We would like to present our solution named “Pink Pathos”. The workload of pathologists has been increasing considerably and the pain point lies in processing huge amount of histopathology images. Also, significant time goes away in computation and carrying out manual tasks. The timeliness in diagnosis of cancerous vs non-cancerous tissues goes for a toss. Our solution lies in keeping the pathologists as our prime focus and we have come up with a predictive model leveraging AI/ML technology to process huge amounts of histopathology images and thereby determine IDC(Invasive Ductal Carcinoma) which is a type of breast cancer. Our AI enabled model(trained using around 100k images) has accuracy rate of 88%, far better than any human professional. The solution is handy for doctors, patients and support staff. In future, we shall enhance the accuracy rate from current 88% to much higher level, deploy an AI based chat bot, which can handle general queries, use speech translation services and tone analyzer services, so that AI enabled solution can keep a constant vigil on the health of patient.

Built With

+ 52 more
Share this project:

Updates