Suna - Physical Therapy Companion

Inspiration

What it does

AI-Powered Exercise Analysis: Suna utilizes advanced AI algorithms to analyze exercises in real-time, providing insights and feedback to physical therapists.

Range of Motion Tracking: The app leverages computer vision technology to accurately track and measure patients' range of motion, enabling therapists to monitor progress and tailor treatment plans accordingly.

Intuitive Reporting: Suna generates comprehensive reports based on patients' range of motion data, allowing therapists to gain deeper insights into their recovery progress and make data-driven decisions.

How we built it

We built Suna using a Nextjs front-end, and tensorflow js and FastAPI for the backend.

Challenges we ran into

A large challenge we ran into was extrapolating and calculating data regarding angles from a tensorflow model. Aside from building a large scale application in a short amount of time, another challenge was working with computer vision and setting it up to work with a front end.

Accomplishments that we're proud of

We are proud of the design, structure, as well as the potential that this application offers. This could potentially save a large volume of patients and physical therapists of money, time, and effort.

What we learned

We learned about working with LLMs such as OpenAI, and also learned about working with computer vision, machine learning, and APIs.

What's next for Suna - Physical Therapy Companion

We look forward to improving the model, deploying, and working with physical therapists to improve virtual monitoring and assessments of their patients' recoveries.

Built With

fastapi
javascript
nextjs
openai
tensorflow

Submitted to

UC Berkeley AI Hackathon

Created by

In 2021, I started attending physical therapy for hip pain resulting from a potential sports-related injury. I continued physical therapy sessions for almost a year before returning to the doctor's office.

Upon my arrival at the doctor's office, they suggested that I simply continue with physical therapy. However, the pain became unbearable due to my inability to perform exercises, prompting the doctor to order further diagnosis that ultimately led to surgery.

If a solution had existed to track my physical therapy sessions based on my range of motion and other metrics, it's possible that I could have been referred back to the doctor sooner, leading to an earlier course of action for my surgery.

Suna serves as a copilot for physical therapists and doctors. It generates a physical therapy exercise/stretch report for users based on the patient's condition and proceeds to track their range of motion over a specified number of weeks/months. It then compiles these metrics into a report, which is sent back to the doctor for reviewing the next course of action.

Paulo Nascimento
sushmullur Mullur

Updates

sushmullur Mullur started this project — Jun 18, 2023 04:58 PM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.