In this project, we use sample of data from ISIC challenge 2017 in order to build a new dataset describing features of image lesions. More speciffically, we focus on Assymetry, border irregularity and color. Finally, we use these features to investigate the following two research questions:
- Given the extracted features, how well can we distinguish between healthy and melanoma lesion?
- Given the extracted features, how well can we distinguish between healthy and keratosis lesion?
See the results in report.
Open the main notebook and read the project setup section - make sure you have installed explicitly mentioned packages. Then you should be able to run the notebook without any problem. In case of any issues, do not hesitate to contact me at [email protected].