Abstract

Image-based classifification of histology sections plays an important role in predicting clinical outcomes. However this task is very challenging due to the presence of large technical variations (e.g., fifixation, staining) and biological heterogeneities (e.g., cell type, cell state). In the fifield of biomedical imaging, for the purposes of visualization and/or quantifification, different stains are typically used for different targets of interest (e.g., cellular/subcellular events), which generates multi-spectrum data (images) through various types of microscopes and, as a result, provides the possibility of learning biologicalcomponent-specifific features by exploiting multispectral information. We propose a multispectral feature learning model that automatically learns a set of convolution fifilter banks from separate spectra to effificiently discover the intrinsic tissue morphometric signatures, based on convolutional sparse coding (CSC). The learned feature representations are then aggregated through the spatial pyramid matching framework (SPM) and fifinally classifified using a linear SVM. The proposed system has been evaluated using two large-scale tumor cohorts, collected from The Cancer Genome Atlas (TCGA). Experimental results show that the proposed model 1) outperforms systems utilizing sparse coding for unsupervised feature learning (e.g., PSDSPM [5]); 2) is competitive with systems built upon features with biological prior knowledge (e.g., SMLSPM [4]).