Original sign of syphilis
Denoised sign of syphilis
Identified zone of interest
Worldwide statistics for STIs are alarming, and yet their stigmatisation and slow healthcare administration often cause delays in treatment.
What it does
From an intuitive iPhone app, the product provides easy, fast, reliable smartphone-based diagnostics for oral forms of STIs, which individuals can use in the comfort of their own homes.
How we built it
The back-end (serverside) is built in Python using OpenCV for computer vision. The front-end is both built using Flask and Swift. Medical accuracy was ensured using the recommended best practice and up-to-date medical textbooks.
Challenges we ran into
The lack of available data made Machine Learning (the prefered choice for such an application) impossible.
Accomplishments that we're proud of
- The medical accuracy of our system
- Successfully built a working iOS app
- We believe this could be a revolutionary creation, both for worried patients around the world struggling to see a doctor for their condition, and for over-subscribed health services.
What we learned
Most of the software was used for the first time, hence heavy learning was involved in every part of this project. We also learnt a lot regarding STIs and the worldwide statistics involved, which only strengthened our resolve that this is an important problem that needs addressing.
What's next for Munsjuk
Compiling a strong dataset that will allow us to build a more robust model using convolution neural networks. Adding more symptoms to our detection library to enable detection of a larger range of oral diseases.