characteristics, diagnostic challenges, and distinctive features while highlighting the incidence of wrong diagnosis for these lesions. These lesions include serous cystadenomas, cystic pancreatic neuroendocrine tumors, mucinous cystic neoplasms, lymphoepithelial cysts, duodenal diverticula, pancreatic schwannomas, chronic pancreatitis, retention cysts, intrapancreatic accessory spleens, pancreatic lipomas, choledochal cysts, and others. Utilizing various imaging modalities, including contrast-enhanced CT, MRI, and EUS, alongside histological and molecular analyses, can aid in accurate diagnosis and appropriate management.”
Misdiagnosis of pancreatic intraductal papillary mucinous neoplasms and the challenge of mimicking lesions: imaging diagnosis and differentiation strategies
Mohammad Yasrab· Stephen J. Kwak · Parissa Khoshpouri· Elliot K. Fishman · Atif Zaheer
Abdominal Radiology (2024, in press) https://doi.org/10.1007/s00261-024-04551-x
Practice Management
Leadership: A Different Approach From a Different Perspective.
Catmull E, Fishman EK, Chu LC, Rizk RC, Rowe SP, Huang JH.
J Am Coll Radiol. 2024 Sep 16:S1546-1440(24)00769-5. doi: 10.1016/j.jacr.2024.08.028. Epub ahead of print. PMID: 39293547.- “There are numerous parallels between the complex leadership responsibilities of people in the technology industry and those in academic medicine. Although it has taken some time to figure out leadership in rapidly evolving fields, one of the key elements is truly loving what you do and understanding why you are doing what you are doing. People can easily tell if you are sincere—or insincere, for that matter. Leading means people know you are genuine. Another crucial component is caring about the people you work with—taking care of them, guiding them, teaching them, and recognizing their contributions.”
Leadership: A Different Approach From a Different Perspective.
Catmull E, Fishman EK, Chu LC, Rizk RC, Rowe SP, Huang JH.
J Am Coll Radiol. 2024 Sep 16:S1546-1440(24)00769-5. doi: 10.1016/j.jacr.2024.08.028. Epub ahead of print. PMID: 39293547. - “Those are the foundational aspects. Ambition is fine—but leading with ambition alone is the wrong approach. You often see people who want to be in charge or become the chief executive officer—but it is essential to realize that achieving greatness is only possible if the people you work with are great. Every great leader will acknowledge their dependence on others’ greatness. It is crucial to recognize that success is a collective effort. It cannot be achieved without everyone being at their best. Additionally, it is important to create an environment free of class structures. Regardless of their role in the company, valuing everyone equally sends a positive message to all employees. Although skills and abilities differ, treating everyone with respect as human beings is vital. Those are the key elements of effective leadership.” .
Leadership: A Different Approach From a Different Perspective.
Catmull E, Fishman EK, Chu LC, Rizk RC, Rowe SP, Huang JH.
J Am Coll Radiol. 2024 Sep 16:S1546-1440(24)00769-5. doi: 10.1016/j.jacr.2024.08.028. Epub ahead of print. PMID: 39293547. - “The key is to embrace the technology, understand its potential, and figure out how to use it effectively. Rather than feeling threatened, we should focus on how to work with it and advance it. That is the optimal approach. Although leadership across various disciplines faces challenges with rapidly evolving technology, staying true to our principles— including being sincere in loving what we do, recognizing the greatness of people around us, and embracing new technology and the people who adapt it—we can continue to make impactful decisions as leaders.”
Leadership: A Different Approach From a Different Perspective.
Catmull E, Fishman EK, Chu LC, Rizk RC, Rowe SP, Huang JH.
