A cutting-edge video-processing technique, pioneered at the University of Florida, is poised to revolutionize the diagnosis and monitoring of Parkinson’s disease by leveraging artificial intelligence capabilities. This innovation will enable neurologists to more accurately track the progression of the condition in patients, ultimately leading to improved care and a better quality of life.
Diego Guarin, Ph.D., an assistant professor in the University of Florida’s Faculty of Health and Human Performance, has developed a system that leverages machine learning to analyze video recordings of patients performing the finger-tapping test, a standard diagnostic tool for Parkinson’s disease that involves rapidly tapping the thumb and index finger 10 times.
“According to Guarin, linked to the Norman Fixel Institute for Neurological Diseases at UF Health, learning from these films could enable the detection of even minute hand action changes that may be indicative of Parkinson’s disease; a visual diagnosis clinicians are likely to find challenging.” “The innovative aspect of this technology lies in its ability to empower patients to record themselves conducting the test, with the software subsequently analyzing their movement patterns and providing clinicians with actionable insights, enabling them to make informed decisions.”
Parkinson’s disease is a neurological disorder characterized by motor symptoms including bradykinesia, tremors, rigidity, and postural instability, which collectively compromise an individual’s overall mobility and quality of life. Typically, signs of a condition begin gradually and progressively deteriorate with time. While there is no single diagnostic test for Parkinson’s disease, a comprehensive clinical evaluation incorporating a series of assessments and examinations allows clinicians to establish a diagnosis and determine the severity of the disorder.
The score scale most widely utilized to track the progression of Parkinson’s disease is the Movement Disorders Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS). While Guarin established a 5-point scale for scores, this restriction hampers the ability to capture subtle changes in progress and leaves it vulnerable to personal biases.
Researchers from UF’s neurology department, including Joshua Wong, M.D., Nicolaus McFarland, M.D., Ph.D., and Adolfo Ramirez-Zamora, M.D., developed a novel approach that leveraged machine learning algorithms to analyze video recordings of Parkinson’s patients, enabling the quantification of subtle motor signs and monitoring disease progression over time.
“We found that by employing a digital camera and computer, we can replicate the exact views clinical professionals aim to achieve,” said Guarin. “With AI assistance, the same examination process becomes faster and more streamlined for all parties involved.”
Gurain noted that the automated system had also uncovered previously unobserved details about motion, leveraging precise data gathered by the digital camera to reveal subtleties such as how swiftly patients open or close their fingers during movement, and the extent to which motion patterns shift between each tap.
According to Dr. Guarin, individuals with Parkinson’s disease exhibit a delayed opening motion when compared to healthy individuals, mirroring the same motion. “This cutting-edge data, only accessible through the video and laptop, significantly enhances our understanding of Parkinson’s disease by allowing for more precise characterization of motor impact, thereby providing valuable markers to gauge therapy efficacy.”
Researchers leveraged the University of Florida’s (UF) powerful HiPerGator supercomputer, initially designed by Guarin for facial recognition applications beyond Parkinson’s disease diagnosis, to train some of its AI models.
“Through the HiPerGator technology, we were able to create a machine learning model that condenses video data into a motion rating,” Guarin explained. “We leveraged HiPerGator to train, inspect, and fine-tune diverse models utilizing massive amounts of video data, ultimately enabling them to execute efficiently on a smartphone.”
Michael S. According to Okun, M.D., director of the Norman Fixel Institute and medical advisor for the Parkinson’s Foundation, automated video-based assessments have the potential to revolutionize medical trials and care.
“According to Okun, the finger-tapping examination plays a crucial role in monitoring Parkinson’s disease progression and assessing illness development.” Currently, interpreting results requires expertise; however, what’s groundbreaking is how Diego and three Parkinson’s neurologists at the Fixel Institute leveraged AI to quantify and standardize disease progression.
As part of a broader effort to empower healthcare professionals, including neurologists, Guarin is collaborating with UFIT to transform this knowledge into a user-friendly mobile application, enabling individuals to monitor and track the progression of their condition remotely from the comfort of their own homes.
