Real-World Evidence in Multiple Sclerosis

Large-scale real-world evidence to advance MS personalized medicine.

Multiple sclerosis (MS) is an autoimmune disease in which the myelin sheath surrounding nerve fibers is damaged. Its heterogeneous presentation makes clinical management challenging at the individual patient level. Quantitative MRI biomarkers—such as lesion volume and measures of brain atrophy—can help clinicians objectively monitor disease progression and guide treatment decisions.

This project involves building the infrastructure to ingest standardized MRI data and neuroperformance test results from patients with MS across 10 leading clinical centers in the United States, Spain, and Germany. The project established a real-world dataset of 20,000 unique patients that has contributed to deepen our understanding of the disease and demonstrated the potential to transform patient care.

From clinical PACS systems to research-ready dataset.

The pipeline ingests standardized, de-identified MRI data from clinical PACS systems into a centralized data repository, performs image quality checks, and processes the scans through an analysis pipeline to derive imaging biomarkers, including lesion burden and measures of brain atrophy.

A relational database (MySQL) is used to manage MRI studies and derived metrics, linking imaging biomarkers with neuroperformance test results, clinical outcomes, and fluid biomarkers. This multimodal real-world dataset is shared quarterly with investigators to support collaborative research.

• Standardization MRI acquisition protocols for T1-weighted and 3D FLAIR sequences are standardized on Siemens 3T scanners. This enables scanner-specific calibration of derived metrics and supports algorithm development to accurately compute biomarkers across different scanner models and clinical sites
• Data Scale The platform supports a real-world clinical dataset of more than 20,000 patients with MS, encompassing heterogeneous imaging protocols across multiple clinical sites and scanner models
• HPC Pipeline High-performance computing infrastructure is used to identify appropriate MRI study time points for analysis, compute longitudinal metrics, and store results in the database
• Research Output The shared real-world dataset enables research in personalized medicine and treatment response