Accuracy and Reproducibility of Blood Clot Burden Quantification With Pulmonary CT Angiography

Abstract

The purpose of our study was to assess the accuracy and reproducibility of clot burden quantification with pulmonary CT angiography (CTA). A semiautomated program was developed for segmentation and volumetric quantification of pulmonary embolus with pulmonary CTA. The accuracy of this measurement method was assessed using two pulmonary embolus phantoms. Reproducibility of the measurement method was assessed using clinical pulmonary CTA in 30 patients (16 women, 14 men; mean age, 62 years) with pulmonary embolism (PE). Two observers segmented and measured the volume of blood clot from pulmonary CTA images twice at two separate sessions. Accuracy was evaluated by the relative volume measurement error. Intra- and interobserver reliability were evaluated using intraclass correlation coefficient (ICC); agreement between measurements within and between the two observers was assessed using Bland-Altman analysis. Mean relative measurement error from the two phantoms was less than 1% for both observers. A total of 60 emboli were measured from the 30 patients. The intraobserver ICC was 0.990 for observer 1 and 0.999 for observer 2; interobserver ICC was 0.994 for session 1 and 0.989 for session 2. ICC for all four clot measurements was 0.988. Mean volume measurement difference for intraobserver agreement was 0.9% for observer 1 and 0.3% for observer 2, and interobserver agreement was -5.1% for session 1 and -5.8% for session 2. Blood clot burden can be quantified with a high degree of accuracy and reproducibility from pulmonary CTA images using a semiautomated segmentation method.