AI in physical therapy
Completed
ARTIFICIAL INTELLIGENCE
COMPUTER VISION
HEALTH & PHARMA
About Project
Muscle injuries can be painful and require a long time to recover from. We developed a solution to effortlessly detect and recover from muscle injuries with AI. Using the camera on the user’s mobile device, the app detects and diagnoses injuries by measuring the range of motion for specific movements. When an injury is suspected, the app recommends personalized recovery plans to get you back to your best in no time.
our role
-
artificial intelligence
-
computer vision
client background
Founded in 1997, Netkinetix specialize in software and business applications to satisfy critical business needs. They provide a wide range of development services and software application products for multiple technology platforms that meet the ever-changing needs and demands of the marketplace.
challenges
Human pose estimation is a well studied problem in machine learning, however it’s always challenging to build a model to be used in the wild.
- Users use the app in different environments with different visual conditions
- Based on the injury, specific body parts need to accurately know which muscle is injured
solution
The mobile app first ask the user which body part is causing the pain, based on that a specific set of movements are given to the user to perform while recording a video
- Users are given instructions about how to face the camera for each movement for better detection
- The muscle range of motion is calculated for each of the given movements
- If the range of motion is not in the normal range, the injured muscle is detected and a recovery plan is provided
outcome
The model was able to successfully detect injuries with high accuracy and showed robustness across different recording conditions
Let's See Our Client's Testimonials