Diagnos : deep learning based cdss for lung disease prediction
This is a clinical diagnosis support system for pulmonologist and radiologist, which intakes x-ray images of lung patients and successfully classifies lung diseases in patients.
Videos
Description
Diagnos : is a an intgration of computer vision in the field of healthcare, application of computer vision holds the power to revolutionize healthcare system, we set to achieve a high performance yet lightweight CDSS clinical diagnosis support system, which will enhance the diagnostics of pulmonologist and radiologist, while complying to healthcare standards of declassified data and transparency of model.
our CNN is based on densenet169 which has 169 layes and uses 7x7 grid for convolution, this model is created in pytorch and then converted to tensorflow.js for execution in backend without a python runtime, for the chat interface we opted for gemini from google as it provide smart responses in comparison to chatgpt and likes of other, this all is completed by the website interface which is created using MERN stack.
Progress During Hackathon
70
Tech Stack
Github Link
https://github.com/Teletetra/DiagnosAIWinner Track
First Place
Hack Krmu
Winner Track
First Place
Hack Krmu