Inspiration
One of our team members was diagnosed with a femur bone dislocation. But that only happened when his condition was looked at by a second doctor. With certain health conditions, time is of the essence. And lost time could cost the person his life. Our goal was to diagnose health conditions as quickly as possible using Machine Learning.
What it does
rays.ml is a web application that diagnoses health conditions using x-rays within seconds with a click of a button. It maintains a medical history of past treatments, offers a smart chatbot to understand the condition, its symptoms and suggests possible treatment. Additionally, it alerts the registered doctor with every updated x-ray report.
How we built it
We used large datasets to train our model. Since the credibility matters, we used datasets offered by NIH and trained them using Google AutoML. Largely, the tech stack includes, firebase, cloud firestore, cloud functions, the Twilio API, image pre-processing with python and compute engine. We also built a chat bot with dialogflow.
Challenges we ran into
We ran into plenty of challenges such as integrating autoML, making the dataset effective and ensuring the standard of data for training. Further, we faced many challenges with end to end integration with JavaScript and the web app.
Accomplishments that we're proud of
We feel really happy to have done something for medical diagnosis and actually being able to enhance detection and healthcare. It is something that is very important and it is critical to have continual innovation in the field.
What we learned
We learned a lot about different diseases, the importance of it and how one can better diagnose it.
What's next for rays.ml
Scale with more health conditions, ultrasound inferences, and other such diseases as well. Most obviously, larger the dataset, better the model. So it would be beneficial to train model on larger datasets for a longer duration.
Log in or sign up for Devpost to join the conversation.