top of page


Medical Image Classification

Tristan Freiberg, Bailey Forster, Henri Antikainen


Melanoma can affect anyone and early detection is a crucial factor affecting survival rates. Machine learning models could assist trained healthcare professionals in screening for skin cancer. The Human Against Machine 10000 (HAM10000) dataset contains images of 7,470 distinct skin lesions, each belonging to one of seven mutually-exclusive classes of skin lesion, including melanoma, as well as other cancerous types and benign types such as nevi (moles). Our goal was to train a convolutional neural network (CNN) to accurately classify images of skin lesions using the HAM10000 dataset. The project includes a Streamlit app to test users’ classification ability against our fine-tuned models.

Screen Shot 2022-06-03 at 11.31.35 AM.png
github URL
bottom of page