Approximately 55 million people worldwide are grappling with dementia, as reported by the World Health Organization, with Alzheimer's disease being the most prevalent form—a degenerative condition without a cure that leads to a decline in brain function.
Beyond the physical toll, Alzheimer's inflicts psychological, social, and economic consequences on both those living with the disease and their caregivers. Given the progressive nature of its symptoms, proactive planning for increased support becomes crucial as the disease advances.
Addressing this need, researchers at The University of Texas at Arlington have developed an innovative learning-based framework. This framework assists Alzheimer's patients in accurately identifying their position on the disease-development spectrum, facilitating better anticipation of the timing of later stages and aiding in planning for future care.
DAJIANG ZHU, AN ASSOCIATE PROFESSOR IN COMPUTER SCIENCE AND ENGINEERING AT UTA
Zhu, an associate professor in computer science and engineering at UTA, highlighted the continuous nature of Alzheimer's development and the transitional stages often overlooked by previous predictive approaches. Supported by over $2 million in grants from the National Institutes of Health and the National Institute on Aging, Zhu's team devised a learning-based embedding framework known as a "disease-embedding tree" (DETree). The DETree not only efficiently and accurately predicts the five fine-grained clinical groups of Alzheimer's development but also offers detailed status information, projecting the patient's position within it as the disease progresses.
Testing their DETree framework involved analyzing data from 266 individuals with Alzheimer's disease from the Alzheimer's Disease Neuroimaging Initiative. The DETree strategy outperformed other widely used methods for predicting Alzheimer's progression in multiple experiments, validating its efficacy using machine learning methods.
Zhu expressed optimism about the framework's accuracy, particularly in accommodating the varying rates at which individuals with Alzheimer's experience worsening symptoms. The team believes that the DETree framework holds promise beyond Alzheimer's and could potentially aid in predicting the progression of other diseases with multiple clinical stages, such as Parkinson's disease, Huntington's disease, and Creutzfeldt-Jakob disease.
JOURNAL: Pharmacological Research