Researchers from the University of Illinois recently suggested an AI-based framework for diagnosing Parkinson’s disease gait dysfunctions using video and deep learning algorithms.
Gait dysfunctions are changes to the typical walking pattern that are commonly connected to disease or physical abnormalities in certain body components. Gait dysfunction is one of the most common causes of falls in older people, accounting for an estimated 17% of all falls.
When a patient may be suffering from a certain neurological condition, such as Parkinson’s disease or multiple sclerosis, doctors usually assess their ability to walk. A person’s solitary walks may provide details regarding a neurological condition that is present.
The study’s conclusions showed that their approach was capable of up to 79% accuracy. In rural or impoverished locations, a wider range of healthcare practitioners may be able to recognise early gait alterations caused by neurological diseases and more quickly deliver a potential diagnosis using the integration of video of individuals walking and AI, according to the research team.
33 volunteers—10 with MS, 9 with Parkinson’s disease, and 14 without neurological disorders—were recruited by the researcher. All of the volunteers walked on a treadmill while being watched by two common RGB cameras.
By examining how these coordinates changed over time, researchers searched for distinctions between those who have MS or Parkinson’s disease and those who do not.
To assess these gait motions, researchers developed and tested 16 alternative AI systems. The algorithms were 79% accurate in determining someone’s neurological condition. If constructed properly, this may, in the researchers’ opinion, transform the game.
