Advanced visualization reach continues into mainstream

June 23, 2010

There was a time when attendees of what was then known as the Stanford Multidetector CT Symposium workstation face-off went expecting to see a train wreck. In the past, that often occurred: a workstation would freeze, an operator would get lost, a network would crash.

There was a time when attendees of what was then known as the Stanford Multidetector CT Symposium workstation face-off went expecting to see a train wreck. In the past, that often occurred: a workstation would freeze, an operator would get lost, a network would crash.

But this year, the first billed as a meeting of the International Society for Computed Tomography (ISCT), none of these things happened. Seven vendors, both companies representing the major modalities and independent firms manufacturing advanced visualization products, walked through three sets of clinical operations without a hitch. It was the smoothest presentation ever, observed my colleague Greg Freiherr, a veteran of many of these events.

It illustrates the situation with advanced visualization today. It's becoming part of the fabric of radiology. What we called advanced visualization yesterday will be called just visualization tomorrow, said Dr. David Hirschorn, an editorial advisor for Diagnostic Imaging, in an article that appeared in June.

In some procedures it's already become indispensable. For example, you could probably read CT colonography in cross-sectional images, but how often do you see it without the ubiquitous pink endoscopic views generated by advanced visualization software? Ditto the filleted colon images that are now usually available. If you do read CT colonography, how comfortable would you feel without the advanced visualization tools?

This will become the case more and more as we move forward. Hirschorn observes that PACS vendors, long comfortable with tools such as multiplanar reformatting (MPR) and maximum intensity projection (MIP), are also incorporating such functions as vessel tracing for angiographic studies, virtual endoscopic views for flying through hollow viscera like the trachea or the intestines, and surface rendering for complex fractures such as those of the tibial plateau-including the ability to remove unwanted bones that obscure the view, like the femoral condyles.

Concurrent with these developments have been efforts to automate the advanced visualization process. Besides being important for time-pressed radiologists, these automated AV algorithms have boosted what PACS and workstations are able to do, increasing reproducibility and reducing dependence on operator skill, as Freiherr observes in this month's cover story. We can expect this process to continue with further adoption of automated algorithms.

This process raises a few issues. One is training. Vendors who supply AV systems are offering courses to explain their use. Users who anticipate adopting these systems would be wise to consider the training aspects.

Another is how these systems fit into workflow. Approaches vary: some facilities are still relying on 3D labs for their AV reconstructions. Others say that radiologists should perform the AV reconstructions, preferably through their PACS. Hirschorn argues that the incorporation of AV into PACS will become a key differentiator among different systems.

Finally, there is the question of validating the imaging data produced by AV algorithms. You can do cross-system comparisons, but what are the standards for ground truth? In a cardiac case at the ISCT Workstation Face-Off, three systems recorded right ventricular ejection fraction values near 45 while the other four ranged from a low of less than 30 to a high above 50, although other measures were fairly consistent.

All these issues, however, are solvable. The fact is that advanced visualization is finding a solid place in imaging practice. Radiologists should follow developments in advanced visualization closely and be prepared to adopt them as their clinical value becomes apparent.