Summary: Researchers developed an innovative learning-based framework called DETree to accurately predict the progression of Alzheimer’s disease. This new tool addresses the continuous nature of Alzheimer’s development.
By efficiently and accurately predicting the disease’s various stages, DETree enables patients and caregivers to better plan for future care needs. This framework, tested using data from the Alzheimer’s Disease Neuroimaging Initiative, surpasses the accuracy of existing prediction models and could potentially be applied to other neurodegenerative diseases.
Key Facts:
- The DETree framework can predict five clinical groups of Alzheimer’s disease development with high accuracy.
- This tool provides valuable insights into the disease’s progression, aiding patients and caregivers in planning future care.
- The research shows promise for applying DETree to other diseases with multiple developmental stages, like Parkinson’s and Huntington’s.
Source: UT Arlington
About 55 million people worldwide are living with dementia, according to the World Health Organization. The most common form is Alzheimer’s disease, an incurable condition that causes brain function to deteriorate.
In addition to its physical effects, Alzheimer’s causes psychological, social and economic ramifications not only for the people living with the disease, but also for those who love and care for them. Because its symptoms worsen over time, it is important for both patients and their caregivers to prepare for the eventual need to increase the amount of support as the disease progresses.
To that end, researchers at The University of Texas at Arlington have created a novel learning-based framework that will help Alzheimer’s patients accurately pinpoint where they are within the disease-development spectrum. This will allow them to best predict the timing of the later stages, making it easier to plan for future care as the disease advances.
“For decades, a variety of predictive approaches have been proposed and evaluated in terms of the predictive capability for Alzheimer’s disease and its precursor, mild cognitive impairment,” said Dajiang Zhu, an associate professor in computer science and engineering at UTA. He is lead author on a new peer-reviewed paper published open access in Pharmacological Research.
“Many of these earlier prediction tools overlooked the continuous nature of how Alzheimer’s disease develops and the transition stages of the disease.”
In work supported by more than $2 million in grants from the National Institutes of Health and the National Institute on Aging, Zhu’s Medical Imaging and Neuroscientific Discovery research lab and Li Wang, UTA associate professor in mathematics, developed a new learning-based embedding framework that codes the various stages of Alzheimer’s disease development in a process they call a “disease-embedding tree,” or DETree.
Using this framework, the DETree can not only predict any of the five fine-grained clinical groups of Alzheimer’s disease development efficiently and accurately but can also provide more in-depth status information by projecting where within it the patient will be as the disease progresses.
To test their DETree framework, the researchers used data from 266 individuals with Alzheimer’s disease from the multicenter Alzheimer’s Disease Neuroimaging Initiative. The DETree strategy results were compared with other widely used methods for predicting Alzheimer’s disease progression, and the experiment was repeated several times using machine learning-methods to validate the technique.
“We know individuals living with Alzheimer’s disease often develop worsening symptoms at very different rates,” Zhu said. “We’re heartened that our new framework is more accurate than the other prediction models available, which we hope will help patients and their families better plan for the uncertainties of this complicated 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 stages of development, such as Parkinson’s disease, Huntington’s disease, and Creutzfeldt-Jakob disease.
About this Alzheimer’s disease research news
Author: Katherine Bennett
Source: UT Arlington
Contact: Katherine Bennett – UT Arlington
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Disease2Vec: Encoding Alzheimer’s progression via disease embedding tree” by Dajiang Zhu et al, Pharmacological Research
Abstract
Disease2Vec: Encoding Alzheimer’s progression via disease embedding tree
For decades, a variety of predictive approaches have been proposed and evaluated in terms of their prediction capability for Alzheimer’s Disease (AD) and its precursor – mild cognitive impairment (MCI). Most of them focused on prediction or identification of statistical differences among different clinical groups or phases, especially in the context of binary or multi-class classification.
The continuous nature of AD development and transition states between successive AD related stages have been typically overlooked. Though a few progression models of AD have been studied recently, they were mainly designed to determine and compare the order of specific biomarkers.
How to effectively predict the individual patient’s status within a wide spectrum of continuous AD progression has been largely understudied. In this work, we developed a novel learning-based embedding framework to encode the intrinsic relations among AD related clinical stages by a set of meaningful embedding vectors in the latent space (Disease2Vec).
We named this process as disease embedding. By Disease2Vec, our framework generates a disease embedding tree (DETree) which effectively represents different clinical stages as a tree trajectory reflecting AD progression and thus can be used to predict clinical status by projecting individuals onto this continuous trajectory.
Through this model, DETree can not only perform efficient and accurate prediction for patients at any stages of AD development (across five fine-grained clinical groups instead of typical two groups), but also provide richer status information by examining the projecting locations within a wide and continuous AD progression process. (Code will be available: https://github.com/qidianzl/Disease2Vec.)
