Algorithms of blood cells recognition

The structure of the medical image analysis system is considered. A schematic diagram of the system for automated microscopic studies as well as the algorithm of recognition of blood cells are shown. The main task to be solved during the morphological analysis of blood is formulated. The requirements are specified for the algorithm used for determining leukocyte counts and detecting blood cells. A model of color and brightness characteristics to describe typical images of a blood smear is offered. The threshold values of the object size at searching the cells are specified. The luminance histogram of a typical field of view is studied. Two-step algorithm for detecting blood cells as well as the algorithm for constructing a separating line are described. The results of the experiments on real samples are given. The causes of detection errors are considered.

References

[1] Sokolinskiy B.Z., Dem’yanov V.L. , Medovyy V.S., Parpara A.A., Pyatnickiy A.M. Automated Sorting Leukocytes in a Blood Smear Using Techniques of Educable Neural Networks and Watershed. Zdravoohranenie i medicinskaya tehnika [Health Service and Medical Technology], 2005, no. 4 (in Russ.).

[2] Dolgov V.V., Menshikov V.V., eds. Nacionalnoe rukovodstvo po laboratornoy diagnostike [National Guidelines for Laboratory Diagnostics], vol. 1,2.

[3] Medovyy V.S., Parpara A.A., Pyatnickiy A.M., Sokolinskiy B.Z., Dem’yanov V.L. Review of Methods of Biomaterial Automated Microscopy. Klinicheskaya laboratornaya diagnostika [Clinical Laboratory Diagnostics], 2006, no. 7, pp. 15-20 (in Russ.).

[4] Medovyy V.S., Pyatnitskiy A.M. Robotic Microscopy and information technology to increase accuracy, sensitivity and availability of blood cell analyses. Current microscopy contributions to advances in science and technology (Microscopy Book Series, Publisher: Formatex Research Center), book 5, vol. 1, pp. 775-781, December 2012.

[5] Albertini M.C., Teodori L., Piatti E., Piacentini M.P., Accorsi A., Rocchi M.B.L. Automated analysis of morphometric parameters for accurate definition of erythrocyte cell shape Cytometry. Part A. 52A:12-18, 2003.

[6] Pratt W.K. Digital Image Processing. N.Y.-Chichester-Brisbane-Torc. John Wiley & Sons, 1978.

[7] Panteleev I., Egorova O., Klykova E. Komp’yuternaya mikroskopiya [Computer Microscopy]. Moscow, Tekhnosfera Publ., 2005.

[8] Bikhet S.F., Darwish A.M., Tolba H.A., Shaheen S.I. Segmentation and classification of white blood cells. Acoustics, Speech, and Signal Processing, 2000. ICASSP’00. Proceedings. 2000 IEEE International Conference. On vol. 6. 2000, vol. 4, pp. 22592261.

[9] Jiang K., Liao Q., Dai S. A novel white blood cell segmentation scheme using scale-space filtering and watershed clustering. Proc. of the Second Intern. Conf. on Machine Learning and Cybernetics. Xi’an, 2-5 November 2003.

[10] Anisimov B.V., Kurganov V.D., Zlobin V.K. Raspoznavanie i tsifrovaya obrabotka izobrazheniy [Image Recognition and Digital Processing]. Moscow, Vyssh. shk. Publ., 1983. 295 p.