In the hours leading up to lunch, the talk here at Stanford MDCT turned to CT colonography, the hotly contested (well, by CMS, anyway) technique for virtually scanning the colon for cancer. C. Dan Johnson, MD, kicked off the CTC presentations with a look at the five requisites for performing high-quality CTC:
* Patient preparation
* Colon insufflation
* Scanning technique
* Training of radiologist
Abe Dachman, MD, of the University of Chicago Medical Center, took the podium to elaborate on colon insufflation, sharing a few "tricks of the trade" with the assembled crowd. Here's one that I thought was particularly smart: spray lidocaine onto the balloon lubricant to reduce bowel pain. And, though Dachman prefers mechanical insufflation, he recommended that everyone keep equipment on hand for manual insufflation, just in case. Finally, he suggested that on low-volume days physicians supervise techs to help reinforce training.
Jacob Sosna, MD, presented on a welcome alternative to the rigorous patient preparation process: spectral CT cleansing. Studies have indicated that patients would be more likely to get screening if they didn't have to go through the bowel catharsis process -- surprise surprise. And electronic cleansing has other benefits: it can be used to digitally remove tagged solid debris and contrast-labeled fluid. As Sosna's example images beautifully displayed, dual-energy cleansing minimizes artifacts. "Non-cathartic CTC with spectral CT electronic cleansing is feasible with improved visualization of the colon compared with standard electronic cleansing," he said.
In a later session, Dachman returned to touch on the issue of radiologist training. New research indicates that fifty studies may not be enough to sufficiently train a radiologist in CTC. "Learning still occurs after your first fifty cases," he said, confirming what many in the audience probably knew from personal experience. "Formal CTC training is critical."
Dachman also provided a summary of recent studies on the sensitivity and specificity of CAD for CTC. Though the software "spell-checker" has been shown to accurately identify polyps that human readers missed, it also comes with decreased specificity and increased reading time, begging the question, will it ever truly be an integrated component of radiologists' CTC workflow?
For Dachman, the answer is yes. "CAD might increase reading time, but it also improves diagnostic performance," he said. "Even if it's a small increase, it's important. CAD is still rapidly developing, and will be very important in moving CTC to national use."