Dynamic Indicators of Mental Illness Captured by Advanced AI in Brain Imaging

Summary: A new AI algorithm has used fMRI data to identify new brain patterns linked to mental health disorders.

Source: Georgia State University

New research from Georgia State University’s TReNDS Center could lead to early diagnosis of devastating illnesses such as Alzheimer’s disease, schizophrenia and autism, in time to help prevent and more easily treat these disorders .

In a new study published in Scientific reports A team of seven Georgia State scientists have built a sophisticated computer program that can sift through massive amounts of brain imaging data and uncover new patterns related to mental health issues.

The brain imaging data comes from scans using functional magnetic resonance imaging (fMRI), which measures dynamic brain activity by detecting tiny changes in blood flow.

“We built artificial intelligence models to interpret the large amounts of information coming from fMRI,” said Sergey Plis, associate professor of computer science and neuroscience at Georgia State and lead author of the study.

He compared this type of dynamic imaging to film – as opposed to a snapshot such as an X-ray or, the more common structural MRI – and noted that “the data available is so much bigger, so much richer than a regular blood test or MRI. But that’s the challenge: this huge amount of data is difficult to interpret.

Moreover, fMRIs on these specific conditions are expensive and difficult to obtain. Using an artificial intelligence model, however, regular fMRI data can be extracted. And these are available in large numbers.

“There are large datasets available on people with no known clinical disorder,” says Vince Calhoun, founding director of the TReNDS Center and one of the study’s authors. Using these large but unrelated available datasets improved model performance on smaller specific datasets.

“New patterns emerged that we could definitely link to each of the three brain disorders,” Calhoun said.

The AI ​​models were first trained on a dataset of over 10,000 people to learn to understand basic fMRI imaging and brain function. The researchers then used multi-site datasets of more than 1,200 people, including those with autism spectrum disorders, schizophrenia and Alzheimer’s disease.

How it works? It’s a bit like Facebook, YouTube or Amazon learning about you from your online behavior and starting to be able to predict your future behavior, likes and dislikes. The computer software was even able to pinpoint the “time” when the brain imaging data was most likely related to the mental disorder in question.

To make these results clinically useful, they will need to be applied before a disorder occurs.

“If we can find markers and predict Alzheimer’s risk in a 40-year-old man,” Calhoun said, “we might be able to do something about it.”

Likewise, if the risks of schizophrenia can be predicted before there are any real changes in brain structure, there may be ways to offer better or more effective treatments.

“Even though we know from other tests or family history that a person is at risk for a disorder such as Alzheimer’s disease, we are still unable to predict exactly when that will happen,” Calhoun said.

Likewise, if the risks of schizophrenia can be predicted before there are any real changes in brain structure, there may be ways to offer better or more effective treatments. Image is in public domain

“Brain imaging could narrow this time window, capturing relevant patterns when they appear before clinical disease is apparent.”

“The vision is that we collect a large set of imaging data, our AI models dig into that and show us what they’ve learned about certain disorders,” Plis said. “We build systems to discover new knowledge that we could not discover on our own.”

“Our goal,” said Md Mahfuzur Rahman, the study’s first author and a PhD student in computer science at Georgia State, “is to connect big worlds and big datasets to small worlds and specific datasets. to diseases and progress to clinically relevant markers for decision-making.”

Funding: This study was supported by seed funding from SMP and in part by NIH grants R01EB006841, R01MH118695, RF1MH121885, and NSF 2112455.

See also

This shows glut3 activation on t cells

About this AI and mental health research news

Author: Noelle Reetz
Source: Georgia State University
Contact: Noelle Reetz – Georgia State University
Image: Image is in public domain

Original research: Free access.
Interpret models interpreting brain dynamics” by Sergey Plis et al. Scientific reports.


Summary

Interpret models interpreting brain dynamics

Brain dynamics is very complex and yet holds the key to understanding brain function and dysfunction.

The dynamics captured by resting-state functional magnetic resonance imaging data are noisy, high-dimensional, and difficult to interpret. The typical approach of reducing this data to low-dimensional features and focusing on the most predictive features comes with strong assumptions and may miss key aspects of the underlying dynamics.

In contrast, introspection of discriminatively trained deep learning models can uncover disorder-relevant cue elements at individual time points and spatial locations. Yet, the difficulty of reliably training on high-dimensional, low-sample-size datasets and the unclear relevance of the resulting predictive markers prevent the widespread use of deep learning in functional neuroimaging.

In this work, we introduce a deep learning framework to learn from high-dimensional dynamic data while maintaining stable and ecologically valid interpretations.

The results successfully demonstrate that the proposed framework can learn resting-state fMRI dynamics directly from small data and capture compact and stable interpretations of features predictive of function and dysfunction.

Comments are closed.