Automated Liver Tumor Segmentation and Diagnosis Using Self-Guided Deep Clustering Networks

The proposed scheme makes use of the capabilities of deep learning in order to be able to deal with the task of automatic processing and subsequent analyzing of the medical images and particularly MRI scans. The SGDCN architecture integrates deep clustering algorithms with selfguided learning strategies to boost tumor detection capability and its robustness even when both minority and majority classes coexist. Altering tumor characteristics and liver anatomy massively and immediately, and with that, the regions of interest (ROIs) that are the cancerous parts are automatically isolated through the use of the method, the so-called, “dynamic opera.” While the potential that false positive reduces is of the highest category and the precision of the diagnosis is notably increased.