Machine Learning Methods for Predicting the Course of the Disease Using the Example of the Development of Severe Pneumonia

The aim of the work is to predict complications in COVID-19 in the form of severe pneumonia based on clinical and laboratory data using machine learning (ML) methods. The severity of COVID-19 disease is based on the results of computed tomography (CT). The patient groups consisted of 31 patients with severe pneumonia (CT 2–4) and 113 patients with mild form (CT 0–1) and without pneumonia. The database included 105 clinical and laboratory parameters. The standard nonparametric criteria χ² and the Mann-Whittney criterion (U-test) with correction for multiple Bonferroni- Holm testing were applied. 13 significant indicators have been identified. Machine learning (ML) methods of the data analysis system («Data Master Azforus») were used and the best of them were applied in the form of an ensemble. ML methods have made it possible to build multifactorial nonlinear models for forecasting. For the entire follow-up period, the prediction result by the method of statistically weighted syndromes (SWS) reached a value of ROC AUC = 0.9. It was possible to make a fairly accurate prediction of severe pneumonia in COVID-19 based on the 26 most significant clinical and laboratory indicators. The clinical signs known to the attending physicians that determine the severity of pneumonia have been confirmed by ML methods. The approbation of the model proved its promise. The introduction of the model into practice will increase the accuracy and efficiency of diagnosis of severe pneumonia. The data analysis system («Data Master Azforus») will allow research doctors to create recommendation systems for predicting and diagnosing diseases

1. Bilichenko T. N. Risk factors, immunological mechanisms and biological markers of severe COVID‑19 course (study overview). Russian Medical Inquiry. Medical review. 2021;5(5):237–244. (In Russ.). DOI: 10.32364/2587-6821-2021-5-5-237-244

2. Kudryavtsev Y. S., Beregov M. M., Berdalin A. B., Lelyuk V. G. Comparison of the Main Staging Systems for Assessing the Severity of Lung Injury in Patients with COVID‑19 and Evaluation of Their Predictive Value. Journal of radiology and nuclear medicine. 2021;102(5):296–303. (In Russ.). DOI: 10.20862/0042-4676-2021-102-5-296-303

3. Inoue A., Takahashi H., Ibe T., Ishii H., Kurata Y., Ishizuka Y. et al Comparison of semiquantitative chest CT scoring systems to estimate severity in coronavirus disease 2019 (COVID‑19) pneumonia. Eur Radiol. 2022;32(5):3513–3524. DOI: 10.1007/s00330–021–08435–2

4. Hemraj S. K., Jacob M. J., Kotian V., K. S.D., G. G.R., Veliath L. B. Chest CT Findings and Their Temporal Evolution in COVID‑19 Pneumonia. Cureus. 2022;14(6):e26021. DOI: 10.7759/cureus.26021

5. Nokiani A. A., Shahnazari R., Abbasi M. A., Divsalar F., Bayazidi M., Sadatnaseri A. COVID‑19 Chest CT Quantification: Triage and Prognostic Value in Different Ages. Clin Med Res. 2023Mar;21(1):14–25. DOI: 10.3121/cmr.2023.1772

6. Szabó M., Kardos Z., Kostyál L., Tamáska P., Oláh C., Csánky E. et al. The importance of chest CT severity score and lung CT patterns in risk assessment in COVID‑19-associated pneumonia: a comparative study. Front Med (Lausanne). 2023;10:1125530. DOI: 10.3389/fmed.2023.1125530

7. Morozov S. P., Gombolevskiy V. A., Cherninа V. Yu., Blokhin I. A., Mokienko O. A., Vlаdzimirskiy A. V. et al. Prediction of lethal outcomes in COVID‑19 cases based on the results chest computed tomography. Tuberculosis and Lung Diseases. 2020;98(6):7–14. (In Russ.). DOI: 10.21292/2075-1230-2020-98-6-7-14

8. Melekhov A. V., Sayfullin M. A., Petrovichev V. S., Agaeva A. I., Golubykh K. Yu., Nikitin I. G. Association of Computer Tomography Features Of COVID‑19 with Outcomes, Clinical and Laboratory Parameters. The Russian Archives of Internal Medicine. 2021;11(6):447–456. (In Russ.). DOI: 10.20514/2226-6704-2021-11-6-447-456

9. Komurcuoglu B., Susam S., Batum Ö., Turk M. A., Salik B., Karadeniz G. et al. Correlation between chest CT severity scores and clinical and biochemical parameters of COVID‑19 pneumonia. Clin Respir J. 2022;16(7):497–503. DOI: 10.1111/crj.13515

10. Wang D., Willis D. R., Yih Y. The pneumonia severity index: Assessment and comparison to popular machine learning classifiers. Int J Med Inform. 2022;163:104778. DOI: 10.1016/j.ijmedinf.2022.104778

11. Holm S. A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics. 1979;6(2):65–70.

12. Senko O. V., Kuznetsova A. V., Matveev I. A., Litvinchev I. S. Chapter VIII. Search of regularities in data: optimality, validity, and interpretability. Advances of Artificial Intelligence in a Green Energy Environment. Edited by Pandian Vasant, Joshua Thomas, Elias Munapo, Gerhard-Wilhelm Weber. Elsevier Inc. 2022;151–171.

13. Senko O. V., Kuznetsova A. V., Voronin E. M., Kravtsova O. A., Borisova L. R., Kirilyuk I. L., Akimkin V. G. Methods of intellectual data analysis in COVID‑19 research. Journal of the Belarusian State University. Mathematics and Informatics. 2022;(1):83–96. (In Russ.). DOI: 10.33581/2520-6508-2022-1-83-96