J Am Coll Radiol. 2024 Sep 16:S1546-1440(24)00769-5. doi: 10.1016/j.jacr.2024.08.028. Epub ahead of print. PMID: 39293547. - “Perhaps the issue is not so much in performance—but a misunderstanding of what is leadership. As Ed Catmull clearly articulated in his contribution to this manuscript—leadership comes down to culture, team, and institution creating an environment in which risks are taken and people’s contributions are recognized. While having trained with the best and the brightest like Alan Kay, Ivan Sutherland, and Gordon Moore during graduate school at the University of Utah, Ed came to realize that it was not the individual people who made the difference—instead, it was the environment they created that allowed for the ability of others to grow and succeed. It was not the willingness to copy a specific individual but to try and create an environment of excellence that led everyone to perform at a higher level than individuals might consider possible.”
Leadership: A Different Approach From a Different Perspective.
Catmull E, Fishman EK, Chu LC, Rizk RC, Rowe SP, Huang JH.
J Am Coll Radiol. 2024 Sep 16:S1546-1440(24)00769-5. doi: 10.1016/j.jacr.2024.08.028. Epub ahead of print. PMID: 39293547. - “In an atmosphere of doom and gloom, there is little any individual can do to change the environment. On the other hand, a strong environment that leads to numerous successful outcomes is something that people will try to translate and duplicate. Whether, in Ed’s case, it was moving the University of Utah environment to Lucasfilm or Pixar or it was the leadership of a John Cameron at Johns Hopkins Department of Surgery in the 1990s, such stories evoke an environment that tested people and brought out the best in so many. None of those stories was based on doing things easier—but instead asking why not be the best? To quote George Bernard Shaw, the real question is, “You see things and say ‘Why?’ But I dream things that never were; and I say why not?”
Leadership: A Different Approach From a Different Perspective.
Catmull E, Fishman EK, Chu LC, Rizk RC, Rowe SP, Huang JH.
J Am Coll Radiol. 2024 Sep 16:S1546-1440(24)00769-5. doi: 10.1016/j.jacr.2024.08.028. Epub ahead of print. PMID: 39293547. - “Perhaps, we need then rethink leadership in the medical environment and need to realize that unless we can change the environment to be positive and productive—to believing the best is yet to come and to believing that we are not being overrun by a leadership team of administrators whose focus is the short-term bottom line rather than the success of the enterprise and its future—we are in serious trouble. The question then is, can we change the environment before it is too late? Change must be timely, or if not, we will end up where many companies like Sun Microsystems, Silicon Graphics, and Eastman Kodak have wound up—footnotes in history on the road to newer and more expansive technologies. Change has never been faster, the stakes never higher, and thetime to succeed never shorter. What then is your plan?”
Leadership: A Different Approach From a Different Perspective.
Catmull E, Fishman EK, Chu LC, Rizk RC, Rowe SP, Huang JH.
J Am Coll Radiol. 2024 Sep 16:S1546-1440(24)00769-5. doi: 10.1016/j.jacr.2024.08.028. Epub ahead of print. PMID: 39293547.
