Inspiration
One of my friend's mom is currently going through breast cancer. She went through so much pain and stress already given that she is only a couple of month in. She wasn't told the side effects of the drug that she was given and experienced the unnecessary pain in her bone. This could be avoided if she was given the info by the care givers. In addition, she often felt so alone and lonely because the patients' family members and friends are not allowed into the chemo therapy room. This app is inspired because we want to help share the burden that each patient and the patient's family is going through.
What it does
Breast Friend is a single comprehensive application designed to allow family and friends to support their loved ones through their treatment journey. It relies on a series of mile stone predictions of what cancer patients need based on their treatment plans, personal health, age, and other metrics. Through advanced AI, a series of tasks and time associated with the tasks will be recommended for friends and family to volunteer for. Simple tasks like the patient walking their dog can be near impossible in certain stages of the treatment. This app will allow family members to volunteer days in advance.
How I built it
We first generated realistic mock datasets based on our user research and interviews with five breast cancer patients as well as public blogs shared by breast cancer patients. Then, we read the text descriptions in some python data structures. Next, we parsed the small dataset of general dialogue using natural language processing tools such as tokenization, stemming, etc. Finally, we trained a classifier on these data and continue improve the accuracy with new additional data such as julio/ispy-1-trial by using data science and deep learning libraries such as scikit learn and pytorch. In addition, we also used breast cancer waiting times data from hospitals in Northern Ireland to understand the waiting time that patients went through to predict the waiting time for patients so that this app can help support other aspects of the patient journey.
Challenges I ran into
How to collect the necessary data and the critical mass needed to derive data driven reccomendations.
Accomplishments that I'm proud of
What I learned
The patient journey is much lonelier than expected. Friends and family have good intentions but often assume what to do and it is not as impactful as they intended.
What's next for Breast Cancer
As we have already tested our app with dummy data. We hope to test it with our friends and their family to understand how can this app best assist them. In the future, we hope to partner with local hospitals and national insurance company to truly ease the pain for the patients financially and emotionally.
Log in or sign up for Devpost to join the conversation.