Objective: To justify the inclusion of input parameters in a neural network designed to predict rehabilitation measures for patients with pulmonary tuberculosis who have lost their permanent work capacity. Material and methods. The study included 335 patients with pulmonary tuberculosis, 212 with a newly diagnosed case and 123 with a chronic form. To predict the effectiveness of rehabilitation measures, a neural network was developed in the Python 3.14 programming language. Its architecture incorporates a multilayer perceptron with direct connections between neurons and backpropagation algorithms. In the context of this study, it was advisable to use a neural network for predicting permanent disability in patients with pulmonary tuberculosis. Results. The following were integrated as input parameters of the neural network: genotype DD of the GSTM1 gene, genotype EE of the GSTT1 gene, genotype ТС of the АВСВ1 gene and polymorphic variants of cytokine genes, including tumor necrosis factor alpha (–308G>A (rs1800629)), interleukin (IL) 1-beta (–31C>T (rs1143627)), IL-4 (–589C>T) (rs2243250)), IL-10 (–592C>A (rs1800872)), and IL-10 (–1082A>G (rs1800896)). Additionally, an adverse recurrent course, multiple and extensive drug resistance, and persistent respiratory and cardiopulmonary failure following a course of rehabilitation were incorporated. Regression analysis showed the statistical significance of all input parameters with respect to the modeling of the dependent variable. As a result of testing the neural network, the prediction accuracy (rehabilitation efficiency) was 88,3%. Conclusion. The use of a neural network provides highly accurate predictions regarding the effectiveness of rehabilitation of patients with pulmonary tuberculosis. This method can be incorporated into clinical pulmonology practice to evaluate persistent disability in such patients. The presented predictive model is expected to form the basis of a comprehensive rehabilitation program that addresses medical, professional and social aspects.ural networks

Objective: To justify the inclusion of input parameters in a neural network designed to predict rehabilitation measures for patients with pulmonary tuberculosis who have lost their permanent work capacity.

Material and methods. The study included 335 patients with pulmonary tuberculosis, 212 with a newly diagnosed case and 123 with a chronic form. To predict the effectiveness of rehabilitation measures, a neural network was developed in the Python 3.14 programming language. Its architecture incorporates a multilayer perceptron with direct connections between neurons and backpropagation algorithms. In the context of this study, it was advisable to use a neural network for predicting permanent disability in patients with pulmonary tuberculosis.

Results. The following were integrated as input parameters of the neural network: genotype DD of the GSTM1 gene, genotype EE of the GSTT1 gene, genotype ТС of the АВСВ1 gene and polymorphic variants of cytokine genes, including tumor necrosis factor alpha (–308G>A (rs1800629)), interleukin (IL) 1-beta (–31C>T (rs1143627)), IL-4 (–589C>T) (rs2243250)), IL-10 (–592C>A (rs1800872)), and IL-10 (–1082A>G (rs1800896)). Additionally, an adverse recurrent course, multiple and extensive drug resistance, and persistent respiratory and cardiopulmonary failure following a course of rehabilitation were incorporated. Regression analysis showed the statistical significance of all input parameters with respect to the modeling of the dependent variable. As a result of testing the neural network, the prediction accuracy (rehabilitation efficiency) was 88,3%.

Conclusion. The use of a neural network provides highly accurate predictions regarding the effectiveness of rehabilitation of patients with pulmonary tuberculosis. This method can be incorporated into clinical pulmonology practice to evaluate persistent disability in such patients. The presented predictive model is expected to form the basis of a comprehensive rehabilitation program that addresses medical, professional and social aspects.

1. Grace A., Mittal A., Jain S. et al. Shortened treatment regimens versus the standard regimen for drug-sensitive pulmonary tuberculosis. Cochrane Database Syst Rev. 2019;12 (12): CD012918. https://doi. org/10.1002/14651858.CD012918.pub2.

2. Dodd P., Yuen C., Jayasooriya S., et al. Quantifying the global number of tuberculosis survivors: a modelling study. Lancet Infect Dis. 2021; 21 (7): 984–92. https://doi.org/10.1016/S1473-3099(20)30919-1.

