Key result
Automated segmentation combining MRI and CTA features with linear discriminant classification and outlier rejection achieved volume errors of 0.9% for calcification, 12.7% for fibrous, and 12.1% for necrotic tissue compared to histology.
Why the study?
Does combining MRI and CTA data with class label uncertainty incorporation improve the accuracy of atherosclerotic plaque component segmentation compared to single modalities?
Population
13 male patients scheduled for carotid endarterectomy with available in vivo MRI, CTA, and corresponding…
Comparison
Supervised voxelwise classification combining in… vs Segmentation using only MRI features, only CTA…
Design
Other
Authors
Loading...
Supports multimodal plaque segmentation feasibility; leaves open clinical validation before practice adoption.
Observational (n=13)
No
Does combining MRI and CTA data with class label uncertainty incorporation improve the accuracy of atherosclerotic plaque component segmentation compared to single modalities?
Combining MRI and CTA features with linear discriminant classification and outlier rejection provides accurate segmentation of calcified, fibrous, and lipid-rich necrotic plaque components in carotid arteries.
Engelen et al. (2014) conducted an observational in Carotid artery atherosclerosis (n=13). Combined MRI and CTA automated segmentation vs. Histology and mCT (ground truth) was evaluated on Volume error per vessel for calcification, fibrous, and necrotic tissue. Automated segmentation combining MRI and CTA features with linear discriminant classification and outlier rejection achieved volume errors of 0.9% for calcification, 12.7% for fibrous, and 12.1% for necrotic tissue compared to histology.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: