Detection and measurement of multiple sclerosis brain lesions from magnetic resonance images using image processing techniques

Multiple Sclerosis (MS) is a disease of the Central Nervous System, i.e. the brain and the Spinal Cord. Magnetic Resonance Imaging (MRI) is a medical imaging technique proven to be a sensitive method for monitoring the progression and changes of this disease due to treatment protocols. Quantification of the severity of the disease through estimation of MS lesion volume via MR imaging is vital for understanding and monitoring the disease and its treatment. This thesis presents a refined methodology called Lesion Detection System for the detection of MS lesions with modest operator involvement using intensity-based techniques. The proposed system consists of two major functions, segmentation and calculation. The problem of segmentation consists of the identification and extraction of substructures of the brain, and the creation of partial volume solutions of the tissues under consideration for lesion detection. Calculations involve the volumetric measurements of the results obtained by the segmentation function.

Based on the qualitative and quantitative results obtained by two human raters, it is concluded that the proposed system demonstrates acceptable intra-rater reproducibility in lesion detection. The quantitative test results are also found to be encouraging enough to warrant further study in the proposed technique of MS lesion detection.