In an effort to gain a better understanding of the development of multiple sclerosis (MS) and diagnose it before the first symptoms appear, scientists are utilizing statistical and artificial intelligence (AI) methods to build models of the evolution of brain structures. This groundbreaking research aims to shed light on the progression of the disease and potentially lead to improved treatment options.
MS is a chronic condition that primarily affects young adults between the ages of 25 and 35, with over 115,000 people in France alone believed to be affected. The disease is characterized by inflammation, autoimmunity, and neurodegeneration, causing damage to the brain and spinal cord. The destruction of myelin, the protective sheath around neuronal axons, disrupts the transmission of information between nerve cells, leading to motor, visual, sensory, and cognitive impairments. While treatments exist to slow its progression, a cure for MS remains elusive.
One of the challenges in studying and monitoring MS is that patients typically only undergo magnetic resonance imaging (MRI) scans once symptoms have already manifested. This limits researchers’ ability to understand the early stages of the disease and how it develops over time. Neurodegenerative diseases, including MS, often affect various brain structures before symptoms become apparent, making early detection and intervention difficult.
Pierrick Coupé, a research professor at CNRS, has dedicated his efforts to analyzing and processing biomedical images related to neurodegenerative diseases. His focus on developing lifespan models, which chronologically situate phenomena, has yielded promising results. By analyzing tens of thousands of brain images from individuals of all ages, Coupé and his team have been able to estimate the average trajectories of brain structure volumes throughout a person’s lifetime, similar to growth curves for monitoring height and weight in children.
Initially intended for investigating normal brain aging, Coupé and his colleagues redirected their research toward conditions like MS and Alzheimer’s disease. By constructing models that compare the normal and pathological trajectories of various brain structures, they have made significant discoveries about the early impact of MS on different regions of the brain.
Through the analysis of MRI scans from nearly 41,000 subjects, including 2,500 MS patients, the researchers focused on understanding when and how MS affects specific brain structures. The study involved an intensive process of brain segmentation using advanced AI techniques. Coupé and his team employed collective AI, where multiple neural networks worked together to achieve precise and reliable results. This approach allowed for the processing of a vast amount of data in a controlled and efficient manner.
The results of the study revealed that the thalamus, responsible for processing sensory data, was the first structure to be affected by MS, followed by the putamen, involved in movement regulation and learning. The brain stem, responsible for vital functions such as heart rate and respiration, showed signs of damage approximately nine years after the thalamus.
These findings are crucial in understanding the early impact of MS on the brain and could potentially pave the way for improved diagnostic tools and personalized treatment strategies. Coupé and his team are actively working on developing a diagnostic support tool for MS, similar to the one they created for Alzheimer’s disease, and hope to expand their research to differentiate between various types of dementia.
The integration of artificial intelligence in the study and diagnosis of neurodegenerative diseases opens up new possibilities for improved patient care. However, Coupé emphasizes the importance of AI systems being explainable and justifiable. Transparency in the decision-making process of AI algorithms is essential to ensure accurate diagnoses and reduce the chances of errors.
The research conducted by Coupé and his team showcases the power of artificial intelligence in transforming our understanding of complex diseases like MS. By harnessing the potential of AI and advanced imaging techniques, scientists are making significant strides toward early detection, personalized treatment, and ultimately improving the quality of life for patients affected by neurodegenerative diseases.
