“We’ve made progress in improving the hardware, the imaging protocols, and the processing tools that allow us to better visualize the spinal cord. These advances could help us detect lesions that are difficult to see with conventional imaging.”
AI may further support MRI analysis by helping process images and identify or segment lesions. However, Cohen-Adad emphasized the need to account for differences in imaging equipment, acquisition protocols, and patient populations. Algorithms developed using data from one setting may not perform equally well in another, underscoring the importance of validating these tools in diverse clinical environments.
To discuss these developments, NeurologyLive® spoke with Cohen-Adad about the evolving role of MRI in MS, the challenges of spinal cord imaging, and the opportunities and limitations of applying AI to clinical imaging. He also emphasized the importance of collaboration among researchers, clinicians, and MRI manufacturers to support the development and responsible implementation of imaging tools that can translate into real-world clinical practice.