MIDS Capstone Project Fall 2022

With your photo, can classify up to 5 categories of skin conditions encompassing ~120 different skin abnormalities - ranging from the most malignant skin cancers to the most common skin diseases

According to the National Institute of Health, skin conditions are one of the most prevalent ailments in America, affecting one in three people at any given time. However, the varying degrees of severity leave many prospective patients wondering if they require physician consultation, or if they would be able to carry on without a costly trip to the doctor’s office.

The goal of is to create a mobile app that could classify different user-submitted skin images to their nearest category of diagnosis, using an ensemble model, in which we combine prediction from a multitude of deep learning abnormality detection models. Users should be able to use the mobile app to diagnose and reduce the amount of money and time spent at doctors offices, and detect possible life-threatening markings that should motivate a trip to their physician. 


Video app demo app demo

If you require video captions for accessibility and this video does not have captions, click here to request video captioning.

Last updated:

December 6, 2022