Multiple sclerosis (MS) is a chronic, inflammatory disease affecting the central nervous system. Approximately 22,000 people in Sweden live with MS, and most begin with the relapsing-remitting form (RRMS), characterized by episodic deteriorations followed by stable periods. Over time, many patients transition to secondary progressive MS (SPMS), where symptoms worsen steadily without clear breaks. Detecting this transition early is crucial, as different forms of MS require distinct treatments. However, the current diagnostic process often detects this shift an average of three years late, leading to ineffective treatments. Researchers at Uppsala University have developed an AI model to improve the early detection of SPMS, significantly enhancing treatment decisions and outcomes. The model leverages clinical data from over 22,000 patients in the Swedish MS Registry, using information such as neurological test results, MRI scans, and ongoing treatments. By analyzing these patterns, the AI can predict with 90% certainty whether a patient's condition has transitioned from RRMS to SPMS. Kim Kultima, the lead researcher, explains that what sets this model apart is its ability to convey confidence levels in its assessments. "This not only helps doctors make more informed decisions but also ensures they understand the reliability of the AI's conclusions," Kultima notes. The model has demonstrated impressive accuracy, correctly identifying or predicting the transition to SPMS in nearly 87% of cases, outperforming traditional diagnostic methods. The earlier detection of SPMS through this AI model allows for timely adjustments in treatment, potentially slowing disease progression and reducing the risk of patients remaining on ineffective medications. According to the researchers, early intervention can significantly improve patient outcomes and quality of life. Moreover, the AI model holds promise beyond clinical diagnostics. It can help identify suitable candidates for clinical trials, contributing to the development of more effective and personalized treatment strategies. "In the long term, this could revolutionize MS management and research," Kultima adds. The open, anonymized version of the model is now accessible to researchers via the web service: https://msp-tracker.serve.scilifelab.se. This accessibility ensures that the scientific community can validate and build upon the model, accelerating advancements in MS research and patient care. Industry experts view this development as a significant breakthrough in the field of neurology and AI-driven healthcare. The AI model's high accuracy and real-world applicability are particularly noteworthy, addressing a critical gap in MS diagnosis and management. Companies like Biogen, which specializes in MS treatments, are closely watching this innovation, as it could enhance their clinical trial recruitment processes and lead to better patient stratification. Uppsala University, a leading institution in scientific research, is known for its contributions to healthcare and biotechnology. This study, published in npj Digital Medicine, underscores the university's commitment to leveraging advanced technologies to improve patient outcomes. The collaboration with SciLifeLab, a national infrastructure for molecular biosciences, further highlights the robust support and expertise behind this groundbreaking research. In summary, the AI model developed by researchers at Uppsala University promises to transform MS diagnosis and treatment by providing early and accurate identification of the disease's progression. Its potential impact on clinical trials and personalized medicine is expected to drive significant advancements in the field.