AUTOMATED COMPARISON OF 3D IMAGES

摘要

A method and apparatus for automated comparison of 3D images, such as specific shapes in images, for example, for detecting vascular changes and aneurysm growth. The method comprises an adaptive geodesic active contour (GAC) approach which can be performed in a single pass, or across multiple iterations. The method uniquely utilizes shape to identify targets in two 3D images so that images can be compared if there is any shape feature. A method is also described for iterative parameter image construction, which beneficially removes false positives. The technology is particularly well-suited for use in comparing geometric changes of aneurysms, tumors, thromboses, inflammations.

主权项

1. A method for performing vascular segmentation of 3D images with adaptive parameter setting in a geodesic active contour (GAC) segmentation process, the method comprising:
(a) acquiring 3D input images containing voxels as units of graphic information defining points in 3D space; (b) performing active contour segmentation on said 3D input images, wherein said segmentation evolves a closed curve/surface through a combination of different forces; (c) mapping gradient and speed function by using shape information as prior knowledge to guide the mapping between the gradient and speed function; (d) enhancing gradients at boundaries of vessel-like regions in the 3D input images; (e) suppressing gradients inside or outside vessel-like regions in the 3D input images; and (f) outputting segmentation contours for vessels found in the 3D input images; (g) wherein said method is performed by executing programming on at least one computer processor, said programming residing on a non-transitory medium readable by the computer processor.