Artificial Intelligence Enhances Detection of Multiple Sclerosis Progression for Better Early Treatment

Researchers at Uppsala University have developed an innovative AI model that significantly improves the ability to identify the progression of multiple sclerosis (MS), potentially allowing for more timely and personalized treatment. MS is a chronic neurological disorder affecting about 22,000 individuals in Sweden, typically beginning with the relapsing-remitting form (RRMS), characterized by episodes of deterioration followed by periods of stability. Over time, many patients transition to the secondary progressive form (SPMS), where symptoms worsen steadily without clear relapses. Recognizing this transition promptly is crucial, as the two forms require different treatment strategies. Currently, clinicians typically diagnose this shift approximately three years after it actually occurs, which can delay the start of effective intervention.

The new AI model leverages clinical data from over 22,000 patients registered in the Swedish MS Registry. It analyzes information gathered during regular healthcare visits, such as neurological examinations, MRI scans, and ongoing treatments. By detecting patterns within this data, the AI can accurately distinguish whether a patient still has RRMS or has transitioned to SPMS, with a confidence level indicated for each assessment. This transparency allows healthcare providers to gauge the reliability of the AI's conclusions.

Published in npj Digital Medicine, the study reports that the AI model identified the transition to secondary progressive MS earlier than documented in patients' medical records in nearly 87% of cases, achieving an overall accuracy of approximately 90%. This advancement means that diagnoses can be made sooner, enabling clinicians to adjust treatments earlier and slow disease progression, ultimately improving patient outcomes. Moreover, the model could be instrumental in selecting suitable candidates for clinical trials, fostering the development of more targeted therapies.

An anonymized, open-access version of the model is available via a web platform, supporting researchers worldwide in their efforts to enhance MS care. This development marks a significant step towards more personalized and timely management of multiple sclerosis.

Source: https://medicalxpress.com/news/2025-04-ai-patient-multiple-sclerosis-early.html