Researchers at University College London (UCL) have developed an innovative artificial intelligence (AI) tool called MindGlide, designed to transform the way multiple sclerosis (MS) treatments are assessed
The new tool uses AI to analyse brain images, like MRI scans and can detect small changes caused by MS that have previously been difficult to measure. With this technology, doctors can better understand the progression of MS and the effectiveness of various treatments.
Understanding MS better
Multiple sclerosis is a condition where the immune system attacks the brain and spinal cord, leading to a range of physical, mental, and emotional challenges.
In the UK, approximately 130,000 people live with MS, which costs the National Health Service (NHS) over £2.9 billion annually. Monitoring the disease and evaluating treatments often rely on MRI scans to observe changes in the brain.
However, specialised scans are usually needed to capture the detailed information required for assessing MS progression, and such scans are not routinely performed.
The future of MRI analysis with MindGlide
MindGlide addresses this limitation by using AI to analyse standard MRI scans that are commonly used in hospitals.
This technology can quickly get important information, such as identifying areas of brain damage, plaques, and even early signs of brain shrinkage, all of which are key indicators of MS.
The AI model is trained on a lot of data, allowing it to detect changes in the brain that traditional methods may miss. Analysing an image takes just five to 10 seconds, a dramatic improvement over the time-consuming manual analysis that often takes weeks, especially with the high volume of scans handled by the NHS.
In a recent study published in Nature Communications, researchers tested MindGlide using over 14,000 images from more than 1,000 MS patients. The results showed that MindGlide outperformed two other AI tools for similar tasks. It was found to be 60% better than one tool, SAMSEG, and 20% better than another, WMH-SynthSeg when it came to detecting brain plaques, a key indicator of disease activity in MS patients. These plaques, or lesions, are often signs of MS progression and are important in assessing treatment effectiveness.
Unlocking valuable insights from existing brain scans
MindGlide can identify lesions on the outer layers of the brain and detects changes in deeper brain areas, creating a more comprehensive view of the disease.
The tool performed well over short-term and long-term scans, making it useful for initial assessments and ongoing monitoring. The technology also confirmed previous high-quality research regarding the effectiveness of various MS treatments, showing how well different therapies worked in slowing down or halting disease progression.
The main breakthrough of MindGlide if the potential to unlock important insights from a large number of existing brain scans. Many of these images were previously underutilised because they were not high-quality enough for expert analysis.
With MindGlide, these images can now be analysed effectively, allowing doctors and researchers to understand better MS and how different treatments impact the disease. This could lead to faster and more accurate diagnoses and better treatment plans personalised to individual patients.
While MindGlide shows a key advancement in MS research and treatment monitoring, its current implementation is limited to brain scans. Spinal cord imaging, which is also critical for assessing MS, is not yet included in the tool’s capabilities. However, researchers plan to expand its scope in future studies to offer a more comprehensive assessment of the entire neural system, including the spinal cord.
