Skin cancer is deadly. On average, one American dies from skin cancer every hour. Yet, it doesn’t have to be this way, because skin cancer, unlike some other cancers, is highly treatable and has a high rate of recovery if recognized early. Curif.ai is an AI-based Android application that fits in your pocket with the power to be able to diagnose various common skin cancers and conditions in a matter of seconds by harnessing the power of deep learning. A patient simply needs to take a photo of their skin and the picture is sent to our backend Firebase server where it is fed through a custom-built Convolutional Neural Network trained from a large image database of skin conditions in order to give a diagnosis. Users can track the pain they feel as well as the disease progression in aesthetic charts, and our algorithm uses logistic regression to give an estimated date of recovery. To prevent the isolated feeling that patients feel after a grim diagnosis, our app creates a support network and facilitates easy inter-patient interaction to expedite recovery. Lastly, we recognize the need for constant invention and innovation in the field of medicine. To promote the process of finding cures for chronic skin conditions we allow patients to opt to share their data with specific research studies looking for a cure. Overall, our app levels the playing field for healthcare and ensures that everyone, regardless of time, location, or socioeconomic status has access to life-saving dermatologist-level diagnosis tools.

What inspires you (or your team)?

Melanoma and other related skin cancers can often be attributed to the extensive UV rays present in sunlight which can wreak havoc on skin cell’s DNA. My (Sreehari) parents spend countless hours in the sunny reserves of national parks walking my two dogs, so when I learned about this deadly disease I was instantly anxious and worried about the health of my parents. The good news for me was that Melanoma and other skin cancers respond well to treatment if detected early; which is where the inspiration for our app came from. We set out to develop an app which people could use to detect the early onset of deadly skin diseases. With a concept in mind, we next set out to determine the features and functionality we would implement. In order to determine the various components of our app we asked our classmates and teachers about the functionalities they would like to see in a healthcare diagnosis app. Unanimously, they responded that they wanted the ability to diagnose themselves using the Android camera, track the progression of their disease, connect with fellow patients for support, and contribute their data toward finding a cure. Lastly, we wanted the User Interface of our app to be clean, minimalistic, and easy to use. So we were influenced heavily by the Google Material Design Spec which outlines best practices for apps aiming to be very user-friendly.