Small Bowel
- ”Hypovolemic shock caused by hemorrhage is a rapidly fatal condition. Acute blood loss causes generalized vasoconstriction and hypoperfusion which results in a cascade of insults at the cellular level, including profound acidosis and coagulopathy. Without timely and effective intervention, this can lead to multi-organ dysfunction and death owing to ischemia of the myocardium or central nervous system . It is estimated that hemorrhage results in more than 60,000 deaths each year in the United States, and roughly 1.9 million deaths each year worldwide. Etiologies of hemorrhage requiring resuscitation via massive transfusion protocols are myriad, with trauma and perioperative bleeding representing the greatest proportions, though causes include obstetric, gastrointestinal, and aneurysmal bleeding, among other. Hemorrhagic shock has been studied extensively in the setting of trauma and is a leading source of both mortality and morbidity for survivors.”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “The appearance of active bleeding on CT is well known. Active hemorrhage is denoted by the extravasation of intravascular contrast, which is seen at CT as a globular or linear focus of contrast outside of the vessels that increases in size and density on delayed imaging. CT is highly sensitive for the detection of contrast extravasation and can demonstrate bleeding below a rate of 0.4 ml/min, which has led to increased use in emergent settings as a first line diagnostic exam for active bleeding.”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “Contrast extravasation seen at arterial phase imaging represents active arterial hemorrhage, classically increasing in size and density on subsequent venous phases. Extravascular contrast visualized only on venous phase imaging characteristically denotes active venous hemorrhage, although very small arterial bleeds or intermittent arterial bleeding owing to the presence of hypovolemic vasospasm may only be visible on a venous phase image acquisition. Whether bleeding is arterial or venous in nature, bleeding seen only on venous phase acquisitions is often assumed to be clinically minor, and less likely to require angiographic intervention. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “A benefit of a dual-phase scan is that it allows for the diagnoses of arterial pseudoaneurysm and arteriovenous fistula and their differentiation from active hemorrhage. An arterial pseudoaneurysm is classically described as an enhancing focus which follows the density of the aortic blood pool on multiple phases of contrast, with its washout on venous phase imaging helping to distinguish it from active bleeding, which can be difficult or impossible on a single-phase study. Additionally, a pseudoaneurysm typically has a well-defined, often rounded morphology, with active bleeding appearing more diffuse and amorphous. An arteriovenous fistula is identified when there is early filling of a vein adjacent to an injured artery on an arterial phase scan. While these lesions do not always present with active bleeding, discovery should nonetheless prompt IR consultation for the purposes of future treatment planning. On a single-phase venous scan, it may be challenging to determine if bleeding is venous or arterial unless extravasation from a known vessel is identified. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “Rectus sheath hematomas (RSH) are commonly encountered in emergency medicine practices and are the source of considerable morbidity and occasional mortality if not promptly treated. The majority of RSH are supplied by the inferior epigastric artery and its branches, which allows for targeted embolization, and contrast extravasation is a predictor of failure of conservative management. Chest wall injuries may also result in active arterial bleeding, which can be effectively embolized if the source vessel is an internal thoracic or intercostal artery as seen in conjunction with displaced sternal and rib fractures, respectively. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “The specific location and characteristics of gastrointestinal hemorrhage have important implications for the interventional radiologist. For example, diverticular sources of hemorrhage are often treated conservatively as they are characteristically self-limiting, while ulcerations or colorectal neoplasms may require embolization, and variceal bleeding may necessitate portal decompression via placement of a transjugular intrahepatic portosystemic shunt (TIPS). Upper endoscopy may still be selected for upper gastrointestinal bleeds, particularly if there is a need for concomitant tissue sampling. It must be noted that even with super-selective catheterization techniques, there is a risk of visceral ischemia if embolization is performed following a bowel resection or bypass surgery.”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “Aortic rupture is a highly morbid, rapidly fatal event. Rupture can be the result of trauma or a spontaneous occurrence in the setting of an acute dissection or a pre-existing aneurysm. It is estimated that at least half of the patients who experience aortic rupture expire on-scene or in transport to the hospital. Abdominal aortic aneurysm (AAA) rupture carries an overall risk of mortality near 90% . A rapid evaluation with CT angiography is crucial to localize the rupture, illustrate anatomy, and identify complicating features such as hemopericardium and hemothorax. Non-contrast CT and point-of-care ultrasound, which has become a common practice in emergency departments, can also identify a ruptured AAA, though these tests provide less vascular detail. