Inspiration
When detected early, the 5-year survival rate for melanoma is 99 percent, yet More than 2 people die of skin cancer in the U.S. every hour. My sister, a dermatologist, often highlights how early detection tools can make a significant difference in saving lives. With recent advancements in AI, we saw an opportunity to combine technology with medicine to create SkinSights, an app that uses machine learning and your phone's camera to estimate the likelihood that a mole could be melanoma. By empowering people with accessible tools and spreading awareness, we hope to encourage proactive skin health and potentially prevent melanoma cases before it's too late.
How we built it
We built our Mobile App using the Flutter framework, and a Flask API backend for our Custom Machine Learning Model. To build our model, we utilized TensorFlow, Numpy, and Kaggle Datasets.
Challenges we ran into
Our main challenge was familiarizing ourselves with the IOS environment for development. Especially with older Operating Systems, we had many dependency-based issues with the modern IOS environment
Accomplishments that we're proud of
Despite the lack of resources and only being a team of 2, we were able to create a Melanoma Classification Model with an accuracy of over 90%.
What's Next for SkinSights
Our next steps for SkinSights involve expanding the scope of the model to identify a wider range of skin conditions, from eczema to psoriasis, making it even more versatile as a diagnostic aid. We plan to collaborate with more dermatologists to gather diverse datasets and refine the accuracy and reliability of our AI model. Additionally, we envision integrating a curated store of dermatologist-recommended skin care products tailored to specific conditions to help users take actionable steps toward healthier skin. By continuing to innovate and partner with medical professionals, we aim to make SkinSights an even more impactful tool in the fight against skin diseases.
Log in or sign up for Devpost to join the conversation.