3. Salhi В., Troosters T., Behaegel M., et al. Effects of pulmonary rehabilitation in patients with restrictive lung diseases. Chest. 2010; 137 (2): 273–9. https://doi.org/10.1378/chest.09-0241.

4. Troosters Т., Janssens W., Demeyer H., Rabinovich R.A. Pulmonary rehabilitation and physical interventions. Eur Respir Rev. 2023; 32 (168): 220222. https://doi.org/10.1183/16000617.0222-2022.

5. Guseva V.A., Rubleva N.V., Kolomiets V.M. Factors of effectiveness of rehabilitation of tuberculosis patients. In: Topical issues in the fight against tuberculosis: materials of the jubilee session dedicated to the 90th anniversary of TsNIIT RAMS. 2011: 59–60 (in Russ.).

6. Zhuk N.A. Principles of treatment and rehabilitation of patients with active tuberculosis. Problems of Tuberculosis and Lung Disease. 2005; 8: 26–29 (in Russ.).

7. Podgaeva V.A., Golubev D.N. The impact of living standards on indicators reflecting the epidemic situation of tuberculosis in the Urals. Tuberculosis and Lung Diseases. 2011; 8: 8–10. (in Russ.).

8. Gerke A. Morbidity and mortality in sarcoidosis. Curr Opin Pulm Med. 2014; 20 (5): 472–8.

9. Alymenko M.A., Valiev R.Sh., Valiev N.R., et al. Association of genotypes of biotransformation enzymes with susceptibility to chronic pulmonary tuberculosis. Consilium Medicum. 2024; 26 (3): 193–8 (in Russ.). https://doi.org/10.26442/20751753.2024.3.202659.

10. Rembovsky V.R., Mogilenkova L.A. Detoxification processes when the body is exposed to chemicals. Saint Petersburg: Polytechnic University Publishing House; 2017: 383 pp. (in Russ.).

11. Shcherbo S.N., Shcherbo D.S., Sokolova N.A., et al. Genetic predisposition and resistance to certain infectious diseases. IV. Tuberculosis. Medical Alphabet. 2022: 1 (6): 7–10 (in Russ.). https://doi.org/10.33667/2078-5631-2022-6-7-10 (in Russ.).

12. An A.R., Rudko A.A., Bragina E.Yu., et al. Study of the association of polymorphic variants of cytokine signaling genes with pulmonary tuberculosis. Tuberculosis and Lung Diseases. 2013; 90 (8): 34–9 (in Russ.).

13. Demyanov A.V., Kotov A.Yu., Simbirtsev A.S. Diagnostic value of the study of cytokine levels in clinical practice. Cytokines and Inflammation. 2003; 2 (3): 20–35 (in Russ.).

14. Simbirtsev A.S. Cytokines as a new system, regulating body defence reactions. Cytokines and Inflammation. 2002; 1 (1): 9–16 (in Russ.).

15. Baggiotini M., Dewald B., Moser B. Human chemokines: an update. Annu Rev Immunol. 1997; 15: 675–705.

16. Mordyk A.V., Tsygankova E.A., Puzyreva L.V., Turitsa A.A. Antitubercular immunity and mechanisms of its formation (literature review). Far Eastern Medical Journal. 2014; 1: 126–30 (in Russ.).

17. Haykin S.S. Neural networks: a comprehensive foundation. Macmillan; 1994: 696 pp. (in Russ.)

18. Kriegeskorte N., Golan T. Neural network models and deep learning. Curr Biol. 2019; 29 (7): 231–6. https://doi.org/10.1016/j.cub.2019.02.034.

19. Pérez J., Cabrera J.A., Castillo J.J., Velasco J.M. Bioinspired spiking neural network for nonlinear systems control. Neural Netw. 2018; 104: 15–25. https://doi.org/10.1016/j.neunet.2018.04.002.

20. Faustova K.I. Neural networks: application today and development prospects. Territory of Science. 2017; 4: 83–7 (in Russ.).

21. Hasanov A.G., Shaybakov D.G., Zhernakov S.V., et al. Neural networks in forecasting disease dynamics. Creative Surgery and Oncology. 2020; 10 (3): 198–204 (in Russ.). https://doi.org/10.24060/2076-3093-2020-10-3-198-204.

