Classification of multicolor fluorescence in-situ hybridization (M-FISH) image using structure based sparse representation model with different constrains

In this paper we propose a structure based sparse model with different constrains by extending the general sparse model to the multiple pixels case, where each pixel together with its neighboring pixels are used simultaneously in the sparse representation of chromosome classes. We use the model to classify multicolor fluorescence in-situ hybridization (MFISH) images. Both the simulation and real data analysis results show that the structure based sparse model penalized with l_p norm (p=0 and p=1) improved the accuracy of classification over the conventional sparse model based classifier, which translates into improved diagnosis of genetic diseases and cancers.