Reworking Multilabel Brain Tumor Segmentation: An Automated Framework Using Structured Kernel Sparse Representation

Advances in the field over the years have made medical imaging an indispensable part of medicine. Today, the use of medical images is often critical for diagnosis and treatment planning. The efficient processing and analysis of the large imaging data sets that have accompanied the rise in popularity of medical imaging have presented significant challenges that have yet to be successfully overcome. Medical image analysis and volume visualization are topics that attract much attention from the image-processing community. The former extracts useful knowledge for specific purposes, such as tumor segmentation, while the latter creates vivid two-dimensional representations of three-dimensional volumetric data. Our research projects have focused on the sparse and compressible properties of signals in some representation systems with the general objective of fully utilizing these properties for the development of effective and efficient algorithms for medical image analysis and volume visualization. In this article, we introduce one of our research projects, which aims to address the significant but challenging task of multilabel brain tumor segmentation. In this work, we proposed a superpixel-based framework for this specific task using structured kernel sparse representation.