MedBot

Creating equal healthcare for all using AI!

Comment

Inspiration

Every day, those of color, lower economic standards, and more face significant challenges in accessing affordable healthcare within their proximity. When we went through a drastic pandemic to make things worse, it brought these disparities in healthcare access into sharper focus and further revealed the troubling and unfair issue: those who needed help couldn't get any. This eye-opening experience inspired me to use what I know and have learned about app development to create a free and accessible tool while making a change in society. I want to help others get the access they need using modern technology and tools like Artificial Intelligence and Machine Learning while giving other resources through the click of a button.

What it does

MedBot gives users access to different AI toolkits and healthcare material. First, users can utilize an AI doctor for proper diagnosis and accurate measurements of their specific symptoms. Users can type in specific symptoms they have and MedBot's NLP model will handle the diagnosis. Furthermore, users can get advice using LLMs and generative-text models, which generate medical advice for the user.

Next, the user can use news features and NewsNow API, connected with MedBot, to seek the latest medical news about a specific topic. It comes with an in-built text reader for each news article that you want to read.

Finally, we used iOS HealthKit and Android Health Tracker to track different metrics such as calories burnt and steps taken.

How we built it

I utilized Flutter for the front-end and back-end of the app. I also utilized Python and different ML tools in building the NLP model for the HealthDiseaseDetector. In reference to Machine Learning, I also utilized OpenAI API services and HuggingFace API services to help with generative text and prediction refinement. I used NewsNow API to get current news and filter medical items for the user's searches. Finally, I used Firebase and Cloud Firestore to maintain and store user documents as well as perform user authentication.

Challenges we ran into

The biggest challenge we ran into was getting the different APIs to work smoothly and not interrupt the other APIs in the process. If we have one API running, we don't have to worry about much. However, since MedBot has multiple APIs running in the background and collecting information for the app, I had to be careful when programming and maintaining the security of the code as well as the processes in the back end. It took many tries and failures to get it working but we were able to get it.

Accomplishments that we're proud of

I think that the entire project -- the lines of code, the different keywords, the multitude of processes, etc. -- was my biggest accomplishment. I was able to get so much done and learn so much about Flutter, Firebase, Python, Machine Learning, etc. At first, I was scared of doing this project, but through my accomplishments, I'm proud of myself for conquering another challenge and creating something fun and helpful.

What we learned

While I did learn a lot about different programming languages such as Flutter and Python or different concepts like Machine Learning and persistent storage, I believe the biggest thing I learned was the enjoyment/fun in conquering a challenge. I looked at this project and hackathon as a big, steep hill and something I was tentative of doing. Yet, as I pushed myself to complete my project, I learned that you cannot pass without failing first (something I learned multiple times on this project).

What's next for MedBot

I want to include a helpful system by creating an important Google Maps interface. I want to give users another feature of locating different hospitals, doctors, therapists, etc. based on the user's needs. This would take a big step in MedBot and something that would really take MedBot to another level in terms of helping others and making an impact on society.

Built With

Share this project:

Updates