Inspiration
I read a news article discussing how AI could be used in the healthcare industry in the near future, the most promising method is image recognition for medical scans. I was curious to try this and started with a basic model to classify MRI scans.
What it does
Classifies MRI brain scans on whether a patient has a brain tumour.
How I built it
Using Teachable Machine: https://teachablemachine.withgoogle.com/
Challenges we ran into
Finding already-classified medical images to use as training data was difficult to find. Initially, the project was to diagnose different degrees of burns to advise patients on the best course of action depending on the severity, such as rinsing it under cold water or visiting the hospital.
Accomplishments
Creating my first AI model, which has an accuracy of 80%
What I learnt
- Using Teachable Machine
- How machine learning works
- How an online hackathon works
What's next for Classifying Brain Tumours using ML
For future development, the model could identify which type of tumour is present, such as malignant or benign.
Log in or sign up for Devpost to join the conversation.