EU Rad: CT computer-aided detection app may rule out stenosis
New computer-aided detection software for coronary CT angiography (CCTA) could help physicians rule out narrowing of the coronary arteries in patients with low to moderate risk of coronary artery disease, according to a study published online Nov. 5 in European Radiology.

Joseph Schoepf, MD, professor of radiology and cardiology, director of cardiovascular imaging at the Medical University of South Carolina in Charleston, S.C., and colleagues evaluated a CCTA computer-aided detection system, COR Analyzer, which was developed by Rcadia Medical Imaging of Auburndale, Mass.

In the study, the researchers enrolled 59 patients without previously diagnosed coronary artery disease, who had been referred for invasive coronary catheter angiography due to atypical chest pain or an abnormal cardiac blood flow study.

Nineteen of the 59 patients had significant coronary stenosis, while significant stenosis was ruled out in the other 40 patients based on invasive coronary catheter angiography, according to the authors. The software correctly identified all 19 patients with significant stenosis in any vessel, and correctly excluded significant stenosis in 26 of 40 patients. 

Schoepf and colleagues reported that the system’s negative predictive was 100 percent on a per patient basis, while its sensitivity, specificity and positive predictive values were 100 percent, 65 percent and 58 percent respectively, compared with invasive coronary catheter angiography.

According to the authors, use of the system as a second reader could give a physician in an emergency room setting “reassuring verification of his or her exclusion of significant stenosis based on a normal or near-normal CCTA study.”

"The findings of this study suggest that if used as a second reader the high negative predictive value demonstrated by this system may enhance the confidence and efficiency of excluding significant stenosis based on a normal or near-normal CCTA study," Schoepf noted.