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “Active bleeding as diagnosed by CT is a clinical emergency that requires swift action driven by efficient communication. Emergency radiologists are central to this process and must provide guidance to emergency medicine and surgical providers regarding the role interventional radiology should play in the management of these critically ill patients. Indications for IR involvement are broad and include trauma to the solid organs and mesentery, intraluminal GI hemorrhage, rectus sheath hematomas, tumoral bleeding, and gynecological bleeding including postpartum hemorrhage. Bleeding conditions which are less likely to be managed by IR include venous bleeding, intramuscular bleeding, mesenteric and bowel hemorrhage in patients with surgically altered anatomy, and small organ hemorrhage such as lower genitourinary or penile bleeding, for which skilled surgical reconstruction will need to be performed in addition to hemostasis. Local and regional practice patterns may vary, however. Emergency radiologists should be familiar with the information our interventional radiology colleagues often need when consulted on these cases, such as surgical history, differentiation of arterial and venous bleeding, anatomic localization of bleeding to a culprit artery, and coagulation status so that we may facilitate expedient decision making and treatment. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z
Syndromes in CT
- ”Hypovolemic shock caused by hemorrhage is a rapidly fatal condition. Acute blood loss causes generalized vasoconstriction and hypoperfusion which results in a cascade of insults at the cellular level, including profound acidosis and coagulopathy. Without timely and effective intervention, this can lead to multi-organ dysfunction and death owing to ischemia of the myocardium or central nervous system . It is estimated that hemorrhage results in more than 60,000 deaths each year in the United States, and roughly 1.9 million deaths each year worldwide. Etiologies of hemorrhage requiring resuscitation via massive transfusion protocols are myriad, with trauma and perioperative bleeding representing the greatest proportions, though causes include obstetric, gastrointestinal, and aneurysmal bleeding, among other. Hemorrhagic shock has been studied extensively in the setting of trauma and is a leading source of both mortality and morbidity for survivors.”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “The appearance of active bleeding on CT is well known. Active hemorrhage is denoted by the extravasation of intravascular contrast, which is seen at CT as a globular or linear focus of contrast outside of the vessels that increases in size and density on delayed imaging. CT is highly sensitive for the detection of contrast extravasation and can demonstrate bleeding below a rate of 0.4 ml/min, which has led to increased use in emergent settings as a first line diagnostic exam for active bleeding.”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “Contrast extravasation seen at arterial phase imaging represents active arterial hemorrhage, classically increasing in size and density on subsequent venous phases. Extravascular contrast visualized only on venous phase imaging characteristically denotes active venous hemorrhage, although very small arterial bleeds or intermittent arterial bleeding owing to the presence of hypovolemic vasospasm may only be visible on a venous phase image acquisition. Whether bleeding is arterial or venous in nature, bleeding seen only on venous phase acquisitions is often assumed to be clinically minor, and less likely to require angiographic intervention. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “A benefit of a dual-phase scan is that it allows for the diagnoses of arterial pseudoaneurysm and arteriovenous fistula and their differentiation from active hemorrhage. An arterial pseudoaneurysm is classically described as an enhancing focus which follows the density of the aortic blood pool on multiple phases of contrast, with its washout on venous phase imaging helping to distinguish it from active bleeding, which can be difficult or impossible on a single-phase study. Additionally, a pseudoaneurysm typically has a well-defined, often rounded morphology, with active bleeding appearing more diffuse and amorphous. An arteriovenous fistula is identified when there is early filling of a vein adjacent to an injured artery on an arterial phase scan. While these lesions do not always present with active bleeding, discovery should nonetheless prompt IR consultation for the purposes of future treatment planning. On a single-phase venous scan, it may be challenging to determine if bleeding is venous or arterial unless extravasation from a known vessel is identified. