Breakthroughs in bioimaging technology have made it possible to acquire vast amounts of biological image information. In addition to three-dimensional spatial information, biological images contain temporal and spectral information along multiple axes. Important challenges in biological image information processing include representing enormous amounts of multidimensional data in a unified manner, revealing relationships among multiple axes, and finding information that forms the core. The present study aims to achieve the systematization of biological image data analysis by multilinear sparse coding (MSC), in which four-dimensional biological image data are treated as a tensor. In MSC, multidimensional data are expressed as a linear combination of tensor bases (those that express individual data or the relationship between them) that have few characteristic patterns (specific meaning and effect). It is expected that the analysis of each base will lead to the identification of components that contribute to segmentation and classification.