OPatience

Improving your connection with your doctor and your health. All from the comfort of your own home.

Comment

Inspiration

We were inspired by several issues users addressed within the healthcare industries and our own current experiences. By browsing through current news and statistics in the healthcare sector, as well as the push towards remote options after the pandemic, we found that there is a lack of remote connectivity options for patients. Furthermore, doctors' diagnoses often lack transparency and accessibility as they are usually kept within the clinic and not with patient. We sought to solve this problem by creating an application that helps user's to better understand their health.

What it does

Our application strives to enhance and improve patients' relationships with their doctors by connecting with them remotely. We target key pain points that users have addressed in our secondary research. Users outline that they have trouble booking appointments with their doctor. They address that they have trouble tracking their own prescriptions and medication. Lastly, they discuss that they have trouble understanding their diagnosis beyond the doctor's office.

To remediate these pain points, we created these key features. The first feature is a booking appointment feature that allows the user to book appointments virtually with their doctor according to both schedules. The second feature is a prescription tracker that reminds users when to take their medications and shows the doctor's instructions for it. The last feature is a monitoring feature that shows the user's diagnosis over time. Using the data from monitoring and methods of machine learning, we also create an essential risk assessment that suggests areas where the users may be at risk and should contact their doctor.

How we built it

We designed the product on Figma and built the backend using Express and Node.js. The database was stored using SQLite3, and the frontend was built using React.

Challenges we ran into

In order to inform our patients ahead about how much at risk they are to develop a heart disease based on their general health metrics, we implemented a logistic regression data model in scikit learn machine learning library in Python. To do so, our model was first trained on a dataset available freely on Kaggle and then the trained model was used on a patient's health records. However, due to lack of time, our model could not be efficiently trained to yield a higher accuracy. Moving forward, we consider training our model on different datasets as well so its well equipped to better predict risks of other health issues.

In addition, during the initial brainstorming stage, we brainstormed many features which could not all be implemented due to the time constraints of the hackathon. To address this, we prioritized the most important features of the application and moved lower priority and more time intensive features to future development. During brainstorming, we also had trouble laying out a clear flow for the app, but the flow improved as we discussed it more.

Structuring the backend was also a challenge, as we had to brainstorm an efficient way to store many types of data in our app, including health metrics, doctor information, prescriptions and appointments. In the end, we developed a database schema to store data effectively and be able to retrieve it easily before coding it in.

Accomplishments that we're proud of

We are very proud that we were able to plan and implement this comprehensive product which helps people connect with their health, with doctors and improve their overall well-being.

Acknowledgements

The cleaned dataset used to train our logistic regression model was available to us from Kaggle: https://www.kaggle.com/datasets/alexteboul/heart-disease-health-indicators-dataset?resource=download which is a cleaned version of https://www.kaggle.com/datasets/cdc/behavioral-risk-factor-surveillance-system.
Both of these dataset are CC0: Public Domain licensed.

What we learned

Teamwork is crucial. We collaborated and strived to create the best product with our respective skillsets. We learned about the entire product creation process, from designing the frontend and backend, research and implementation of the product. We learned how to collaborate well with a tight deadline and to create a cohesive product.

What's next for OPatience

  • Further training of the ML algorithm to improve accuracy of prediction
  • Enhancing security of data storage in the backend
  • User onboarding process

Built With

Share this project:

Updates