Patalah

At-risk Breast Cancer patients participate in multiple tests and frequently experience misdiagnosis. Patalah aims to eradicate this by providing an accurate solution that requires only one test.

Comment

Motivation:

When we came across the ELC 2020 Virtual Hackathon we couldn’t resist participating. We know someone personally who has been diagnosed with Breast Cancer and therefore also knew the challenges that came along with the disease. We wanted to help at-risk patients overcome said challenges.

What it does

We focused on three problems: Inaccuracy and unnecessary rounds of Mammogram testing (increasing costs) Misdiagnosis and delay by doctors (prognosis) The pandemic restricting access to quality healthcare

We developed an ML model and integrated it with the Patalah website, the website also has information about Breast Cancer in an easy-to-understand format. By feeding in FNAC values the model will accurately predict whether your cancer is benign or malignant.

The site also has a map that will show Breast Cancer Clinics near your area.

How we built it

We decided to use an ML model to overcome the problem with doctors. We opted for Ensemble Learning as it has better accuracy and precision than any single model. The website has information about Breast Cancer and its symptoms in an easy-to-understand format.

We chose a dataset that contained FNAC values (statistically and clinically more accurate). Based on 30 attributes the ML model can predict whether a patient’s cancer is benign or malignant (96-99% accuracy). Our aim was to overcome the challenges of false-positive reports, false-negative reports, and overdiagnosis.

Challenges we ran into

Embedding the ML model into the website was tedious. Coming up with the best model was also difficult. We had to make a lot of changes to the UI to make it aesthetic. Putting the video together was time-consuming and hard.

Accomplishments that we're proud of

The fact that we have a working model is an achievement in itself.

What we learned

We learned efficient division of labour despite the pandemic restricting communication. We learned how to work smart rather than hard and gained knowledge about Breast Cancer and its effects.

What's next for Patalah

We believe that this is a practical solution that will certainly reduce costs and ensure more accuracy when it comes to detection. Keeping in mind Social Distancing norms, Patalah aims to keep the battle against Breast Cancer alive.

+ 20 more
Share this project:

Updates