According to the World Health Organization, approximately 55 million people worldwide live with dementia. The most common form of dementia is Alzheimer’s disease, an incurable condition that leads to a decline in brain function.
In addition to the physical implications, Alzheimer’s disease has psychological, social, and economic repercussions, not only for the individuals living with the disease but also for those who love and care for them. As the symptoms worsen over time, it is important for both patients and their caregivers to prepare for potential needs and increase support as the disease progresses.
To address this, researchers at the University of Texas in Arlington have developed a novel learning-based framework to help Alzheimer’s patients accurately determine where they are on the spectrum of disease progression. This can enable better prediction of later stages and make future treatment planning for advanced disease easier.
„Various predictive approaches have been proposed and evaluated for their predictive capability of Alzheimer’s disease and its precursor, mild cognitive impairment.“
Dajiang Zhu, Associate Professor of Computer Science and Engineering, UTA
He is the lead author of a new peer-reviewed article published in Open Access Pharmacological Research. „Many of these previous prediction instruments overlooked the continuous development of Alzheimer’s disease and the transitional stages of the disease,“ says Zhu.
Supported by grants from the National Institutes of Health and the National Institute on Aging totaling over $2 million, Zhu’s Medical Imaging and Neuroscientific Discovery Research Laboratory and Li Wang, Associate Professor of Mathematics at UTA, developed a new learning-based embedding framework that encodes the different stages of Alzheimer’s disease progression in a process they call a „disease embedding tree“ or DETree. Using this framework, DETree can efficiently and accurately predict each of the five granular clinical groups of Alzheimer’s disease progression and provide more detailed status information by projecting where the patient will be in the course of the disease.
To test their DETree framework, the researchers used data from 266 individuals with Alzheimer’s disease from the multi-center Alzheimer’s Disease Neuroimaging Initiative. The results of the DETree strategy were compared with other widely used methods for predicting the progression of Alzheimer’s disease, and the experiment was validated using machine learning methods.
„We know that people with Alzheimer’s disease often progress with symptoms deteriorating at very different rates,“ says Zhu. „We are excited that our new framework is more accurate than the other available prediction models, and we hope it will help patients and their families better plan for the uncertainties of this complex and devastating disease.“
He and his team believe that the DETree framework has the potential to help predict the progression of other diseases that have multiple clinical developmental stages, such as Parkinson’s disease, Huntington’s disease, and Creutzfeldt-Jakob disease.
University of Texas in Arlington
Zhang, L., et al. (2024). Disease2Vec: Encoding Alzheimer’s disease progression through the disease embedding tree. Pharmacological Research. doi.org/10.1016/j.phrs.2023.107038.