Medi Scanner

We created a website where you can upload existing or take pictures of human injuries, and a ML model will identify the type and provide recommendations on further actions that should be taken.

Comment

LINK TO THE REPOSITORY

Inspiration

Long wait times for non-ER-related injuries and the desire to reduce recovery time so we can get back to life adventures!

What it does

Determines the stage/severity of an injury based on a photo uploaded by the user. The image is compared to a trained ML model that has been fed injury image datasets. A list of treatment recommendations are generated based on the stage of the injury.

How we built it

  • Front end built with ReactJS, HTML, CSS, JS
  • Back end built with Python, Flask REST API
  • ML Model trained with TensorFlow

Challenges we ran into

  • Merging front-end and back-end functionalities
  • Creating a Flask REST API
  • Resolving merge conflicts

Accomplishments that we're proud of

  • Creating a functioning web app within a 24-hour timespan
  • Learning new tech stacks and debugging on the fly
  • Creating our own data set and training our own ML model

What we learned

  • Importance of task assignment and management
  • New tech stacks
  • Training ML models
  • Resolving and debugging merge conflicts
  • Creating simple UI/UX designs

What's next for Medi Scanner

  • Increase ML model accuracy
  • Progression images of the recovery phase
  • Display medical centers near the user

Built With

Share this project:

Updates