An algorithm based on a deep probabilistic architecture referred to as tree-structured sum-product network (t-SPN) is considered for cells classification. The t-SPN is a rooted acyclic graph constructed as a tree of several sum-product networks where each network is constructed over a subset of most confusing class features. The constructed t-SPN architecture is learned by maximizing the margin which is defined to be the difference in the conditional probability between the true and the most competitive false labels. To enhance generalization, l(2)-regularization (REG) is considered along with the maximum margin (MM) criterion in the learning process. To highlight cell features, this paper investigates the effectiveness of two generic high-pass filters: ideal high-pass filtering and the Laplacian of Gaussian (LOG) filtering. On both HEp-2 and Feulgen benchmark datasets, the t-SPN architecture learned based on the max-margin criterion with regularization produced the highest accuracy rate compared to other state-of-the-art algorithms that include convolutional neural network (CNN) based algorithms. Ideal high-pass filter was more effective on the HEp-2 dataset which is based on immunofluorescence staining while the LOG was more effective on Feulgen dataset which is based on Feulgen staining.