Parkinson's is a progressive neurological disorder that affects movement, balance and coordination. It gets progressively worse over time and currently has no definitive cure.
Each year, nearly 90,000 people in the United States are newly diagnosed. The disease often begins with subtle symptoms—slight tremors, changes in walking pace—that can be easy to miss. Accurate monitoring is crucial, yet many patients only see their specialists a few times a year, making it harder to track these gradual changes.
Diego Guarin, a University of Florida researcher exploring how artificial intelligence can transform the way Parkinson’s disease is monitored, has been studying how simple video recordings can capture what the human eye might miss.
His work, which started with a single research connection in 2018, has now grown into an app, VisionMD, designed to help patients track their symptoms at home. Allowing those subtle symptoms to be caught earlier.
WUFT’s Krystal Felix sat down with Guarin to discuss how this research could benefit patients in the future.
This interview has been edited and condensed for clarity. Listen above or read a slightly longer version below.
Q: What inspired you to make those first steps to apply AI to detect early movements in changes linked to Parkinson's?
GUARIN: So I've been working with people with movement disorders for several years now. I started with Amyotrophic lateral sclerosis (ALS) [a fast-progressing disease that damages nerve cells, making it harder to control movement], but ALS progresses too fast.
So during that time, I also connected with researchers working on Parkinson's disease. Basically we observed that first, there's very little understanding of the motor symptoms. There’s mostly personal assessments, so there's very little objective assessment. And also people with Parkinson’s disease, most of the time they progress relatively slowly. So, we can really see through the time how the disease changes.
And that is really interesting to me, because we have a way to measure the brain, and then we measure as this disease progresses and there's a correlation with the brain.
Q: How does the AI technology analyze the hidden signs, how does it work?
Guarin: So when a clinician sees a patient with Parkinson's, clinicians have worked for many years developing a set of measurements that they can then reliably assess. Specific guidelines on how to assess the movement of a Parkinson’s patient it’s called the Unified Parkinson's Rating Scale. And it provides a set of tests that clinicians can do to evaluate the condition of the patient on the severity and they receive a number based on that. But it's the clinician making the assessment and I would say it’s completely subjective, but that is not true it is very much influenced by the experience of the clinician.
So, we record videos of patients performing exactly the same task the patients do in the regular clinical evaluation. And then we use our techniques to analyze those exact patients and obtain information that the clinician is already familiar with, so they can easily interpret.
Q: Since this has been implemented in other places globally, can you speak to how it’s affecting or helping patients already?
Guarin: Yeah, we were very lucky some researchers in Germany were doing something very similar to us just in a slightly different way. So, we started working together and they adopted our approach.
Now, what they are doing is pushing the boundary a little bit more. They are trying to see if using these videos, they can suggest the appropriate therapy for different patients. So, kind of giving the clinician more objective information that then will guide therapy because they have been able to find how these measurements that we obtained from video are affected by, for instance, by medication [one of which being Levodopa, a common drug that replenishes dopamine in the brain], deep brain stimulation, which is another treatment that people with Parkinson’s take.
So, if we can quantify how this subjective measurement is affected by these therapies then maybe we can suggest a therapy.
Q: How is this development being turned into an app and what does that mean and how can it reach more people?
Guarin: We created an app where people can record videos at home. We were careful to make sure a person with Parkinson’s, who has limitations with movement, can actually work with the app. It has automatic recording starting and stopping based on conditions we were able to make.
So with this app, basically the person puts the phone in a tripod, and they sit in front of the app and they can record multiple movements, which are then assessed automatically using the different algorithms that we have created. And they get this tracking of the different motor symptoms from home.
We are currently testing the app. We have roughly 30 people involved who are recording themselves at home and seeing how their motor symptoms change over time. We want to see if this again coincides with the clinician, maybe the app will give more information. Maybe the app will be more sensitive to changes, and maybe help people to be more proactive to tackle changes that are occurring instead of just waiting for an appointment.
Q: What is the most rewarding part of this research and what makes you so eager to continue this research?
Guarin: We know that patients have problems getting clinicians to understand how they're feeling. What is going on with them is not the patient's fault. It's not the clinicians fault, it's just the lack of a common language, to talk about what is going on with the patient. We believe that this might be a solution. So that is very inspiring to us because we say, okay, if we can help these patients to get the medication that they need, to get the treatment that they need to better communicate with their patient, then that would be fantastic.
And I believe that this solution is going to help with that. So, yeah we are trying to make it a reality.