Computer Science Department
Scancer
Nate Wilhelm
Soaking up the sun can be very relaxing; however, skin cancer is a great risk of sun exposure. Scancer, an IOS application, will allow users to monitor their sunspots. Scancer will utilize machine learning to determine if a mole or freckle is potentially cancerous using the HAM10000 dataset, a dataset of 10,000 images of skin lesions. Using this dataset, the model will be able to predict if the skin lesion is one of 7 types of skin lesion.
Upon opening the app, users will be directed to a home screen which will have 4 options: take/upload a photo, view profile, a list of dermatologists, and a map view of dermatologists. By selecting one of these options the user will then be redirected to the corresponding page. By clicking the camera button, the user will be taken to a page to classify an image. On this page users can either upload or take a picture to be classified. Upon classification, the image and its label will appear on the screen. Additionally, this image and label will be added to the profiles page for later viewing and continual monitoring. By selecting the list view of dermatologists, users will see a list of the 20 closest dermatologists. In this list users will find a link to their yelp page, whether they are open, a rating out of 5, and the name of the office. Alternatively, users may use the map view to see on a map where the nearest dermatologists are.