The team uses digital cameras, machine learning to predict neurological diseases

IEEE Journal of Biomedical and Health Informatics (2022). DOI: 10.1109/JBHI.2022.3208077″ width=”800″ height=”530″/>

credit: IEEE Journal of Biomedical and Health Informatics (2022). DOI: 10.1109/JBHI.2022.3208077

In an effort to simplify the process of diagnosing patients with multiple sclerosis and Parkinson’s disease, researchers used digital cameras to capture changes in gait — a symptom of these conditions — and developed a machine-learning algorithm that can distinguish between people with MS and PD from people without these neurological conditions.

Their findings are reported in IEEE Journal of Biomedical and Health Informatics.

The goal of the study was to make the process of diagnosing these diseases more accessible, said Manuel Hernandez, a professor of kinesiology and the University of Illinois at Urbana-Champaign. public health who led the work with graduate student Rahnit Kaur and professor of industrial and enterprise systems engineering and mathematics Richard Sowers.

Currently, patients must wait — sometimes years — to see a neurologist for a diagnosis, Hernandez said. And people in rural communities often have to travel long distances to a facility where their condition can be assessed. To be able to collect gait information using no more than a digital camera and online evaluation of these data may allow clinicians to perform rapid screening that refers only those likely to have neurological conditions to a specialist.

To conduct the study, the team videotaped adults with and without MS or Parkinson’s disease as they walked on a treadmill, focusing digital cameras on the participants’ thighs and lower limbs. Those without neurological conditions were comparable in age, weight, and gender to participants with MS and PD. The walking exercise also included trials in which participants walked while reading each letter of the alphabet in turn. This additional task was designed to simulate real-world walking challenges while performing other potentially mentally distracting tasks, Sowers said.

“This is new research because we were trying to draw attention to the fact that the lab is different from how people behave in the wild,” he said. “When you’re at home, you’re doing whatever you’re doing, but you’re also thinking, ‘Did I close the garage door? Did I turn off the stove?’ So there is an additional cognitive load.”

The researchers used an open-source video analysis tool to obtain data on how the participants moved during the walking exercise.

“We looked at body coordinates for the hips, knees, ankles, big toes and heels,” said Kaur, who developed a method to analyze how these coordinates changed over time to reveal differences between adults with and without MS or the disease. Parkinson’s.

She tested the accuracy of her approach using more than a dozen traditional machine and deep learning algorithms. The team also tested the method on new study subjects to see if it could identify people with MS, Parkinson’s disease, and those without either.

The study found that some of the algorithms were more than 75% accurate in detecting these differences.

“This study demonstrates the viability of low-cost vision-based systems for the diagnosis of certain neurological disorders,” the researchers wrote.

According to scientists, it will probably take several years to make the new tools available to the public.


Machine learning helps detect walking problems in people with multiple sclerosis


Additional information:
Rahneet Kaur et al., Vision Frameworks for Predicting Multiple Sclerosis and Parkinson’s Gait Dysfunction—A Deep Learning Approach, IEEE Journal of Biomedical and Health Informatics (2022). DOI: 10.1109/JBHI.2022.3208077

Citation: Team Uses Digital Cameras, Machine Learning to Predict Neurological Diseases (2022, October 11) Retrieved October 11, 2022, from https://medicalxpress.com/news/2022-10-team-digital-cameras-machine-neurological .html

This document is subject to copyright. Except in good faith for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.