Classify a skin lesion into malignant or benign.
Melanoma cancer is a type of skin cancer which is not very common but can be fatal, if not detected in early stages. From all skin cancers, melanoma represents just 1% of cases, but 75% of deaths. However, early detection of Melanoma cancer is very critical, as the estimated 5-year survival rate for melanoma drops from over 99% if detected in its earliest stages to about 14% if detected in its latest stages. In this project we tried to build a classifier which, given a skin lesion image, could classify whether it is malignant or benign.
We used the dataset provided by the ISIC — International Skin Imaging Collaboration. To download the dataset
1. Upload the .pkl dependencies to your Google drive.
2. Open the downloadDataset.ipynb in Google Colab.
3. Mount your drive.
4. Run the cells.
Since the dataset is huge, it will take some time to download even using Colab's GPU, so be patient.
Trained the dataset using Inception-V3 model then applied trasfer learning on it.
usingInception.ipynb