HealthTogether

An efficient online platform that connects doctors and patients.

Comment

HealthTogether

An efficient online platform that connects doctors and patients

My Inspiration

  • Unequal access to health care, especially for those in rural communities
    • For instance, those in rural areas sometimes have to travel up to 200 miles to meet a obstetrician or gynecologist AAMC
  • The shortage in physicians
    • In the next 12 years, the US will have a predicted shortage of between 37,800 and 124,000 physicians AAMC
  • Current telehealth systems can cost between $20,000 and $30,000 SOFTERMII, further disadvantaging underfunded hospitals

My Solution

  • A simple online platform that matches patients with doctors and one another and allows for efficient and safe communication regarding the patient's condition
    • Clean, minimalist Web UI interface allows users to efficiently interact with one another
    • Proprietary recommendation engine matches patients with doctors based upon an advanced set of features
    • ML algorithm that interacts with patients during off hours, allowing patients to receive feedback and assistance when no doctors are present
    • A second recommendation algorithm identifies other patients who have had similar ailments and allows them to chat with one another, providing assistance and comfort from someone who has had a similar condition and allowing for a more intimate and personal connection than with a doctor
    • Database of prescription drugs that have successfully combated certain ailments are recommended to the doctor, helping to increase their efficiency
    • A toxicity algorithm vetts unproductive and rude conversations between patients, ensuring a kind and comforting online space

Development process

  1. Utilized ReactJS for the UI interface, and python for the various machine learning components
  2. Integrated Firebase Realtime database with the UI
  3. Trained a custom toxicity classification model using Cohere
  4. Drug recommendation algorithm created - Colab Notebook
  5. Flask API created to host ML algorithms
  6. Integrated the APIs into the project

User Experience

  • By matching users with others with similar conditions, users can empathize with one another, fostering a comforting online environment
  • Simple UI reduces technical complexities and provides users with a more efficient experience
  • Chat algorithm allows users to receive feedback whenever they need it

User Privacy

  • All data is stored using Firebase, which encrypts data in transit using HTTPS Firebase
  • Additional security measures are necessary and I hope to add these in future iterations of this project

Future Improvements

  • Adding additional security measures
  • Improving the accuracy of the toxicity classifier
  • Further streamlining the user experience (from both the patient and physician perspectives)
  • Creating an app to provide users with a more native experience on mobile devices

Challenges during development

  • Limited time made it difficult to implement all the features that I wanted
  • Using many different languages and APIs made development more difficult

Accomplishments

  • I'm proud of the UI and how it ensures an efficient user experience

What I learned

  • I learned more about designing and creating minimalist UIs
  • Additionally, I learned more about firebase and its various real-time capabilities

Sources

Drug Review Dataset

Felix Gräßer, Surya Kallumadi, Hagen Malberg, and Sebastian Zaunseder. 2018. Aspect-Based Sentiment Analysis of Drug Reviews Applying Cross-Domain and Cross-Data Learning. In Proceedings of the 2018 International Conference on Digital Health (DH '18). ACM, New York, NY, USA, 121-125. DOI: [Web Link]

Built With

Share this project:

Updates