Atherosclerotic carotid lumen segmentation in combined B-mode and contrast enhanced ultrasound images

摘要: Patients with carotid atherosclerotic plaques carry an increased risk of cardiovascular events such as stroke. Ultrasound has been employed a standard for diagnosis atherosclerosis. To assess atherosclerosis, the intima contour artery lumen should be accurately outlined. For this purpose, we use simultaneously acquired side-by-side longitudinal contrast enhanced ultrasound (CEUS) and B-mode (BMUS) images exploit information in two imaging modalities accurate segmentation. First, nonrigid motion compensation is performed on both BMUS CEUS image sequences, followed by averaging over 150 time frames to produce improved signal-to-noise ratio (SNR). After that, segment from these using novel method based dynamic programming which uses joint histogram pair distinguish between background, lumen, tissue artifacts. Finally, obtained improved-SNR mean transformed back each frame original sequence. Validation was done comparing manual segmentations independent observers automated 9 arteries 7 patients. The root square error 0.17±0.10mm average segmentation 0.19±0.06mm. In conclusion, present robust overcomes complexity anatomical structures, noise artifacts echolucent exploiting combined modality.