Automated coronary CT angiography software demonstrates potential in emergency triage of patients with chest pain

Physicians at Beth Israel Deaconess Medical Center, Boston reported results of a study demonstrating the potential for automated analysis of coronary CT angiography (CCTA) by Rcadia's COR Analyzer SystemR to assist in rapidly ruling out coronary artery stenosis in hospital emergency department (ED) patients with chest pain. The ability of the automated system to enhance the use of CCTA has potential to significantly reduce unnecessary utilization of coronary observation beds in the low to moderate probability coronary artery disease (CAD) patient. The study was presented in Chicago at the 2009 annual meeting of the Radiological Society of North America (RSNA).

"In recent years, CCTA has proven to be an effective, non-invasive procedure for coronary artery analysis," said Girish Tyagi, M.D., the study's principal investigator, radiologist at the medical center and Instructor at Harvard Medical School. "However, coronary CT angiography is under-utilized in the ED because the procedure relies on expert readers who may not be immediately available during 'off hours'. In our study, we asked whether this fully automated software tool could assist in patient triage by comparing the automated results with those of expert readers."

"The results of this initial study, particularly the system's high negative predictive value of 98 percent, suggest that this automated system can be a very useful tool," he continued. "The system may provide non-expert readers with confidence to rule out significant stenosis, leading to decreased length of stay for patients in the ED and improved throughput in the ED." He added that positive results from the automated analysis require further interpretation by an expert reader.

In the retrospective study, images from CCTAs performed on 115 low to intermediate risk patients who entered the ED with suspected CAD were studied with the COR Analyzer System. The researchers compared analyses of the CCTA studies by the COR Analyzer System with the interpretation of the studies by consensus opinion of two expert readers who served as the gold standard. For 100 analyzable studies, the automated results from COR Analyzer yielded a negative predictive value of 98 percent, meaning that 98 percent of the patients determined by the COR Analyzer System to not have significant stenosis (>50 percent reduction in lumen area), were in agreement with the expert readers. The COR Analyzer System identified five of six patients determined by the expert readers to have significant stenosis, for a sensitivity of 83 percent. The specificity was 82 percent.

In addition to Dr. Tyagi, the study's authors are: Atif Khan, M.D., Faisal Khosa, M.D., Sheryar Sarwar, M.D., Maya Tyagi, Marc Camacho, M.D., and Melvin E Clouse, M.D., all of Beth Israel Deaconess Hospital, Boston.