CT flexes muscle in coronary disease detection

November 29, 2005

New research conducted at the University of Erasmus in Rotterdam may erase any lingering doubts about the superiority of 64-slice CT in detecting coronary artery disease.

New research conducted at the University of Erasmus in Rotterdam may erase any lingering doubts about the superiority of 64-slice CT in detecting coronary artery disease.

Dr. Francesca Pugliese, a radiology researcher at the university, studied 153 symptomatic patients and found that 64-slice CT is faster, more comprehensive, and more accurate than 16-slice; ultrafast, second-generation 16-slice; and four-slice CT for evaluating the coronary arteries for diagnosing stenoses. The research was presented Sunday at the RSNA meeting.

Three sets of 51 patients each were scanned on each device. Significant disease was defined as 50% or greater stenoses. Two radiologists read the CT studies, and a cardiologist provided lesion classification from follow-up coronary angiography.

Considering coronary branches measuring >2 mm in diameter, the sensitivity, specificity, positive predictive value, and negative predictive values for 16-slice CT were 92%, 93%, 64%, and 99%, respectively. For ultrafast 16-slice CT of the same region, researchers found sensitivity, specificity, PPV, and NPV to be 95%, 98%, 87%, and 99%, respectively. For 64-slice CT, sensitivity, specificity, PPV, and NPV were 92%, 93%, 69%, and 99%, respectively.

Vessel wall calcification was less extensive in the ultrafast 16-slice group. In the 64-slice CT group, including the assessable

Previously reported problems with visualization of small, distal segments for 16-slice, eight-slice, and four-slice CT appear to be largely solved by moving up to 64-slice scanning. Only 2% of the segments were judged to be of low diagnostic confidence with the 64-slice device. In comparison, poorly imaged segments accounted for 26% of the total for the four-slice CT population. Seven percent of the segments were scored with low user confidence, although all segments assessed with conventional angiography were deemed accessible.

The average scanning time on the 64-slice device was 313 seconds. That time is 64 seconds faster than a second-generation 16-slice CT device capable of half-second rotations, 117 seconds faster than a first-generation 16-slice CT performing one-second rotations, and 187 seconds faster than a four-slice CT scanner.

"We can conclude that 64-slice CT might broaden the application of multidetector CT coronary angiography in terms of the anatomical information it can protect and its clinical implementation," Pugliese said.