This paper describes the development of MS-Analyze, a system framework to analyze and detect patterns in brain pathology of multiple sclerosis (MS) as the disease progresses over time. We are building the system to collect, analyze and integrate disparate data extracted from observing the behavior of MS lesions and surrounding pathology on magnetic resonance imaging (MRI) over time. Multiple sclerosis (MS) is a brain disease that affects over 250,000 people in the USA alone. Various MRI sequences are used to monitor brain changes during the natural progression of the disease, and as different drug treatments are explored to slow the disease. The outcome is a set of disparate data streams that need to be correlated efficiently to discover patterns of MS pathology and plan treatment. However, MS data analysis faces the same computational problems as many other scientific domains with heterogeneous data streams: the need for integration of and access to large amounts of data, beyond what is normally available to any one given laboratory. MS-Analyze addresses both of these challenges by combining data collection, data fusion, data analysis, and secure data sharing. MS is a good application to demonstrate the system because it offers rich data that challenge the system development.