Imaging Pearls ❯ Colon ❯ Artificial Intelligence (AI)
|
-- OR -- |
|
- “Evaluation of class activation maps can help radiologists gain confidence in how an AI algorithm assesses the imaging findings that contribute to the tool’s final decision. In the present study, the activation maps indicated that the AI results were based on alterations in pericolonic fat (e.g., extracolonic tumor extension, fluid collection, and pericolic fat stranding), consistent with classic secondary findings, rather than on bowel wall thickening. This explanation is crucial for radiologists’ uptake of CADx tools. However, the authors do not indicate the time required by the AI algorithm to analyze each CT examination, and they acknowledge that clinical use of the tool requires a user interface for the radiologist to manually define the pathologic colon segment in 3D. Depending on implementation, this step could be too time-consuming to be accepted by radiologists in a clinical setting. Nonetheless, the present work should inspire radiologists and engineers to create their own AI solutions to address such issues.”
Beyond the AJR: Artificial Intelligence Helps Radiologists to Improve Their Performance in Differentiating Colon Carcinoma From Acute Diverticulitis on CT.
Martín-Noguerol T, Luna A.
AJR Am J Roentgenol. 2023 Sep 20:1. doi: 10.2214/AJR.23.29466. - ”Use of an AI algorithm trained to differentiate CC and AD may significantly improve radiologists’ performance in the assessment of patients with focal large-bowel wall thickening.”
Beyond the AJR: Artificial Intelligence Helps Radiologists to Improve Their Performance in Differentiating Colon Carcinoma From Acute Diverticulitis on CT.
Martín-Noguerol T, Luna A.
AJR Am J Roentgenol. 2023 Sep 20:1. doi: 10.2214/AJR.23.29466.
- “Artificial intelligence using computer-aided diagnosis (CADx) in real with images acquired during colonoscopy may help colonoscopists distinguish between neoplastic polyps requiring removal and nonneoplastic polyps not requiring removal. In this study, we tested whether CADx analyzed images helped in this decision-making process.”
Real-Time Artificial Intelligence–Based Optical Diagnosis of Neoplastic Polyps during Colonoscopy
Ishita Barua, et al.
NEJM Evid 2022; 1 (6) - “Real-time polyp assessment with CADx did not significantly increase the diagnostic sensitivity of neoplastic polyps during a colonoscopy compared with optical evaluation without CADx.”
Real-Time Artificial Intelligence–Based Optical Diagnosis of Neoplastic Polyps during Colonoscopy
Ishita Barua, et al.
NEJM Evid 2022; 1 (6) - BACKGROUND Artificial intelligence using computer-aided diagnosis (CADx) in real time with images acquired during colonoscopy may help colonoscopists distinguish between neoplastic polyps requiring removal and nonneoplastic polyps not requiring removal. In this study, we tested whether CADx analyzed images helped in this decision-making process.
CONCLUSIONS Real-time polyp assessment with CADx did not significantly increase the diagnostic sensitivity of neoplastic polyps during a colonoscopy compared with optical evaluation without CADx.
Real-Time Artificial Intelligence–Based Optical Diagnosis of Neoplastic Polyps during Colonoscopy
Ishita Barua, et al.
NEJM Evid 2022; 1 (6) - “Implementation of AI in cancer screening and clinical diagnosis requires proof of benefits from high-quality clinical studies. Our international multicenter study assessed the incremental gain of a specific CADx AI system for real-time polyp assessment during colonoscopy. Our study indicates that real-time AI with CADx may not significantly increase the sensitivity for small neoplastic polyps. However, CADx may improve specificity for optical diagnosis of small neoplastic polyps and increase colonoscopist confidence with visual diagnosis of polyps.”
Real-Time Artificial Intelligence–Based Optical Diagnosis of Neoplastic Polyps during Colonoscopy
Ishita Barua, et al.
NEJM Evid 2022; 1 (6) - “In conclusion, real-time assessment with CADx did not significantly increase sensitivity for neoplastic polyps during colonoscopy. There are promising signals for increased specificity and improved confidence of optical diagnosis, but our statistical approach precludes us from making any definitive statements about the identification and removal of small rectosigmoid polyps using the colonoscopy system we employed.”
Real-Time Artificial Intelligence–Based Optical Diagnosis of Neoplastic Polyps during Colonoscopy
Ishita Barua, et al.
