RASHionale is an AI driven app that has been trained on images of various skin rashes to, given a user’s image of their rash, return possible diagnosis for their condition. The app itself is coded in Xcode using Swift, and the machine learning model (a Convolutional Neural Network) is coded in Python. To use the app, the user first inputs their basic information, such as name, age, gender, and medical history. To take a scan of an area of concern, the user will press the “New Scan” icon and follow the prompts. Upon completion, a results page will show up, which shows the likelihood of having each disease, and returns a probability scale of 1 to 100. These are calculated using a convolutional neural network, or CNN, a trained machine learning model, to make accurate decisions based on the information given by the patient. If the probability exceeds a certain threshold, users will be recommended to contact trusted physicians in the area, and will be put higher on the priority list because they have a potential issue. Users can also contact local physicians directly through the “Healthcare” option. The RASHionale app solves the issues of accessibility and cost of dermatologists, as it is available across all platforms and is free to use and users only need to pay if they need further dermatologists’ assistance. It will help save more lives and help people live a fuller and healthier lifestyle.

What inspired you (or your team)? 

Upon further research into the accessibility of dermatologists, we were appalled to find that 42% of Americans lived in areas underserved by dermatologists, and many avoid going to the doctors due to the high costs and long wait times to schedule appointments (average of 56 days). Because of this lack of accessibility, many people are unable to obtain an affordable medical opinion on their skin conditions, which means they do not get treatment and their condition often worsens. We focused on four of the most commonly confused rash presenting diseases: lupus, lyme disease, psoriasis, and exanthematous drug-induced rashes. In creating Rashionale, we wanted to help bridge the gap between patient and medical experts by fast tracking the diagnosis process and ensuring that patients have access to the information that they need.