A Study on the Suitability of Applying Active Contour Evolution Models in Segmenting and Delineating Boundaries in Medical Images

Active contour model has been used to segment and delineate boundaries in different types of medical images. It is a popular class of image segmentation and boundary delineation method due to its ability to fit a curve to an object boundary by iteratively expanding or contracting its boundary estimate. In this study, we provide an analysis of the suitability of applying active contour models in segmenting and delineating boundaries in medical images. As a case study, we used morphological Chan-Vese and morphological Geodesic Active Contour models to improve the accuracy of manually developed label images of axial view lumbar spine MRI images. The images contain labels for intervertebral disc, posterolateral element, and thecal sac regions and the context of the experiment is to improvement the segmented regions accuracy so that it can be used to diagnose lumbar spinal stenosis. Our experiment shows that morphological Geodesic Active Contour model performs better than morphological Chan-Vese, however both models still produce worse segmentation results than are needed for lumbar spine stenosis detection algorithm.