A University at Buffalo-led team reported in Communications Medicine that AI can identify previously invisible cortical lesions in multiple sclerosis by reprocessing existing MRI datasets. The work targets a known limitation of conventional imaging, where gray-matter (cortical) lesions have been difficult to quantify. The researchers said the approach enables clinicians and researchers to see cortical lesions using multi-contrast post-processing and deep learning. The team’s results are positioned to support both MS research and clinical decision-making, including tracking disease activity linked to disability and cognitive impairment. The report underscores the growing use of AI in imaging analytics as a “software” pathway to unlock additional value from legacy scans already embedded in cohorts.