NEJM Evid 2022; 1 (6)
- “In this cohort study, AI systems showed higher assistance ability in late sessions per half day, which suggests the potential to maintain high quality and homogeneity of colonoscopies and further improve endoscopist performance in large screening programs and centers with high workloads.”
Assessment of the Role of Artificial Intelligence in the Association Between Time of Day and Colonoscopy Quality
Zihua Lu, MD
JAMA Network Open. 2023;6(1):e2253840. doi:10.1001/jamanetworkopen.2022.53840 - IMPORTANCE Time of day was associated with a decline in adenoma detection during colonoscopy. Artificial intelligence (AI) systems are effective in improving the adenoma detection rate (ADR), but the performance of AI during different times of the day remains unknown.
Assessment of the Role of Artificial Intelligence in the Association Between Time of Day and Colonoscopy Quality
Zihua Lu, MD
JAMA Network Open. 2023;6(1):e2253840. doi:10.1001/jamanetworkopen.2022.53840 - Conclusions
“In conclusion, our results suggest that later sessions per half day were associated with a decline adenoma detection. Furthermore, AI systems could eliminate the time-related degradation of colonoscopy quality. In the future, the application of AI systems has the potential to maintain high quality and homogeneity of colonoscopies and further improve endoscopist performance in large screening programs and centers with high workloads.”
Assessment of the Role of Artificial Intelligence in the Association Between Time of Day and Colonoscopy Quality
Zihua Lu, MD
JAMA Network Open. 2023;6(1):e2253840. doi:10.1001/jamanetworkopen.2022.53840
- BACKGROUND AND CONTEXT Colonoscopy is highly operator dependent, and endoscopists miss up to 30% of adenomas. Adjunct computer-aided detection devices, trained to detect preneoplastic polyps, may improve procedure quality through real-time data outputs.
NEW FINDINGS Computer-aided detection colonoscopy increases overall adenoma detection (27% relative increase) in average-risk patients among high-quality US-based endoscopists, reflecting 1 additional adenoma resected among every 4.5 patients screened.
LIMITATIONS Long-term follow-up studies are needed to evaluate the effect of computer-aided detection on clinical outcomes.
IMPACT Computer-aided polyp-detection devices are ready to be introduced into routine clinical practice. Such increases in adenoma detection may reduce the incidence of interval colorectal cancers
Computer-Aided Detection Improves Adenomas per Colonoscopy for Screening and Surveillance Colonoscopy: A Randomized Trial
Aasma Shaukatet al.
Gastroenterology 2022;163:732–741 - “Adenoma detection rate (ADR) is a recognized quality indicator for colonoscopy, but it can lead to a false interpretation of a thorough examination by the “one and done” phenomenon in which after identification of 1adenomatous polyp, the endoscopist may subsequently conduct the examination with less intensity.”
Computer-Aided Detection Improves Adenomas per Colonoscopy for Screening and Surveillance Colonoscopy: A Randomized Trial
Aasma Shaukatet al.
Gastroenterology 2022;163:732–741 - “In the current study, although we found an increase in small lesions of 1 to 4 mm with the use of the CADe device (0.74 vs 0.88; 20% relative increase), we also found an increase in larger lesions of 5 to 9 mm (0.53 vs 0.68; 29% relative increase), primarily driven by increased detection in the proximal colon (proximal colon, 5 to 9 mm lesions: 0.30 vs 0.43, 44% relative increase; distal colon, 5 to 9 mm lesions: 0.23 vs 0.25, 10% relative increase). This reflects an area of considerable interest and clinical relevance, as prior studies on the efficacy of colonoscopy have shown limitations in the prevention of proximal cancers, calling into focus improved detection in the segment and part of colonoscopy quality improvement programs.”
Computer-Aided Detection Improves Adenomas per Colonoscopy for Screening and Surveillance Colonoscopy: A Randomized Trial
Aasma Shaukatet al.
Gastroenterology 2022;163:732–741 - “In summary, in high-volume endoscopists performing screening and surveillance colonoscopies in the United States, this novel CADe device improved APC, an importantquality indicator for colonoscopy that is of relevance to endoscopists, without decreasing THR.”
Computer-Aided Detection Improves Adenomas per Colonoscopy for Screening and Surveillance Colonoscopy: A Randomized Trial
Aasma Shaukatet al.
Gastroenterology 2022;163:732–741