Researchers at the University of Massachusetts Amherst have been awarded a two-year grant of $278,118 from the National Institutes of Health (NIH) to develop innovative deep-learning models for the early detection of Alzheimer’s disease using clinical data, including real-world brain MRIs.
The main objective of this research is to identify Alzheimer’s patients at an early stage, ideally two or more years before symptoms manifest, and to identify populations at risk of developing the disease using MRI data. This will allow researchers to test interventions and medications that can potentially disrupt the progression of the disease. To achieve this, the researchers will rely on multimodal clinical data, including brain MRIs.
Dr. Madalina (Ina) Fiterau, the principal investigator and project leader of the study, highlights the importance of identifying brain changes early on, stating, “Sixty percent of a patient’s brain matter disappears by the time of diagnosis, and at that stage, it’s irretrievable. What we would like to do is identify those changes early, at least two years before onset, and then, based on that, figure out which treatments work.”
Unlike previous research that used specialized data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), this study aims to develop models that utilize standard MRIs instead of engineered data. Dr. Joyita Dutta, another principal investigator in the study, explains that while specialized data is useful, it lacks generalizability to real-world scenarios. The goal is to train the deep-learning models to extract relevant features from standard brain MRIs, specifically focusing on regions known to be affected by Alzheimer’s such as the hippocampus, cerebral cortex, and ventricle cavities.
Additionally, the research team aims to overcome model biases that arise from the demographic gaps present in ADNI data. The ADNI data predominantly consists of white, highly educated individuals, which does not accurately represent the diverse population affected by Alzheimer’s. By addressing these biases, the models can be applied more effectively to diverse patient populations.
This grant-funded research is a significant step toward earlier Alzheimer’s detection and intervention, providing hope for improved patient outcomes and paving the way for potential disease-modifying therapies.
The source of the article is from the blog yanoticias.es