Accuracy of a semi-automatic technique for segmentation of the carotid arteries from 3D ultrasound images

A semi-automatic 3D segmentation technique has recently been developed and is being used to extract carotid artery lumen from 3D ultrasound images. The accuracy of the segmentation technique has been assessed by comparing 3D surfaces derived by the semi-automatic technique, to surfaces constructed from manual segmentation. Preliminary results show that for a sample free-hand 3D ultrasound image, typical manual and automatic surfaces are separated from each other with a mean error of 0.4 mm.