MRI, Big Data, and Artificial Intelligence: Rewards vs Risks

Multiple sclerosis (MS) is a chronic, inflammatory, neurodegenerative condition of the CNS characterized by highly variable clinical features. This variability makes it difficult for neurologists to predict the clinical course in each individual and, as a result, results in significant uncertainty for patients. Disease variability also makes prediction of treatment response a challenge, posing a major barrier to implementation of personalized medicine approaches in MS. These are major issues, as there are up to 3 million people with MS worldwide and approximately 1 million in the United States,1 and many go on to develop significant physical disability or disabling symptoms. We have an incomplete understanding of the reasons for this heterogeneity and variability. The pathology of MS is heterogeneous and involves inflammatory and neurodegenerative processes.2 In MS lesions, there are varying amounts of myelin content, with some patients and lesions demonstrating little demyelination.3 Similarly, the inflammatory profiles of patients with MS show significant heterogeneity, with T cells,4 B cells,5 and microglial cells6 involved. To gain further understanding, we need tools that help us understand the disease process, improve our diagnostic accuracy, better predict disease course, and select optimal treatments.

Read article at journal's website

Related Articles


Your email address will not be published.