Description of Invention
This abstract presents the development of a real-time system for the segmentation and classification of IHC breast cancer images. The system utilizes a custom StarDist algorithm integrated with a hybrid deep-learning model known as Resnet50_U-Net. A user-friendly graphical user interface (GUI) has been designed to display segmented images alongside real-time camera images. The system also calculates a score for the classification and presents it within the GUI. Comparative analysis with the built-in StarDist method demonstrates superior performance in terms of image segmentation and classification.