MediAi

Inspiration

The main inspiratiion of these project comes from seeing people who suffer and wait in long queues to see doctor to conform the type of difficulty the patient faces like fracture, blood clotes and other illnessness that are visible even in minute way. It allows users or patients and doctors to easly locate the illness that they wish to find out in the xray.

What it does

It uses computer vision basically trained YOLOv8 model is being used to detect the illness like fracture, blood clots..etc, It has a trained model which was trained using 25000+ images for fractures and clood clots. Upon uploading futhur scanned image of a human body, it detects even a minute fracture, dislocation, blood clots ets. Uplon uploading the scanned image of a patieent, it stores the image in blockchain technology. The furthur images uploaded are furthur used to train the model into furthur accuracy.

How we built it

It was build by taking reference from doctors, listening to their difficulty and toughness they face in detecting minute facture or dislocations or blood clotes. It reduces the time for patient seeing orthopedics, waiting in the queue to solve their problem. It helps doctors to give treatment in an easy way, reducing their difficulty.

Challenges we ran into

Some of the challenges we ran into are getting patient data and scanned images for training the model. We also neede permission from higher authority and permission from hospital authorities and different patients to train the model. We also faced some problems for getting the trained model ready, and also got problem solving images, as we got restriction from higher authorities.

Accomplishments that we're proud of

We are prooud of accmplishing that the model in compilence with various other trained model and medical research papers and information stored in the database, helping providing primary care to patients like wrapping the fractured area etc.. Deciding on the amount of information that the patient provides.

What we learned

We learned the use of various usages of API keys, various self trained models, prompt engineering, and how computer vision works in the most way. We also learned the proper working of blockchain technology, computer vision etc..

What's next for MediAi

MediAi is set for furthur updates like finding out tumorrs and the point of orgin of the tumors. It also provide an extension for medical sites or hospital managemtn tools to help doctors t get a proper result soon after the scan.

Built With

• blockchain
• computervision
• etherium
• firebaseml
• reaact.js
• supabase