VisionMD: An Open-Source Tool for Video-Based Analysis of Motor Function in Movement Disorders
1Movement Estimation and Analysis Laboratory, Department of Applied Physiology and Kinesiology, University of Florida.
2Department of Neurology, University of Würzburg.
3Fixel Institute for Neurological Disease, University of Florida.
4Crayton Pruitt Family Department of Biomedical Engineering, University of Florida.
5Department of Neurology, University of Florida.
Abstract
VisionMD, an open-source software for automated video-based analysis of MDS-UPDRS Part III motor tasks, offers precise, objective, and scalable assessments of motor symptoms in Parkinson’s disease and other movement disorders. Leveraging deep learning, VisionMD tracks body movements to compute kinematic features that quantify symptoms severity and supports longitudinal monitoring. VisionMD’s user-friendly, customizable framework empowers clinicians and researchers to objectively evaluate motor symptoms in persons with movement disorders without specialized hardware.
Pipeline

Tutorials
Installation Instructions - MacOs
-
Unzip the File
Once downloaded, unzip the file. -
Launch the Application
Double click onVisionMD
. -
Open the Terminal
A Terminal window will open, showing some progress while the program starts. -
View the Startup Message
After a few seconds, you will see the following message in the Terminal:Django version 4.2.11, using settings 'backend.settings' Starting development server at http://127.0.0.1:8000/ Quit the server with CONTROL-C.
-
Open Google Chrome
This program has only been tested with Google Chrome; other browsers might provide unexpected behavior. -
Access the Application
In the URL bar, type:http://127.0.0.1:8080
and hit Enter.
VisionMD will open, and you are now able to use the software. -
Close the Application
Once you are done using VisionMD, return to the Terminal window and hitCONTROL + C
.
Close the Terminal window and then close Google Chrome.
Sample Videos
Click to get sample videos to process using VisionMD
How to download videos from GitHub
- Select the video you want to download
- Click the download button

Kinematic Measures Sample Data
This data was collected from two cohorts of Parkinson’s Disease patients to illustrate VisionMD’s capability in assessing the impact of different therapies. Each patient underwent two recording sessions: one with their therapy “on” and another with it “off.” The therapies assessed included:
- Deep Brain Stimulation (DBS)
- Dopaminergic Medication
The Excel file contains each row as a separate recording. For more details and to download the data, visit the Kinematic Data GitHub Page.
A sample code analyzing the data to reproduce results presented in the paper is provided. Analysis includes a paired t-test for each kinematic measure to compare “on” and “off” therapy sessions. This analysis is performed separately for patients undergoing Deep Brain Stimulation and those receiving Dopaminergic Medication, allowing users to assess the effects of each therapy on motor function.