Well-known for its varied and unpredictable symptoms, Multiple Sclerosis is a big challenge for medics, as well as those living with the condition. But artificial intelligence is fast emerging as a promising ally in the fight against MS. We look at recent advances in AI research for MS, and its potential for diagnosis, monitoring and treatment.
AI in healthcare settings
Alongside all the scare stories about AI in the news, we often hear of its potential to transform healthcare for the better. This optimism is not unfounded. Programmers are developing AI machine learning models that can detect patterns in huge complex datasets, such as health records, medical images, genetic information, and medical research papers. These can automate tasks to make light work of admin and help clinicians make faster, better decisions.
AI is already assisting in many clinical settings, from early diagnosis of cancers to assessing patient data in home-based ‘virtual’ wards. AI-assisted analysis of MRI scans has already been credited with revolutionising stroke care in the UK. Since 2020, the UK government has been funding the most promising technologies for health and social care. According to the NHS AI Lab Roadmap, there are now 83 live AI research projects across the NHS.
AI in MS Research
Within the MS research community, there’s also a lot going on with AI. A recent literature review of research papers from the last four years concluded that AI technologies could bring a transformation in personalised MS care. With access to large amounts of patient data, AI models can offer faster, more accurate diagnosis and make better predictions of how the condition will progress. Research is also starting to give us a glimpse of how AI could design effective bespoke treatment plans.
Interpreting MRI scans
MRI scans of the brain and spinal cord are an important tool for diagnosing and monitoring MS. Skilled clinicians inspect sets of MRI images to look for lesions – the small spots of scarring that give multiple sclerosis or ‘many scars’ its name. It’s a time-consuming process, prone to human error.
One of the NHS AI projects uses icobrain ms AI software to assess annual MRI scans find out how well people with are responding to disease modifying drugs. There’s now mounting evidence, including from recent research in Australia, suggesting AI models can now detect, analyse and quantify lesions in MRI scans more accurately than humans in a fraction of the time.
As more is discovered about MS and the brain, the demand for MRI scans is likely to increase significantly, making fast, automated AI analysis even more valuable.
Predicting the effects of MS
AI has helped researchers define new classifications of MS from MRI scans, and make predictions about how people with these forms will progress. Researchers have also been able to predict MS outcomes with AI models looking at other types of clinical data, such as real-time data from wearable sensors. New research from the US suggests AI may even be able to predict MS from blood cells and other biological markers years before any symptoms appear.
Glimpsing the future
AI technologies are helping engineers develop robotics for MS treatment and therapies. New-generation nanorobotics could cross the blood-brain barrier and target anasthetics directly to inflamed lesions in the central nervous system. And robot-assisted gait training could help people with MS improve their mobility.
AI technologies are developing at high speed. Advances that seemed like science fiction a few years ago are now almost within reach. There are big ethical and logistical challenges, but research into MI and AI is an exciting frontier. It gives us hope that some of these emerging technologies could indeed dramatically improve the lives of people with Multiple Sclerosis.