22. Streltsov V.A., Baranova V.G., Stolbun Yu.V. The need to assess the psychological status of patients with pulmonary tuberculosis. Tuberculosis and Lung Diseases. 2011; 5: 176–77 (in Russ.).

23. Adams D. The biology of the granulomas. Pathology of granulomas. New York. 1983: 1–19.

24. Kantemirova B.I., Sychev D.A., Karpishchenko V.N. Study of glutathione as a marker of the second phase of xenobiotic biotransformation in children with various somatic pathologies against the background of the treatment. Biomedicine. 2013; 1 (2): 103–7 (in Russ.).

25. Kaur M., Shahil Rahman T.K., Dolma S., et al. Xenobiotic metabolizing gene variants and the risk of male infertility – a systematic review, meta-analysis and in silico analysis. Toxicol Rep. 2025; 14: 102019. https://doi.org/10.1016/j.toxrep.2025.102019.

26. Bleasby K., Castle J.C., Roberts C.J., et al. Expression profiles of 50 xenobiotic transporter genes in humans and pre-clinical species: a resource for investigations into drug disposition. Xenobiotica. 2006; 36 (10–11): 963–88. https://doi.org/10.1080/00498250600861751.

27. Maurice C.F., Haiser H.J., Turnbaugh P.J. Xenobiotics shape the physiology and gene expression of the active human gut microbiome. Cell. 2013; 152 (1–2): 39–50. https://doi.org/10.1016/j.cell.2012.10.052.

28. Rendina M., Turnbaugh P.J., Bradley P.H. Human xenobiotic metabolism proteins have full-length and split homologs in the gut microbiome. G3. 2025; 15 (9): jkaf131. https://doi.org/10.1093/g3journal/jkaf131.

29. Lakin K.M., Krylov Yu.F. Biotransformation of medicinal substances. Moscow: Meditsina; 1981: 344 pp. (in Russ.)

30. Singh V., Dziwornu G.A., Chibale K. The implication of Mycobacterium tuberculosis-mediated metabolism of targeted xenobiotics. Nat Rev Chem. 2023; 7 (5): 340–54. https://doi.org/10.1038/s41570-023-00472-3.

31. Liu Y., Li H., Dai D., et al. Gene regulatory mechanism of mycobacterium tuberculosis during dormancy. Curr Issues Mol Biol. 2024; 46 (6): 5825–44. https://doi.org/10.3390/cimb46060348.

32. Mehta V.K., Deb P.S., Rao D.S. Application of computer techniques in medicine. Med J Armed Forces India. 2017; 50 (3): 215–8. https://doi.org/10.1016/S0377-1237(17)31065-1.

33. Thacharodi A., Singh P., Meenatchi R., et al. Revolutionizing healthcare and medicine: the impact of modern technologies for a healthier future – a comprehensive review. Health Care Sci. 2024; 3 (5): 329–49. https://doi.org/10.1002/hcs2.115.

34. Junaid S.B., Imam A.A., Balogun A.O., et al. Recent advancements in emerging technologies for healthcare management systems: a survey. Healthcare. 2022; 10 (10): 1940. https://doi.org/10.3390/healthcare10101940.

35. Penny W., Frost D. Neural networks in clinical medicine. Med Decis Making. 1996; 16 (4): 386–98. https://doi.org/10.1177/0272989X9601600409.

36. Toma A., Diller G.P., Lawler P.R. Deep learning in medicine. JACC Adv. 2022; 1 (1): 100017. https://doi.org/10.1016/j.jacadv.2022.100017.

37. Miotto R., Wang F., Wang S., et al. Deep learning for healthcare: review, opportunities and challenges. Brief Bioinform. 2017; 19 (6): 1236–46. https://doi.org/10.1093/bib/bbx044.

38. Arntz A., Weber F., Handgraaf M., et al. Technologies in home-based digital rehabilitation: scoping review. JMIR Rehabil Assist Technol. 2023; 10: e43615. https://doi.org/10.2196/43615.

39. Celian C., Redd H., Smaller K., et al. Use of technology in the rehabilitation setting: therapy observations, mixed methods analysis, and data visualization. Arch Rehabil Res Clin Transl. 2025; 7 (1): 100425. https://doi.org/10.1016/j.arrct.2024.100425.