carlotta-marchis/breast-cancer-detection

The scope of my project is to develop an automated diagnosis support system for the detection of the exact regions where Invasive Ductal Carcinoma (IDC) is present inside a whole mount slide.  IDC is the most common type of all breast cancers. Breast cancer diagnosis usually consists of several steps, including palpation, mammography or ultrasound imaging and breast tissue biopsy. In the particular case of IDC a very important step consists in grading its aggressiveness. To do so, pathologists usually focus on the regions of the mount sample where IDC is present. The automation of the detection of the exact regions of IDC inside a whole mount slide could help reduce costs and time of the test. As a first step in the development of this project I will train a simple CNN model to classify breast histopathology images.