By: Grace Huff
At the Norman Fixel Institute for Neurological DIseases, researchers are pioneering the future of movement disorder assessments with VisionMD, an AI-powered, open-source software that allows clinicians and researchers to analyze motor symptoms objectively, precisely, and efficiently. Developed as an accessible and scalable solution, VisionMD removes the subjectivity of traditional motor assessments and offers a privacy-conscious, user-friendly platform for evaluating Parkinson’s disease, multiple system atrophy (MSA), progressive supranuclear palsy (PSP), and other movement disorders.
Dr. Diego Guarin, one of the lead researchers behind VisionMD, emphasizes how this tool differs from existing video-based analysis platforms. “VisionMD is the first open-source tool of its kind; it also works fully offline, so clinicians and researchers do not have to think about ‘Where is my data going?’ as VisionMD runs locally on the user’s computer,” he explains. “We optimized the machine learning models so they can work on almost any computer; there is no need for specialized hardware.”
Traditionally, clinicians assess motor function using visual observation and rating scales, such as the MDS-UPDRS Part III, which, while widely accepted, remains prone to human variability and inconsistency. VisionMD eliminates these limitations by using AI-driven video analysis to track and quantify body movements, allowing for more standardized and reliable evaluations.
Revolutionizing Clinical Care with AI-Driven Video Analysis
VisionMD was designed to bridge the gap between AI and neurology, making it easier for clinicians to monitor disease progression, track treatment effectiveness, and personalize patient care. The software processes standard video recordings, tracking specific motor tasks such as finger tapping, leg agility, and hand movements, and extracts kinematic data like speed, amplitude, and variability. These insights help clinicians detect subtle changes in motor function over time, something that would be difficult to measure through visual observation alone.
Beyond these capabilities, VisionMD is built with usability in mind. “We worked closely with clinicians and researchers to make sure that VisionMD makes sense for them and that the results we provide are useful,” Guarin shares. “Our goal is to make video-based motor assessments more accessible and standardized, and VisionMD is a stepping stone toward this goal.”
Gabriela Acevedo, a key researcher in the development of VisionMD, highlights the importance of making AI-driven assessments seamless for both clinicians and patients. “VisionMD was designed with clinicians, researchers, and patients in mind. Our lab is leveraging AI to ensure high-quality data collection while automating key steps to make the process seamless for both patients and healthcare providers,” she explains. “The goal is to make video-based motor assessments more accessible, efficient, and standardized, bringing cutting-edge technology into real-world clinical and research settings.”
With its open-source and offline functionality, VisionMD removes key barriers that have previously limited AI adoption in clinical neurology. Unlike proprietary commercial platforms that require cloud storage and expensive hardware, VisionMD:
- Runs entirely offline, ensuring data privacy and security.
- Works on almost any computer, eliminating the need for high-end systems.
- Is open-source, meaning researchers can modify and adapt it for specific needs.
- Processes videos recorded in clinics or at home, making remote monitoring a reality.
Bringing AI into the Future of Movement Disorder Assessments
As Telehealth becomes more common, VisionMD is positioned to support remote monitoring. “With the popularity of TeleVisit, we also hope that VisionMD will become a standard tool to analyze all videos recorded during Telehealth appointments,” Guarin explains. “The video data is already available; it is just a matter of using VisionMD to start analyzing all these videos.”
Acevedo expands on this, explaining the bigger impact of AI-powered assessments in clinical practice. “VisionMD has the potential to reshape how we track motor symptoms. In clinical settings, it could enable remote monitoring, giving clinicians more frequent, objective insights without requiring constant in-person visits,” she says.
This means that instead of requiring frequent in-person visits, patients can submit home-recorded videos for AI-powered evaluation, allowing for more frequent and convenient tracking. This tool has the potential to revolutionize movement disorder care by making disease monitoring more accessible, scalable, and data-driven.
An International Collaboration
This tool was developed as part of an international collaboration between the University of Florida and the University of Würzburg, demonstrating a global effort to improve movement disorder assessments.
Acevedo describes the broader potential of video-based assessments, saying, “In everyday clinical practice, this technology has the potential to enable earlier detection of symptom changes and support more personalized treatment adjustments, ultimately improving patient care and quality of life.”
By combining AI, machine learning, and user-centered design, VisionMD stands apart from other tools. It is more than just a technology breakthrough—it is a practical, scalable solution that directly improves patient care.
Bringing VisionMD to Patients and Clinicians Everywhere
VisionMD is not just new technology, it’s a game-changer for patients and clinicians alike. Instead of subjective evaluations, doctors now have a tool that measures movement speed, consistency, and variability with AI-driven precision. This means earlier intervention, better treatment tracking, and more effective care.
VisionMD is now available for clinicians and researchers worldwide. Learn more here.
To read the paper, click here.