seraogianluca / breast-cancer-image-classification

Breast abnormalities classification and diagnosis using TensorFlow developed for Computational Intelligence and Deep Learning course of the MSc AIDE at the University of Pisa.

Breast Cancer Classification

The project aims to develop different models to solve binary classification problems on the CBIS DDSM: Curated Breast Imaging Subset of Digital Database for Screening Mammography dataset. In particular, we will perform two binary classification tasks, breast abnormalities classification and breast abnormalities diagnosis, using different CNNs. In particular, we will perform the tasks with CNNs trained from scratch, with pre-trained CNNs, with a siamese network and using an ensemble technique.

Project files

Abnormality classification

• Scratch CNN
• Pretrained CNN - VGG
• Pretrained CNN - INC
• Baseline CNN
• Ensemble

Abnormality diagnosis

• Scratch CNN
• Pretrained CNN - VGG
• Pretrained CNN - INC
• Ensemble

Credits

@seraogianluca, @ragnar1002