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “Rectus sheath hematomas (RSH) are commonly encountered in emergency medicine practices and are the source of considerable morbidity and occasional mortality if not promptly treated. The majority of RSH are supplied by the inferior epigastric artery and its branches, which allows for targeted embolization, and contrast extravasation is a predictor of failure of conservative management. Chest wall injuries may also result in active arterial bleeding, which can be effectively embolized if the source vessel is an internal thoracic or intercostal artery as seen in conjunction with displaced sternal and rib fractures, respectively. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “The specific location and characteristics of gastrointestinal hemorrhage have important implications for the interventional radiologist. For example, diverticular sources of hemorrhage are often treated conservatively as they are characteristically self-limiting, while ulcerations or colorectal neoplasms may require embolization, and variceal bleeding may necessitate portal decompression via placement of a transjugular intrahepatic portosystemic shunt (TIPS). Upper endoscopy may still be selected for upper gastrointestinal bleeds, particularly if there is a need for concomitant tissue sampling. It must be noted that even with super-selective catheterization techniques, there is a risk of visceral ischemia if embolization is performed following a bowel resection or bypass surgery.”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “Aortic rupture is a highly morbid, rapidly fatal event. Rupture can be the result of trauma or a spontaneous occurrence in the setting of an acute dissection or a pre-existing aneurysm. It is estimated that at least half of the patients who experience aortic rupture expire on-scene or in transport to the hospital. Abdominal aortic aneurysm (AAA) rupture carries an overall risk of mortality near 90% . A rapid evaluation with CT angiography is crucial to localize the rupture, illustrate anatomy, and identify complicating features such as hemopericardium and hemothorax. Non-contrast CT and point-of-care ultrasound, which has become a common practice in emergency departments, can also identify a ruptured AAA, though these tests provide less vascular detail. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z - “Active bleeding as diagnosed by CT is a clinical emergency that requires swift action driven by efficient communication. Emergency radiologists are central to this process and must provide guidance to emergency medicine and surgical providers regarding the role interventional radiology should play in the management of these critically ill patients. Indications for IR involvement are broad and include trauma to the solid organs and mesentery, intraluminal GI hemorrhage, rectus sheath hematomas, tumoral bleeding, and gynecological bleeding including postpartum hemorrhage. Bleeding conditions which are less likely to be managed by IR include venous bleeding, intramuscular bleeding, mesenteric and bowel hemorrhage in patients with surgically altered anatomy, and small organ hemorrhage such as lower genitourinary or penile bleeding, for which skilled surgical reconstruction will need to be performed in addition to hemostasis. Local and regional practice patterns may vary, however. Emergency radiologists should be familiar with the information our interventional radiology colleagues often need when consulted on these cases, such as surgical history, differentiation of arterial and venous bleeding, anatomic localization of bleeding to a culprit artery, and coagulation status so that we may facilitate expedient decision making and treatment. ”
Clinical management of active bleeding: what the emergency radiologist needs to know
Ryan T. Whitesell · Cory R. Nordman · Sean K. Johnston · Douglas H. Sheafor
Emergency Radiology https://doi.org/10.1007/s10140-024-02289-z
- “Erdheim-Chester disease (ECD) is a rare multiorgan histiocytosis with diverse clinical manifestations. Initially described as “lipoid granulomatosis” by William Chester and Jakob Erdheim in 1930 , it was named after its founders by Henry Jaffe in 1972. ECD was previously considered inflammatory histiocytosis, but the discovery of activating mutations in the mitogen-activated protein kinase (MAPK) signaling pathway has established ECD as a histiocytic neoplasm, although inflammation contributes to organ damage . Hence, ECD was included in the “histiocytic and dendritic cell neoplasms” category in the 2016 World Health Organization classification of hematopoietic tumors. Similarly, ECD along with Langerhans cell histiocytosis (LCH) has been included in the Langerhans group in the 2016 revised Histiocyte Society classification of histiocytosis, as both entities share mutations in the MAPK pathway.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 - - The pattern of bilateral symmetric metadiaphyseal regions of osteosclerosis, especially around the knees, is virtually pathognomonic for ECD. These regions reveal intense uptake at scintigraphy and FDG PET/ CT, again a finding nearly exclusive to ECD.
- ECD in the retroperitoneum manifests as infiltrative soft-tissue masses in bilateral perirenal (hairy kidneys) and posterior pararenal spaces.
- Myocardial infiltration has a predilection for the right atrium and atrioventricular sulcus and manifests as a focal mural mass (pseudotumor).
