An automated system designed for the interpretation of coronary CT angiography (CCTA) helps radiologists determine which cases are high priority and should be read first, according to a study presented today in Boston at the 2009 annual meeting of the American Roentgen Ray Society (ARRS).
Researchers from Thomas Jefferson University in Philadelphia said the study included 100 CCTA studies submitted for evaluation using a new automated system (COR Analyzer II, Rcadia Medical Imaging). Interpretation of CCTA studies with the application was compared with a consensus interpretation of two experienced cardiac imagers. The 100 CCTA exams were selected from a series of more than 1,000 cases performed on a 64-slice CT system from Philips Healthcare during 2006-2007.
"The software detected 10/13 patients with artery stenosis 50 percent or greater and correctly identified 59/80 patients with less than 50 percent artery stenosis," said Shaoxiong Zhang, MD, from the department of radiology at Thomas Jeffereson. "The negative predictive value for the study population was high, 95 percent."
Among the 100 cases submitted for evaluation, Zhang reported that the software failed to process seven cases, leaving 930 segments in 930 patients for the current analysis. They also found that the software correctly identified stenosis in 8/22 coronary segments (sensitivity, 36 percent) and correctly identified less than 50 percent stenosis in 869/908 segments (specificity, 96 percent). Segments in which a stenosis was missed included the left main coronary artery (one patient), the left anterior descending artery (eight patients), the left circumflex artery (three patients) and the right coronary artery (two patients).
"The new technology offers promising results for the diagnosis of stenosis in the major coronary arteries, although "more research needs to be done to further improve the accuracy of the software in identifying significant disease," said Zhang. "Study interpretation using the automated interpretation of CCTA studies will allow for shorter reading times in negative cases and reduced response times in positive cases," he said.
"In urban areas there are high demands to provide 24/7/365 day coverage and cardiac imaging may not be available during nights and weekends. Rural area, outpatient diagnostic centers may be lacking cardiac imaging experts. Preliminary study interpretation using the new rapid, automated interpretation system may potentially allow for better prioritization of cases for evaluation by the expert leader," Zhang said.