Every year, NHS radiologists spend thousands of hours scrutinising MRI scans from over 130,000 people with multiple sclerosis (MS), trying to spot tiny new brain lesions by eye—a process that is slow, tiring, and prone to error. This project tests whether an artificial intelligence tool called icobrain-ms can do that job better. The problem is that MS is a disabling central-nervous-system disease costing the NHS more than £1 billion annually. Patients on disease-modifying treatments need regular MRI monitoring to detect hidden disease activity, which signals that a drug change is needed. But visual detection alone misses some changes and takes too long. If icobrain-ms proves superior to eyeball assessment in a cluster randomised trial, it could transform a quiet but critical part of NHS infrastructure: the radiology workflow. Faster, more accurate scan review would mean quicker treatment decisions, fewer relapses, and lower healthcare costs. The team will also model the health economics—measuring impacts on radiologist time, drug costs, and follow-up care. Success could lead to a phase 4 study and a NICE health technology assessment, with the goal of rolling out AI-assisted MRI reading across the NHS.
View original technical description
BACKGROUND Multiple sclerosis (MS) is a disabling disease of the central-nervous-system. Over 130,000 people with MS (pwMS) in the UK; annual cost to the NHS >£1billion. Disease modifying treatment (DMT) standard of care. However, effective monitoring essential to detect sub-clinical disease activity. Magnetic resonance imaging (MRI) established tool. However, eyeball detection time consuming, tiring and error-prone. icobrain-ms is a validated AI technology complementing visual assessment, thereby supporting clinicians to decide whether or not DMT needs to change. AIMS To test icobrain-ms for superiority in detecting disease-activity; to undertake health-economic analysis to study impact on resources and disease course. PRIMARY OBJECTIVE To establish whether there is superiority of icobrain-ms over visual detection of MRI disease activity in pwMS on DMT. SECONDARY OBJECTIVES CLINICAL To establish whether pwMS with disease-activity detected by eyeballing only, whose DMT was subsequently changed, encounter less disease-activity or clinical deterioration than pwMS in whom only icobrain-ms revealed changes, but no change in DMT was triggered. HEALTH ECONOMICS To establish whether icobrain-ms impacts on healthcare resources, including time for MRI review and from MRI acquisition to review; DMT change and cost; follow-up care requirements. WORK PLAN Six work packages (WP1) set-up, protocol development, submissions, (WP2) audit of current MRI use, (WP3) installation of icobrain-ms, (WP4) clinical trial comparing icobrain-ms vs. eyeball assessment (WP5) health-economic study, (WP6) communication of results, implementation in practice. METHODS Cluster randomised trial 1:1 of icobrain-ms vs eyeball assessment. Health-economic modelling. TIMELINES Project duration 36 months: 6 months set-up; 12 months recruitment; 12 months follow-up; 6 months closeout, analysis, reporting. IMPACT Personalised DMT; improved outcomes for pwMS; reduced healthcare costs DISSEMINATION Peer-reviewed publications; PPI STRATEGY AFTER FUNDING Plan phase 4 study; submission for NICE HTA with a view to roll out technology across the NHS.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know