- Chest radiographic findings may be normal. Pulmonary involvement is, however, seen at CT in 50% of patients and often shows smooth interlobular septal thickening. The histiocytic infiltrate is perilymphatic in distribution and causes smooth thickening of the interlobular septa, peribronchovascular interstitium, and visceral pleura.
- Intracranial lesions in ECD, including intra-axial, meningeal, perivascular, and pituitary infundibulum lesions, are rarely isolated, and patients who harbor these lesions almost always have osteosclerosis of the paranasal sinus walls and/or orbital disease.
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011- “Retroperitoneal and renal involvement in ECD is seen in 50%–60% of patients. Although patients with renal involvement are frequently asymptomatic, they may present with abdominal pain and dysuria . ECD in the retroperitoneum manifests as infiltrative soft-tissue masses in bilateral perirenal (hairy kidneys) and posterior pararenal spaces. On CT images, this soft tissue is isoattenuating relative to skeletal muscle and shows mild postcontrast enhancement. At MRI, it is iso- to hypointense relative to skeletal muscle on T1- and T2-weighted images and shows mild postcontrast enhancement. Because the normal renal parenchyma shows intense uptake of FDG, the sensitivity of FDG PET/CT for detection of retroperitoneal involvement, in particular perinephric ECD, is not as high as contrast-enhanced CT.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 - “Perinephric soft-tissue infiltrates, other than ECD, can also be seen in lymphoma, retroperitoneal fibrosis (RPF), RDD, amyloid, and extramedullary hematopoeisis (EMH) . ECD and lymphoma may involve the perinephric and retroaortic spaces and show uptake at FDG PET/ CT. Contiguous involvement of the kidneys or perinephric space from retroperitoneal masses or lymph nodes is a common pattern of renal lymphoma. Contrarily, ECD rarely affects the lymph nodes.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 - “Although ECD and RPF cause hydronephrosis, the former is epicentered around the renal hila and proximal ureters, whereas the latter is epicentered around aortic bifurcation and hence affects the distal ureters. As RPF extends more superiorly, it lies anterior and lateral to the aorta and rarely infiltrates the renal hila and perinephric spaces. Immunoglobulin G4 (IgG4)–related RPF may also show pancreatic involvement. Renal involvement in RDD most commonly manifests as bilateral hilar masses.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 - “Cardiac ECD can affect the myocardium, pericardium, or endocardium. Myocardial infiltration has a predilection for the right atrium and atrioventricular sulcus and manifests as a focal mural mass (pseudotumor). Up to 40% of patients may have a right atrial pseudotumor . The pseudotumor appears as a soft-tissue mass on CT images, which is isoattenuating relative to the myocardium and shows mild enhancement.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011- “Neurologic disease is seen in 25%–50% of patients and most commonly affects the hypothalamic-pituitary axis (up to 53%). Other common locations include the meninges and neuroparenchyma. Intracranial lesions in ECD, including intra-axial, meningeal, perivascular, and pituitary infundibulum lesions, are rarely isolated, and patients who harbor these lesions almost always have osteosclerosis of the paranasal sinus walls and/or orbital disease.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011 - “ECD is a multiorgan inflammatory histiocytic neoplasm that most frequently affects the skeletal system and produces virtually pathognomonic osteosclerosis around the knees that shows increased radiotracer uptake on nuclear medicine studies. Other typical imaging findings comprise perinephric infiltrates (hairy kidneys), periaortic soft tissue (coated aorta), and a right atrial pseudotumor. However, radiologic findings are interpreted in conjunction with clinical and histologic features to establish the diagnosis. In addition to aiding in diagnosis, imaging also helps to determine the extent of disease, target sites for tissue sampling, assess treatment response,and predict the prognosis.”
Imaging in Erdheim-Chester Disease
Yashant Aswani, et al.
RadioGraphics 2024; 44(9):e240011
