AI predicts how multiple sclerosis patients will progress in ten years with 90% certainty

Using data from the first MRI scans of patients with multiple sclerosis, obtained from the neurology service of the Santiago de Compostela Hospital, the program uses artificial intelligence to predict the development of these patients very accurately over 10 years, with almost 90 percent, according to a new study by the Translational Research Group in Neurological Diseases (ITEN) of IDIS and IIS Galicia.

The study, published in the journal Plos One, analysed a total of 446 records of affected people over a follow-up period of at least one year.

This machine learning model predicts disability progression in patients with multiple sclerosis using baseline magnetic resonance imaging (MRI) data and clinical assessments using the Expanded Disability Status Scale.

“The work offers new models to describe the progression of patients using artificial intelligence programs that predict their trajectories using these descriptors, and also gives us an idea of ​​what factors contribute to this evolution, such as age of onset or injury,” explains researcher and first author of the work, Silvia Campagnioni.

The study will therefore allow for the optimization of the dosage of MS treatment in terms of dose and duration of treatment, as well as the optimization of its use depending on each patient’s profile or treatment methods, while improving the trajectory through the use of individual drugs. ML Predictors

Among the most notable findings, the study found that “age of onset” was one of the most influential characteristics of the developed regressor models. Moreover, the number of brain lesions greater than or equal to nine on the initial MRI was the most influential variable in the classifier model decisions.

“This work has a significant impact not only in scientific and technical terms, but also in economic and social terms, as it affects health, quality of life and development cooperation,” says IDIS researcher Roberto Agis, the final author of the project. “We could obtain objective data and indicators of preventive measures that would help predict the therapeutic effectiveness of treatment,” he explains.

Multiple sclerosis (MS) is a chronic inflammatory neurodegenerative disease that causes demyelination and accumulation of long-term disability. In MS, autoreactive activation of the adaptive immune system plays a leading role. Although the cause of the disease is unknown, everything points to an interaction between genes and environmental risk factors, viruses, and lifestyle.

According to the findings of the research team, research using artificial intelligence (AI) can provide versatile and powerful tools for the treatment of multiple sclerosis. “Artificial intelligence technologies such as deep learning and machine learning can facilitate the integration of biological, psychological and social factors in the prevention, diagnosis and treatment of multiple sclerosis, including other diseases,” explains Cesar Veiga.

Treatment decisions in multiple sclerosis are still based on the integration of the same demographic, clinical and paraclinical patient variables that were observed many years ago, such as resonance images and the presence of oligoclonal bands. “There are still many open problems in this field, and improvements are coming from several areas of convergence, such as the integration of data sets that can improve personalization and the predictive power of AI algorithms in healthcare,” says Jose Maria Prieto, head of the IDIS ITEN Group.