H1

The lack of accessibility to healthcare information and increase in misinformation has greatly impacted our team’s members. Siya had to book an hour-long appointment just to get information about her iron deficiency. Aadi tried to research drugs that could improve his strength for bodybuilding but couldn’t find a lot of information. Natalie got misdiagnosed for a stomach bug and Shreyas got misdiagnosed for an eye disease.

In addition, doctors are struggling with an intense data overload, and need strong data analytics solutions to make use of all the data that a patient provides. In a talk by Dr. Paul Tang at Treehacks, we learned that 81% of the time doctors cannot find answers to important questions that come up during diagnosis. Some of these include “What medications have been prescribed to this patient” (54% deficiency) or “What were the results of a specific diagnostic test or procedure” (36% deficiency). By creating a full biological profile of an individual patient, we could bridge the gap to help doctors and patients make more informed decisions together.

This is why our project tackles accessibility to healthcare information. It had four major parts: Information (built by applying few-shot learning to an LLM) search engine to review recent and relevant scientific literature to make it easy for the average person to receive medical information (such as potential side-effects to drugs they are diagnosed and a summary/more information about diseases) Forum / Discussion board to connect people to form communities around different diseases and drugs Personalized health profile based on medication tracking and surveys related to nutritional/dietary data Video “office hours” from real doctors to write prescriptions on-the-spot based on user research and/or providing support during treatment All our teammates learned a lot as we worked on aspects we have not worked on before. Natalie and Aadi learned a lot about how to use and implement HTML. Siya learned how to use cloud databases and openAI by using InterSystems. Shreyas learned how to implement additional features into the API. Our biggest challenge was integrating all of the aspects of our project together. This required a lot of communication. For example, Natalie and Aadi had to communicate with each other a lot to implement the forum into the website’s main HTML code. Shreyas and Siya had to work with each other to implement backend into the API. Shreyas and Aadi worked together to implement the backend and front end. A lot of this code also required a lot of debugging which was also a difficult process. Nonetheless, we all learned a lot about teamwork and aspects of software development that we have never learned about before.

We used Dolby IO to create the communication functionality between doctors and patients. We also working on using Intersystems to create our data health profiles.

We also believe there is huge potential for the future of this project. Data profiles will help solve a pressing need in healthcare, and allow for better diagnosis and patient safety. Data analytics as we have started doing them will revolutionize the future of medicine.

Fun fact: We all known each other for almost two years. We met as high school juniors when we were accepted to the Research Science Institute at MIT. We chose to go to Stanford because of each other and are now super close friends. Our experience at TreeHacks helped us bond together as we ideated, coded, debugged, attended events/workshops, and had fun together.

Built With

• css
• dolby
• express.js
• html
• javascript
• node.js
• openai
• pyscript
• python