Boobee

A data driven approach to personalized Breast Cancer Detection

Comment

Inspiration

This project was built out of the lack of data-driven recourses that help monitor women’s health. Clue, an app that tracks women’s menstrual health was a source of inspiration at the start. The people we hope to affect with our project are the women in developing countries who do not have the same level of healthcare as the US. 1 in 8 women will develop breast cancer in their lifetime, so helping even a small percentage of women would be impactful.

What it does

The final product—a web application—allows users to input personal information, including age, race/ethnicity, family history, age at menarche, weight, height, and menopausal status. The information is then used in an API call to a statistical model, which returns the user's probability of having breast cancer.

How we built it

Trista and Archie had a strong technical foundation in data analytics and statistical modeling, and Wendy and Andrew had experience with product development. Trista and Archie conducted statistical modeling and figured out what variables can enhance the accuracy of the model. After finalizing the statistical model, Trista and Archie uploaded it to Domino Data Labs, where the model could be accessed as an API. From there Andrew developed the UI/UX in React hosted it via Netlify, allowing us to apply the statistical model to the user's data.

Challenges we ran into

One of the main challenges we faced was to connect the statistical model written in R with the Front End written in JavaScript. Since each of us specializes in different skills, this was our first time to connect different languages to complete one project. After we researched and tried different approaches, we managed to connect the two, seemingly incompatible, pieces of software by turning the statistical model into an API that JavaScript could call.

Accomplishments that we're proud of

Overcoming our greatest challenge was our greatest accomplishment. In addition to that, we are proud of completing the Boobee website that is accessible and interactive, figuring out the appropriate statistical model to predict breast cancer probability, and putting it all together

What we learned

Through working with teammates from diverse backgrounds, we gained deep insights into how to complete a project that integrates data science, user interface design, and computer science. Not only did we learn new coding skills to achieve integration, such as building an API, but also, we began to appreciate the communication and coordination required between data scientists, software engineers, and UX designers to complete one project.

What's next for Boobee

We plan to complete the Webapp and build an IOS/Android App to help women better understand their breast health. Specifically, we envision implementing a machine-learning algorithm to facilitate the early detection of breast cancer and provide an individualized guideline for breast care. Moreover, we aim to include more features, such as breast self-care tutorials, cancer survivor stories feed breast health analysis report and medical provider resource matching. The mission is to constantly monitor women’s breast health and make care more accessible. Additionally, we hope to connect the users of Boobee to healthcare providers so that they can better understand the breast condition of the patients based on the data.

Built With

Share this project:

Updates