Lesion assessment criteria can contribute to higher levels of inter-observer agreement of measurements from optical coherence tomography (OCT) images, according to the results of research presented Oct. 31 at TCT 2016 and simultaneously published in the Journal of the American College of Cardiology.

Annapoorna Subhash Kini, MBBS, FACC, et al., assessed the inter-observer variability of fibrous cap thickness by intraclass correlation coefficient to determine whether OCT image measurements in vivo could be reproduced. 170 OCT images were analyzed by two independent observers and the main sources of variability were incorporated into lesion assessment criteria. A third independent observer then interpreted all 170 OCT images using the developed criteria.

The results of the investigation found that careful consideration of OCT image features led to significantly higher levels of inter-observer agreement. The original intraclass correlation coefficient of the fibrous cap thickness was 0.56 (95 percent confidence interval: 0.38 - 0.69). After the lesion assessment criteria was developed, the intraclass correlation coefficient of the fibrous cap thickness was 0.88 (95 percent confidence interval: 0.80 - 0.93).

The authors of the study note that "the inter-observer variation can be partially resolved by development of standard interpretation algorithms."

Keywords: Transcatheter Cardiovascular Therapeutics, Diagnostic Imaging, Algorithms, Cardiology, Confidence Intervals, Observer Variation, Publications, Research, Tomography, Optical Coherence, Angiography, Diagnostic Imaging

< Back to Listings