DiaSense

App that predicts diabetes, provides data analysis, and summarize trial data using AI and Machine Learning

Comment

Inspiration

There is a need to accelerate clinical trials to bring potential new treatments to patients more quickly in term of:

  • Data Analysis
  • Data Result Prediction
  • Trial Study Summarization

What it does

The apps offers different services that speed up different stages of the clinical trial:

  • Trial Data Vizualization: Data Analysis Charts for Trial Data
  • Diabetes Prediction: Based on Patients' Data
  • Trial Notes Summary: Speedy Summary for Long Trial Notes
  • Consent and Recontact: The app informs patients that they give consent for us to gather data and recontact them in the future in exchanging for using the services

How we built it

  • The first service is trial Note Summary that summarizes the notes of long trial observations for fast and effective conclusion which was built using AWS and Claude LLM.
  • The second service is diabetes prediction that cut down the time to analyze and find conclusion thanks to the use of pythons and machine learning algorithms.
  • The third service fastens the data analysis process by offering a data visualization board that was built using Tableau and patients' data.

Challenges we ran into

We had challenges to brainstorm different ideas to speed up the clinical trials. We also tried to find functions that are feasible for accelerate the process. We had to first design the whole app, then build different functions and then connect them together to a completed app.

Accomplishments that we're proud of

We were able to build different functions that accelerate various stages of clinical trial and bring effective as well as established treatments to the patients.

What we learned

We learnt to build app using Tableau, AWS AI, Claude LLM, python, and machine learning algorithms

What's next for DiaSense

  • We would like to enhance the dashboard to take in more dynamic data from patients
  • We want to improve the prediction models to return more accurate results
  • We want to enhance the conversation of the AI to make it more human like.
Share this project:

Updates