The new results of the research in the field of automation of hematopoietic tumor diagnostics by analysis of the images of cytological specimens are presented. The main result is a new information technology for the morphological analysis of the lymphoid cell nuclei of patients with hematopoietic tumors based on the combined use of pattern recognition and image analysis techniques. The principal characteristic of the proposed technology is that the features used for description of lymphocyte nuclei are chosen and calculated from the images of specimens by image processing and analysis methods, and also by methods of mathematical morphology and Fourier analysis. The proposed technology provides transition from the diagnostic analysis of lymphocyte nuclei to diagnosing the patients with hematopoietic tumors by means of pattern recognition techniques. Experimental check of the technology shows that it can be successfully used in program system for automated diagnostics of the hematopoietic tumors.