Inspiration
According to a Harvard study, for every hour of face-to-face patient time, doctors spend another TWO hours on paper work. We want to reduce this time by automating the process.
What it does
Chiron is the medical assistant which converses with the user and asks for all the details including disease description, name, age, gender, pain intensity and the image of the affected area. Chiron uses the image to identify the cause and uses other details to prepare a brief report. This report can be very useful for the doctors.
How I built it
We have used google's dialogflow to build a chatbot that converses with the user. Integrated the chatbot with the Google Assistant. The whole code is deployed on google cloud. We have trained a model using CNNs to identify the cause using the image such as rashes, acne, measles, etc.
Challenges I ran into
Building an efficient model which identifies the disease using the small image dataset.
Accomplishments that I'm proud of
What I learned
Working on the image classification was new and enjoyed the task. And also working on google cloud was fun.
What's next for Chiron
The model can be made more efficient with more data. More features can be added such as virtual meeting with the doctor, appointment history, etc.
Built With
- flask
- google-cloud
- google-datalab
- google-dialogflow
- machine-learning
- natural-language-processing
- neural-network
- python
Log in or sign up for Devpost to join the conversation.