Health is one of the most valuable assets of human lives. The way the healthcare system is organized can hugely impact the quality of life. The current healthcare system is far from being perfect and we see an opportunity to improve the transparency of medical services. We recognize that advances in technology have great potential to take preventive care to a new level.

What it does

We propose a new tool that helps a patient to efficiently upload, store, structure, access, analyze and visualize their own medical records based on their patient history. Using our framework, one can:

  • Take pictures and upload all the medical records (summaries of doctor visits, lab and test results) to their private accounts
  • Access and visualize data from personal electronic devices
  • Get alerts and recommendations based on their data (preventing screening, appointments, yearly physicals)
  • Track changes in medical history (weight, cholesterol level, vaccinations)

How I built it

Front end (JavaScript, React, HTML/CSS) :

  • Incorporation of modern UX design principals
  • Adopting best practices for website development
  • Creating responsive webpages with scalable content following Bootstrap techniques

Back end (Python, Flask):

  • AI-based optical recognition of uploaded images
  • State-of-the-art clinical named entity recognition software for automatic detection of medications and tests from raw record text
  • Detection of general entities (organizations, names, locations, dates) from note summaries using natural language processing libraries
  • Usage of external resources and rules for proposing recommendations based on the detected data

Challenges we ran into

  • Lack of experience with the used frameworks
  • Limited time to create and fully test the prototype
  • The challenge of narrowing the scope of the project and prioritize our tasks

Accomplishments that I'm proud of

  • Productive discussion and coming to an agreement about project ideas
  • Ability to break the project down into functional units and parallelize the work
  • Development of a functional web application prototype

What I learned

  • To better understand and take part in all stages of full-stack development
  • To apply machine learning techniques to a real-life problem
  • To work efficiently as a team in a time-constrained environment

What's next for MedUnited

While we have ...we understand there is still a lot needs to be done to improve and sharpen our idea. One of the main concerns is user privacy that naturally comes with our project. Our highest priority for the next step will be to make sure that our user data is reliably stored. We intend to achieve this by using the secure multi-party computation approach to ensure the integrity of data storage and access. Once successfully implemented, we can additionally apply AI-driven data analytics for more targeted patient recommendations and better overall user experience. Other features that we would be excited to see in the product are:

  • Ability to share information with family members to document family medical history (especially useful in case of emergency situations)
  • Management and visualization of medical expenses and insurance information
  • Development of the application for mobile devices and using their cameras
  • Take into account genetic information to perform a various health risk assessments
Share this project: