1.5 million incorrect prescriptions are sent out to pharmacies every year. Some are caused by misdiagnoses and others are caused by a lapse in the transfer of a patient's health records. We decided to solve these problems by creating a web app that uses several APIs to confirm doctors' diagnoses and simplifies the transfer of patient records from one doctor to another.

What it does

Pulse confirms doctors' diagnoses of various diseases by listing known symptoms of the diseases which can often be overlooked by doctors. Pulse also makes switching doctors easier by ensuring electronic health records are securely transferred over.

How I built it

To build Pulse, we used the IBM Bluemix Watson API to find symptoms that match the doctor's predictions. We also used the Epic FHIR API to collect sample data about patients, and to better model our storage and analysis of patient data. Finally, we used MongoDB to store user accounts and patient data, Linode to host our application, Python and Flask for the backend, and HTML/CSS/JavaScript/Bootstrap for the frontend of our application.

Challenges I ran into

At first, we had trouble figuring out the most important data points to collect, and how to model the storage of our data. Luckily, the Epic FHIR API, which provides sample data about patients, helped us overcome this obstacle.

Accomplishments that I'm proud of

We successfully figured out how to generate dynamic PDFs that contain the most crucial elements of the patient's medical record. Additionally, we managed to extract meaningful descriptions of various health conditions from the IBM Bluemix API. Finally, we came up with an efficient data model that has the most useful information about patients, which was largely the foundation of our project.

What I learned

We learned how to use data technology to help doctors make more meaningful decisions for their patients. We also realized the true importance of easy information flow in the healthcare field.

What's next for Pulse

We want to expand our project to collect data from more resources for more accurate readings. We also want to add other features to enhance the doctor and the patient's user experience.

Share this project: