RehabTrack

Routine List - Patient Dashboard
Patient Profile and NFT Display
NFT (minted) - proof of exercise completion & personal collectible
Gemini AI accepted exercise and gives suggestions for improvements
Doing something other than the prescribed exercise results in a fail
Complete all exercises in routine to claim an NFT, viewable in PolygonScan
Doctor is able to see patients' NFTs to verify exercise completion
Doctor Is able to administer AI generated exercises to personalize routines per user

Inspiration

Studies show that up to 70% of physio patients do not fully complete their prescribed exercises. Low adherence to prescribed home physiotherapy exercises leads to stalled recoveries, recurrent injuries, and wasted clinician time. Tracking compliance remotely is difficult, and patients lack motivation without timely feedback or tangible rewards. RehabTrack was born to bridge that gap by combining AI-validated exercise monitoring with gamified, blockchain-based incentives - turning rehab routines into a trackable, rewarding experience.

What it does

Doctors prescribe exercise routine to patients. Patients verify routine with AI and get rewarded a minted, personalized NFT to show as proof and a token of completion. Patients can proudly show off their NFTs.

AI Video Validation: Uses real-time camera analysis to confirm correct form, count repetitions, and score performance.
NFT Rewards: Automatically mints a unique collectible token whenever a patient completes a session to a predefined accuracy threshold. These NFTs serve proof of exercise completion and incentivizes patients to be proactive in their healing journey.
Personalized NFTs: Gemini image generation API generates a custom made NFT based on exercise done
Doctor Dashboard: Provides clinicians with a timeline of completed exercises, performance metrics, and proof-of-completion NFTs.
AI Routine Generation: Generates selectable exercise for doctors to add to a custom made routine
Patient Dashboard: Delivers exercise instructions, form corrections, and a gallery of earned NFT badges.

How we built it

Front End: Next.js for both the clinician portal and patient interface. Back End & Database: Supabase for user authentication, exercise prescriptions, activity logs, and NFT metadata. AI Integration: Google Gemini Video Understanding API for pose estimation and form analysis, wrapped in serverless functions for real-time feedback. Google Gemini to generate custom exercises based on user medical conditions. Image generation API to generate custom made NFT for users. Blockchain: Thirdweb SDK to generate and mint NFTs on Polygon Testnet, with MetaMask wallet integration on the client.

Challenges we ran into

NFT integrations and functionality
First time using Supabase
Gemini integration (3 different APIs!)

Accomplishments that we're proud of

We are very proud that we were able to integrate many new technologies that none of us have worked with before. It was our first time working with Gemini API, Supabase, and NFTs in general. We are especially very proud of our perseverance to keep pushing forward, even when we were discouraged at times.