A video processing technique developed at the University of Florida that uses artificial intelligence will help neurologists better track the progression of Parkinson's disease in patients, and ultimately improve their care and quality of life. will increase.
The system, developed by Diego Garin, Ph.D., assistant professor of applied physiology and kinesiology in the UF College of Health and Human Performance, uses machine learning to analyze video recordings of patients performing the finger tapping test, a standardized test. applies to For Parkinson's disease involves tapping the thumb and forefinger 10 times quickly.
“By studying these videos, we can also detect small changes in hand movements that are characteristic of Parkinson's disease but can be difficult for clinicians to visually identify.” . “The beauty of this technology is that a patient can record themselves performing the test, and the software analyzes it and informs the physician how the patient is moving so the physician can make a decision.”
Parkinson's disease is a brain disorder that affects movement and can result in tremors, tremors, stiffness, and problems with balance and coordination. Symptoms usually start slowly and get worse over time. There is no specific lab or imaging test that can diagnose Parkinson's disease, but a series of exercises and exercises performed by the patient help clinicians identify and assess the severity of the disorder.
The most commonly used movement disorder classification scale used in Parkinson's disease is the Unified Parkinson's Disease Classification Scale. Garin explained that, despite its reliability, the rating is limited to a 5-point scale, which limits its ability to track subtle changes in development and is prone to interpretation.
The research team, which included UF neurologist Joshua Wong, MD; Nicholas McFarland, MD, PhD; And Adolfo Ramirez-Zamora, MD, has developed a more objective method for quantifying motor symptoms in Parkinson's patients by analyzing videos and capturing proportional changes in the disease over time. Machine learning algorithm was used for this.
“We found that we could see the same features that clinicians are trying to see using cameras and computers,” Garin said. “With the help of AI, the same exam is easier and less time consuming for everyone involved.”
Garin said the automated system also revealed previously overlooked details about movement using precise data collected by the camera, such as how quickly the patient opens a finger during movement. on or off and how much the movement characteristics change during each tap.
“We've seen that, with Parkinson's disease, the opening movement is delayed, compared to the same movement in individuals who are healthy,” Garin said. “This is new information that would be nearly impossible to measure without video and computers, letting us know that this technology can help better characterize how Parkinson's disease affects movement and guide treatment options. provides new markers to help assess effectiveness.”
To complete the system, which Garin originally designed to analyze facial features for conditions other than Parkinson's disease, the team tapped into UF's HyperGator — the world's largest One of the big AI supercomputers.
“HyperGator enabled us to develop a machine learning model that simplifies the video data into motion scores,” explained Garin. “We used Hypergrator to train, test, and improve various models with large amounts of video data, and now those models can run on a smartphone.”
Michael S. Okon, MD, director of the Norman Fixel Institute and medical advisor to the Parkinson Foundation, said automated video-based diagnostics could be a “game changer” for clinical trials and care.
“The finger tapping test is one of the most important elements used to diagnose and measure disease progression in Parkinson's disease,” Okon said. “Today, an expert is needed to interpret the results, but what is revolutionary is how Diego and three Parkinson's neurologists at the Fixel Institute can use AI to object to the progression of the disease.”
In addition to putting this technology in the hands of neurologists and other care providers, Garin is working with UFIT to develop it into an app for mobile devices, allowing individuals to spend time at home. You will be able to assess your illness as well.