MICSurv: Medical Image Clustering for Survival risk group identification

Medical image processing is an exceptional methodology for cancer diagnosis as well as for the guidance of medical interventions such as surgical planning. Some studies have introduced the survival risk prediction using medical images, however, the number of research papers that address the problem of identifying groups of subjects that have similar survival probability distributions utilizing medical images is very limited. In this study, we demonstrate a simple yet powerful approach that can be used in a set of biomedical images dataset along with survival annotations in order to identify various risk groups with regards to the survival of the subjects.