NickMandylas/cnn-project

Platform to perform skin lesion image analysis with a convolutional neural network to determine whether the lesion is malignant, and the likely variant.

CNN Project

A proposed online all-in-one platform that aims to easily allow medical professionals to upload images of skin lesions to a convolutional nerual network which will analyse the likelihood of whether the lesion is a malignant form of cancer, as well as the likely variant.

---

This project has three distinct packages in order for the platform to run:

Server

Backend API server built with Typescript.

Technologies utilised:

• Fastify (Web Framework)
• Mercurius (GraphQL Middleware)
• MikroORM (PostgreSQL ORM)
• ioredis (Redis Client)

Web

Frontend Website built with Typescript.

Technologies utilised:

• ReactJS (Frontend Framework)
• NextJS (Frontend Framework)
• Apollo Client (GraphQL Client / State Management)
• Bumbag (Components Library)

Machine Learning

TBD

---

Instructions to run each individual package can be found in the README file of each package.