Researchers at Queen Mary University of London have developed a machine learning model that can forecast the onset of dementia up to nine years before clinical diagnosis by analysing fMRI scans. The model’s high accuracy and focus on the brain’s default mode network offer promising potential for early detection and interventions.
Researchers at Queen Mary University of London have developed a machine learning model capable of predicting the onset of dementia up to nine years before clinical diagnosis. The model, which boasts over 80% accuracy, analyzes functional magnetic resonance imaging (fMRI) scans to identify early signs of all-cause dementia and Alzheimer’s disease.
The study utilized 1,111 fMRI scans from the UK Biobank, involving participants who had not been diagnosed with dementia at the time of their scans but developed the condition within nine years. The model specifically examines changes in the brain’s default mode network (DMN), which is active during rest and self-referential thought. Through machine learning, researchers identified disconnects between ten key DMN regions.
The study revealed associations between DMN disconnectivity and dementia risk factors, such as social isolation and genetic predispositions. The senior author, Professor Charles Marshall, emphasized the model’s potential in early detection, potentially allowing for timely interventions once effective treatments are developed.
The findings, published in Nature Mental Health, highlight the need for further research to validate these results across diverse populations and different types of dementia. Dr. Claire Sexton from the Alzheimer’s Association and Dr. Clifford Segil of Providence Saint John’s Health Center also noted the importance of continued advancements and standardizations in diagnostic tools used for dementia prediction.
