Evaluation of the clinical applicability of artificial intelligence systems for detecting infiltrative changes on chest x-rays in inpatient care

Aim: to investigate the capabilities of artificial intelligence systems (AIS) in an inpatient setting, specifically in identifying signs of infiltration in chest radiographs (RG OGK).

Materials and methods. The initial data were chest X-rays (CXRs) in a frontal projection of 1,537 inpatients aged 18 to 102 years, obtained from the Unified Radiological Information System of the Unified Medical Information and Analytical Service (ERIS EMIAS). The presence of infiltration on the chest X-ray was determined by a radiologist (Radiologist 1) and an expert radiologist (Radiologist 2) with more than 5 years of experience. Radiologist 2 annotated the studies in two stages: at the first stage, only the chest X-rays were provided; at the second stage, in addition to the radiological data, access to information about the patient’s condition was provided.

Results. The accuracy rates for patients with a shifting state were 0.755 for AIS, 0.756 for Radiologist 1, and 0.856 for Radiologist 2, respectively. No statistically significant differences in accuracy were found between AIS and Radiologist 1 (p>0.05). When considering the scenario of using the AIS as a second opinion the accuracy for patients with a shifting state was 0.833, which is statistically significantly higher than the accuracy of Radiologist 1 and the AIS (p<0.05).

Conclusion. AIS can be used as a second opinion when detecting signs of infiltration based on RG OGK data in a hospital setting. However, since the diagnostic accuracy of the AIS is still insufficient for practical use, further research is needed to train algorithms using both visual data and electronic medical record data.

1. Gusev A.V., Artemova O.R., Vasiliev Yu.A., et al. Integration of AI-based software as a medical device into Russian healthcare system: results of 2023. National Health Care (Russia). 2024; 5: 17-24. doi: 10.47093/2713-069X.2024.5.2.17-24.

2. Vasilev Y.A., Sychev D.A., Bazhin A.V., et al. Autonomous artificial intelligence for sorting results of preventive radiological examinations of chest organs: medical and economic efficiency. Digital Diagnostics. 2025; 6: 5-22. doi: 10.17816/DD641703.

3. Hwang E.J., Park J., Hong W., et al. Artificial intelligence system for identification of false-negative interpretations in chest radiographs. European Radiology. 2022; 32: 4468-4478. doi: 10.1007/s00330-022-08593-x.

4. De Groot P.M., Carter B.W., Abbott G.F., et al. Pitfalls in Chest Radiographic Interpretation: Blind Spots. Seminars in Roentgenology. 2015; 50: 197-209. doi: 10.1053/j.ro.2015.01.008.

5. Pilyus P.S., Drokin I.S., Bazhenova D.A., et al. Otsenka perspektiv ispol’zovaniya tekhnologiy iskusstvennogo intellekta dlya analiza KT-izobrazheniy organov grudnoy kletki s tsel’yu vyyavleniya priznakov zlokachestvennykh novoobrazovaniy v legkikh. Meditsinskaya vizualizatsiya. 2023; 27(2): 138-146. (In Russ.)]. doi: 10.24835/1607-0763-2023-2-138-146.

6. Hofmeister J., Garin N., Montet X., et al. Validating the accuracy of deep learning for the diagnosis of pneumonia on chest x-ray against a robust multimodal reference diagnosis: a post hoc analysis of two prospective studies. European Radiology Experimental. 2024; 8: 20. doi: 10.1186/s41747-023-00416-y.

7. Servisy iskusstvennogo intellekta v luchevoi diagnostike. Available at: https://mosmed.ai/. Accessed: 05.09.2025. (In Russ.)

8. Vasil’ev Yu.A., Vladzimyrskiy A.V., editors. Iskusstvennyy intellekt v luchevoy diagnostike: Per Aspera Ad Astra. Moscow: Izdatel’skie resheniya; 2025: 491. (In Russ.)

9. Morozov S.P., Abuladze L.R., Andreychenko A.E., et al. Bazovye rekomendatsii k rabote servisov iskusstvennogo intellekta dlya luchevoy diagnostiki: metodicheskiye rekomendatsii. Seriya «Luchshiye praktiki luchevoy i instrumental’noy diagnostiki». Vyp. 119. Moscow: GBUZ «NPKTs DiT DZM»; 2022: 68. (In Russ.)]

10. Schoonjans F. MedCalc’s Sample size for Area under ROC curve. Available at: https://www.medcalc.org/calc/sample-size-area-under-roc-curve.php. Accessed April 05.09.2025.

11. Andropova P.L., Gavrilov P.V., Kazantseva I.P., et al. Interexpert Agreement Between Emergency Neuroradiologists with Different Levels of Experience in the Rating of ASPECTS. Radiology — practice. 2022; 5: 10-25. doi: 10.52560/2713-0118-2022-5-10-25.

12. Wetcher-Hendricks D. An Expanded McNemar Approach for Evaluation of Longitudinal Data 2025. Preprints 2025. doi: 10.20944/preprints202503.2111.v1.