Twacha is a mobile app available on both Android and iOS platforms, designed for the early detection of skin cancers. Imagine you've just returned from a vacation spent on the beaches, and a few weeks later, you notice a peculiar dark spot on your skin that wasn't there before.
In most cases, it might be something harmless, but there's a less probable chance that it could be a form of skin cancer. Twacha can assist in the early detection of skin cancers, providing you with valuable information to help you decide if consulting a doctor is necessary. While consulting a doctor is always the first step, Twacha can help you make sure whether the skin lesion is indeed a potential concern, allowing you to schedule that doctor's appointment you might have postponed due to a busy schedule.
It's worth noting that Twacha's results may not always be definitive. Therefore, it is essential to use the app as a supplementary tool and not as a substitute for professional medical advice.
Twacha is built using kotlin multiplatform support and compose multiplatform
Almost all code is shared and in kotlin.
Twacha is a kotlin based project. However, it has authentication included as well as it stores data on mongoDB. So there is a backend written in node js for Twacha as well. The backend is deployed on vercel.
You can find the backend code here https://github.com/raehat/twachabackend.
The code is deployed on https://skinappkmmbackend.vercel.app/
To analyze image and produce results, a ML model is written in python and compiled with pytorch. You can find the jupyter notebook here https://drive.google.com/drive/folders/1DjtGW9PrzQrFjsMVBr1db6qtj3sjGc5J?usp=sharing
ML Model is deployed in a flask app. Flask app is deployed on railway and Twacha app communicates with this flask app to analyze images with skin lesions. Flask App code can be found here https://github.com/raehat/twacha-ML-backend.
Flask app is deployed on https://flask-production-9f74.up.railway.app/
To run mobile app, follow through
- Open terminal
- Run command
git clone https://github.com/raehat/twacha - Run command
cd twacha - Open project in Android Studio
You can build the android app as well as ios app from android studio.
If you want to run the backend locally, follow through
- Open terminal
- Run command
https://github.com/raehat/twachabackend - Run command
cd twachabackend - Run command
npm install - Run command
npm start
Node js app will run on http://localhost:3000/. Open Twacha Kotlin Application in Android Studio. Go to Twacha -> composeApp -> src -> commonMain -> kotlin -> Data -> Network.kt
You will find AUTH_URL here
Replace original value of AUTH_URL with http://localhost:3000/
If you want to run flask app with ML Model locally, follow through
- Open terminal
- Run command
https://github.com/raehat/twacha-ML-backend - Run command
twacha-ML-backend - Run command
pip install requirementx.txt - Run command
python app.py
Flask App will run on http://localhost:5000/. Open Twacha Kotlin Application in Android Studio. Go to Twacha -> composeApp -> src -> commonMain -> kotlin -> Data -> Network.kt
You will find ML_URL here
Replace original value of ML_URL with http://localhost:3000/