Enhancing Multiple Sclerosis Lesion Segmentation in MRI Images
Biediger, Dan 1974-
MetadataShow full item record
Multiple sclerosis (MS) is an autoimmune disease of the central nervous system that causes damage to the insulating myelin sheaths around the axons in the brain. It affects over 2.5 million people world-wide. The disease progresses at different rates in different people and can have periods of remission and relapse. A fast and accurate method for evaluating the number and size of MS lesions in the brain is a key component in evaluating the progress of the disease and the efficacy of treatments. MS lesion segmentation usually requires the expertise of a trained physician. Manual segmentation is slow and difficult and the results can be somewhat subjective. While many automated methods exist, they do not provide sufficiently accurate segmentation results. There exists a need for a robust, fast, and accurate method for automatically segmenting MS lesions. This thesis presents the results of an effort to improve the segmentation results of an existing system for lesion segmentation in MRI images. It includes two different strategies to improve the segmentation results by addressing opportunities missed in the existing approach. The first strategy leverages the current processing system at a granularity finer than the whole-brain to detect lesions at a local level. The existing system makes global estimates on the tissue intensities. Because these intensities vary across the brain, the global assumption provides inaccurate estimates in some cases. The first improvement combines a series of local results to produce a whole-brain lesion segmentation. This approach better captures the local lesion properties and produces encouraging results, with a general improvement in the detection rate of lesions. The second method looks at the individual voxel level and the local intensity neighborhood. As a post-processing method, it selects seed points from the results of the previous step. It uses a region growing method based on cellular automata to expand the lesion areas based on a local neighborhood similarity in intensity. While it provides some benefit, it is sensitive to initial conditions and the results depend on the implementation details.