Liver: Three-Dimensional Imaging: Will it Come of Age in the New Millenium

Elliot K. Fishman, M.D
Professor of Radiology and Oncology
Johns Hopkins University School of Medicine

Over the past two decades we have seen numerous improvements in CT scanner technology bringing us from an era where we obtained a slice every 10-15 seconds to an era where a study is completed in 10-15 seconds. Whether it be scanner resolution, selectable collimation, possible mAs and kVp choices or tube and generator capacities for essentially endless continuous scanning we have seen unparalleled progress. In contrast, the techniques we use to look at these images has changed very little over these past two decades. Axial images have always been the traditional mode for reviewing images and as we enter the millennium, most radiologists are still reading films (hard copy) while others are beginning to transform into the era of electronic or soft copy reading. Yet, both of these techniques still focus on axial images. In the past, numerous authors have described the value of looking at images in other perspectives including the use of multiplanar reconstruction with coronal, sagittal and oblique imaging, holography, as well as various forms as three-dimensional imaging. Modified 3D techniques including perspective rendering, virtual angioscopy, fly thru’s and fly arounds have all been explored. Many articles have shown very promising results and have shown that the impact of these newer display techniques goes beyond diagnosis to patient management and outcomes. Yet, both in most academic practices and in the general radiologic community, little has changed in terms of viewing.

As we stand on the dawn of Multidetector CT scanning where literally hundreds to thousands of images can be generated on a single patient study, we are approaching I believe what Andy Grove, co-founder of Intel calls a "strategic inflection point." We are at a point where there is great promise for change and change is in fact inevitable. The only question is how we deal with this change and how we use it for advantage because nothing will ever again be quite the same. The promise of 3D imaging and display is not just in the select case be it a pelvic fracture, a CT renal angiogram, or virtual colonography study. Rather, its promise lies in becoming the primary modality for looking at images as they are generated by the scanner. No longer will the radiologist view or cine through a series of slices but will instantly be inside or around the volume of information provided for that patient. The potential of this type of volume imaging are many and would include increased speed of review, decreased cost of documentation of information, increased accuracy and better patient management decisions. While these are all lofty goals and seem more like a political statement than a clinical reality, it is our belief that we are at that moment when change is inevitable.

The area where this may in fact place first is in the emergency room. The Emergency Room has become a key triage zone of the hospital for radiologist and nonradiologist alike. Decisions as to patient management need to be made in a timely and accurate fashion in order to optimize both patient care and minimize cost to the system. It is in this environment where the rapid acquisition and analysis of information can have most profound impact. In a patient who has fallen down a flight of steps and undoubtedly has a fracture it will be quicker and more accurate to obtain a CT scan of the injured region and view this as a volume dataset. Not only can the fracture be defined with more accuracy but it will have impact on patient management decisions. The images acquired by multidetector scanners are true volumes and are best displayed as volumes. Newer workstations must have real-time capabilities for looking at these large volumes of information. The next generation of workstations based around processors like the Intel Merced chip (Intel Corp., Santa Clara, CA) can when running in parallel provide gigabyte processing speeds which will allow for larger datasets to be viewed interactively and with better rendering techniques.

Volume rendering technique has been described in numerous articles and does provide the capability of imaging any tissue type and is ideal in a situation where a volume is to be viewed interactively. A fracture of bone, concurrent injury of vascular structures and soft tissue can all be assessed in a single examination in what may be a 5-15 second period. Elimination of unnecessary studies, decreased examination time and increased accuracy are the potential benefits of such an approach. Similarly in a patient with suspected vascular pathology such as pulmonary embolism, aortic aneurysm or dissection, a single CT scan provides the entire volume of information that is needed to make an accurate diagnosis of the presence or absence of pathology, as well as to determine extent of disease and direct patient care. The CT data can be used to determine triage (surgical vs. medical care), and if surgery is needed, decide on the procedure (open repair vs. endovascular stent) and if a stent is needed to help design its dimensions. The ability to view these volumes of highly accurate data will surely go a long way to eliminating duplicate or unnecessary studies. The advantages of a true volume approach are endless and their impact will be significant.

In order to do this type of imaging, a radiologist skilled in the analysis of volume datasets will be necessary. This may place more burden on the radiologist as the need for this information will be in what many would call ‘Internet time.’ It will be impossible to take advantage of this system unless the radiologist is present at the site of that acquisition or be electronically connected to that site. Potential changes that may free us from the need for an absolute physical presence and allow us to have a "virtual" presence include the increase in the bandwidth of the Internet including Internet 2, new innovative image compression schemes, and high speed networks using DSL or cable modem. There will be low cost (eventually) gigabyte speed systems which will allow physicians to rapidly view the images and consult on these datasets with the referring physician’s a few feet or a thousand miles away.

There is no doubt that the developments on the imaging acquisition side, the 3D rendering side, and image transfer protocol will have an impact on the delivery of services in the second generation of radiology. It is imperative that radiology designs these pathways rather than becoming a victim of others who design them to function the way they believe radiology should be done. As Alan Kay (Apple Fellow) said, "The best way to predict the future is to invent it." Our goal is to understand the needs of the radiologic environment, master the new technologies and help create this brave new world.