• Latest news

NMH: Neural network developed that predicts onset of dementia with 80% accuracy 

July 10, 2024  13:15

British scientists from Queen Mary University of London have created a new method of predicting dementia nine years before diagnosis. According to experts, the method surpasses the usual memory tests and measurements of the degree of brain shrinkage. The study is published in the scientific journal Nature Mental Health (NMH).

Dementia is a collective term used to describe a variety of conditions characterized by gradual cognitive decline severe enough to interfere with daily life and independent functioning. It affects memory, thinking, orientation, comprehension, numeracy, learning ability, language and judgment.

The experts used data from UK Biobank, a large UK biomedical database. The researchers focused on a subset of participants who had undergone functional magnetic resonance imaging (fMRI) scans and either already had a diagnosis of dementia or received it later.

The sample consisted of 148 dementia cases and 1030 matched controls, providing a robust comparison group by matching for age, gender, ethnicity, lead arm, and geographic location of the fMRI scanning center.
Participants underwent resting-state fMRI scans (rs-fMRI), which measures brain activity by detecting changes in blood flow.

The researchers targeted the passive brain mode network (DMN), a brain structure active during rest and involved in high-level cognitive functions such as social cognition and self-referential thinking.

Using a technique called dynamic causal modeling (DCM), the scientists analyzed rs-fMRI data to estimate the effective connectivity between different regions in the DMN.

The experts then used these connectivity estimates to train a machine learning model. The goal of this model was to distinguish between people who would later develop dementia and those who would not.

Neuroseed identified 15 key connectivity parameters in DMN that differed significantly between future dementia cases and control groups. Among these, the most notable changes included increased inhibition from the ventromedial prefrontal cortex (vmPFC) to the left parahippocampal formation (lPHF) and from the left intraparietal cortex (lIPC) to the lPHF, and impaired inhibition from the right parahippocampal formation (rPHF) to the dorsomedial prefrontal cortex (dmPFC).

In addition to diagnostic capabilities, the researchers developed a model to predict time to dementia diagnosis. The predictive power of these models is 80% and suggests that changes in the DMN may serve as early biomarkers of dementia, opening a window into the disease process years before clinical symptoms appear.

Follow NEWS.am Medicine on Facebook and Twitter

  • Video
  • Event calendar
  • Archive