The study of texture features for the recognition problems of bone marrow cells in information-measuring systems of oncohematology

The influence of the parameters of the construction of the spatial-dependence matrices on textural features in the tasks of recognizing bone marrow cells in information and measurement systems for the diagnosis of acute leukemia is studied. Bone marrow preparations were obtained from patients with B- and T-cell acute lymphoblastic leukemias. 100 images of blast cells of B- and T-types were involved. Five textural features are considered – energy, inertia moment, local uniformity, maximum probability, entropy. The features were calculated on the basis of the spatial dependence matrices. The type of the color components of the RGB color image model, the adjacency distance and the direction of adjacency were analyzed as variable parameters when constructing the specified matrices. For a given sample of images of blast cells of type T and B a range of adjacency distances from 1 to 11 pixels was revealed, in which the greatest change in the values of texture features is observed. For different types of signs the change ranged from 20 % to 1700 %. The maximum information content among the studied texture features was obtained for the G-component of a color image in the texture feature “local uniformity” (information content coefficient 0.48) with an adjacency distance equal to one pixel. For practical application, it is recommended to use four directions of adjacency when constructing spatial adjacency matrices. The obtained results are important for specialists working in the field of designing information and measurement systems of oncohematology (diagnostics of dangerous oncological diseases – acute leukemias).

1. Gematologiya: rukovodstvo dlia vrachey, ed. N. N. Mamaev, St. Petersburg, SpetsLit. Publ., 2019, 639 p. (In Russ.)

2. Samorodov A. V., Biomedical Engineering, 2019, vol. 52, no. 6, pp. 387–390. https://doi.org/10.1007/s10527-019-09853-9

3. Sahlol A. T., Kollmannsberger P., Ewees A. A., Scientifi c Reports, 2020, vol. 10, no. 1, pp. 1–11. https://doi.org/10.1038/s41598-020-59215-9

4. Kutlu H., Avci E., Ozyurt F., Medical hypotheses, 2020, vol. 135, 109472. https://doi.org/10.1016/j.mehy.2019.109472

5. Chernykh E. M., Mihelev V. M., Kompjuternaya sistema klassifi katsii leykocitov na izobrazheniyah kletok krovi, Nauchniy rezultat. Informatsionnie tehnologii, 2019, vol. 4, no. 3, pp. 38-47. (In Russ.)

6. Amin M., Kermani S., Talebi A., Oghli M., Recognition of acute lymphoblastic leukemia cells in microscopic images using k-means clustering and support vector machine classifi er, Journal of Medical Signals & Sensors, 2015, vol. 5, no. 1, pp. 49–58.

7. Nikitaev V. G., Measurement Techniques,2015, vol. 58, no. 2, pp. 214–218. https://doi.org/10.1007/s11018-015-0688-0

8. Pratt W., Digital image processing: book 2, Hoboken, Wiley, 1982, 480 p.

9. Panteleev V. G., Slaev V. A., Chunovkina A. G., Measurement Techniques, 2008, vol. 51, no. 1, pp. 107–112. https://doi.org/10.1007/s11018-008-0019-9

10. Haralick R. M., Proceedings of the IEEE, 1979, vol. 67, no. 5, pp. 786–804. https://doi.org/10.1109/PROC.1979.11328

11. Dmitriyeva V. V., Tupitsyn N. N., Polyakov Ye. V., Samsonova A. D., Issledovaniye eff ektivnosti klassifi katsii izobrazheniy kletok kostnogo mozga v komp’yuternykh sistemakh diagnostiki ostrykh leykozov i minimal’noy ostatochnoy bolezni, Modelirovaniye, optimizatsiya i informatsionnyye tekhnologii, 2020, vol. 8, no. 3, pp. 1–9. (In Russ.)

12. Nikitaev V. G., Tupitsyn N. N., Pronichev A. N., Serebryakova I. N., Palladina A. D., Biomedical Engineering, 2021, vol. 54, no. 5, pp. 354–356. https://doi.org/10.1007/s10527-021-10038-6

13. Nikitaev V. G., Pronichev A. N., Polyakov E. V., Chernysheva O. A., Serebryakova I. N., Tupitsyn N. N., Procedia Computer Science, 2020, vol. 169, pp. 353–358. https://doi.org/10.1016/j.procs.2020.02.229

14.

15.