Improved vessel morphology measurements in contrast-enhanced multi-detector computed tomography coronary angiography with non-linear post-processing.

Multi-detector computed tomography (MDCT) permits detection of coronary plaque. However, noise and blurring impair accuracy and precision of plaque measurements. The aim of the study was to evaluate MDCT post-processing based on non-linear image deblurring and edge-preserving noise suppression for measurements of plaque size. Contrast-enhanced MDCT coronary angiography was performed in four subjects (mean age 55 +/- 5 years, mean heart rate 54 +/- 5 bpm) using a 16-slice scanner (Siemens Sensation 16, collimation 16 x 0.75 mm, gantry rotation 420 ms, tube voltage 120 kV, tube current 550 mAs, 80 mL of contrast). Intravascular ultrasound (IVUS; 40 MHz probe) was performed in one vessel in each patient and served as a reference standard. MDCT vessel cross-sectional images (1 mm thickness) were created perpendicular to centerline and aligned with corresponding IVUS images. MDCT images were processed using a deblurring and edge-preserving noise suppression algorithm. Then, three independent blinded observers segmented lumen and outer vessel boundaries in each modality to obtain vessel cross-sectional area and wall area in the unprocessed MDCT cross-sections, post-processed MDCT cross-sections and corresponding IVUS. The wall area measurement difference for unprocessed and post-processed MDCT images relative to IVUS was 0.4 +/- 3.8 mm2 and -0.2 +/- 2.2 mm2 (p < 0.05), respectively. Similarly, Bland-Altman analysis of vessel cross-sectional area from unprocessed and post-processed MDCT images relative to IVUS showed a measurement difference of 1.0 +/- 4.4 and 0.6 +/- 4.8 mm2, respectively. In conclusion, MDCT permitted accurate in vivo measurement of wall area and vessel cross-sectional area as compared to IVUS. Post-processing to reduce blurring and noise reduced variability of wall area measurements and reduced measurement bias for both wall area and vessel cross-sectional area.