Sparsity techniques in medical imaging

With the advent of the age for big data and complex strucure, sparsity has been an important modeling tool in compressed ensing, machine learning, image processing, neuroscience and tatistics. In the medical imaging field, sparsity methods have been uccessfully used in image reconstruction, image enhancement, mage segmentation, anomaly detection, disease classification, and mage database retrieval. Developing more powerful sparsity modls for a large range of medical imaging and medical image analysis roblems as well as efficient optimization and learning algorithm ill keep being a main research topic in this field. The goal of this pecial issue is to publish original and high quality papers on innoation research and development in medical imaging and medical mage analysis using sparsity techniques. This special issue will elp advance the scientific research within the field of sparsity ethods for medical imaging. This special issue is composed of nine high quality research artiles that are selected from over 20 submissions based on rigorous eer reviews. These papers cover a broad spectrum of research topcs in medical imaging and medical image analysis, including image egistration, segmentation, reconstruction, estimation, modeling, lassification and visualization. Specifically, Belilovsky et al. in collaborative research among entrale Supelec, Inria Saclay, Research Center Athena, Stony Brook niversity, Mount Sinai and MIT have proposed the k-support norm o predictively model fMRI data for both classification and regresion tasks. Chen and Srinivas from Ventana Medical System have eveloped a stain unmixing algorithm for brightfield multiplex mmunohistochemistry (IHC) images using a group sparsity model. u et al in a collaborative efforts among Nanjing University of nformation Science and Technology, Rutgers Cancer Institute of ew Jersey, University of Pennsylvania and Case Western Reserve niversity have developed an unsupervised sparse non-negative atrix factorization (SNMR) approach for color unmixing in digial pathology image analysis. Deshpande, Maurel and Barillot from niversity of Rennes, INSERM, CNRS and Inria have developed n adaptive dictionary learning method to automatically classify ultiple sclerosis (MS) lesions in MR images. Zheng et al. in colaborative research between University of Pennsylvania, Shandong