T.M. Magrupov1, R.R. Akhmadzhanov2, Z.M. Rakhimberganov3

1,3 Tashkent State Technical University n.a. Islam Karimov (Tashkent, Republic Uzbekistan)
2 "Scientific Center" Medical Equipment and Technology "(Tashkent, Republic of Uzbekistan)
1 talatmt@rambler.ru, 2 akhmadjonovr.98@gmail.com, 3 raximberganovaz@gmail.com

Lung diseases cause various problems for human health, and their early detection, diagnosis and treatment are very important in maintaining human working capacity in all countries of the world. Identifying the type of lung diseases by sound signals is fundamental in determining the patient's diagnosis.

The work purpose – Development of an algorithm for processing and analyzing lung sounds based on deep learning methods – convolutional neural networks to improve the accuracy of diagnosis of respiratory diseases.

A technique for processing and analyzing sounds of lung diseases for deep learning of neural networks is proposed. To optimize the processing of sound data, methods of extracting low-frequency cepstral coefficients, adding noise, stretching over time, and shifting the pitch of sounds were used to create a deep learning model.

The proposed technique for processing and analyzing lung sounds using deep learning models provides opportunities for medical professionals to obtain reliable results in the diagnosis of respiratory diseases with higher accuracy, contributing to early intervention and treatment of the patient.

Magrupov T.M., Akhmadzhanov R.R., Rakhimberganova Z.M. Methods of processing and analyzing data on sounds of pulmonary diseases based on deep learning of neural networks. Biomedicine Radioengineering. 2025. V. 28. № 5. P. 115–119. DOI: https:// doi.org/10.18127/j15604136-202505-23 (In Russian)

1. Arnold J.E., Figueroa W.Q. Enhancing Lung Sound Auscultation with Smart Stethoscopes. Journal of Clinical Engineering. 2021. V. 47.№ 3. P. 14–148.
2. Levitan R., Stephen J. The Importance of Lung Sound Analysis in Medical Diagnosis. Journal of Pulmonary Medicine. 2020. V. 23. № 5. P. 34–357.
3. Singh P., Kumar A. Challenges in Respiratory Sound Data Collection for Disease Classification. Respiratory Medicine and Research. 2021. V. 42. № 2. P. 103–110.
4. Magrupov T.M., Talatov Y.T., Magrupova М.Т. Method and algorithms for measuring the signs of biomedical indicators of medical diagnostics. 2021 Int. Conf. on Information Science and Communications Technologies (ICISCT). Tashkent. Uzbekistan. 2021. P. 214–219.
5. Abdumalikova F., Giyasova M., Usmanov H. Osobennosti techeniya serdechno-sosudistyh zabolevanij pri Covid-19. Vestnik Tashkentskoj medicinskoj akademii. 2021. S. 82–84 (In Russian).
6. Magrupov T.M., Nurillaeva N.M., Hidoyatova M.R., Abdumalikova F.B., Zubajdullaeva M.T. Metody analiza i obrabotki biomedicinskih signalov pri zabolevaniyah legkih. Vestnik molodyh uchenyh. 2024. № 24. S. 71–79 (In Russian).
7. Magrupov T.M., Nurillaeva N.M., Ahmadzhonov R.R., Zubajdullaev Sh.Sh., Gaibnazarov S.S., Semenova E.A. Algoritmicheskaya i programmnaya realizaciya tekhnologii klassifikacii biomedicinskih izobrazhenij legochnyh (biomedical engineering) Medicinskaya tekhnika. 2025. №2. S. 18–21 (In Russian).
8. Lee K.S. Neural Networks for Automated Respiratory Sound Analysis. Bio medical Signal Processing and Control. 2019. V. 51. P. 79–85.
9. López-Cuenca P., Vargas V.S., Rueda L. Deep Learning Techniques for Automated Classification of Respiratory Patterns. IEEE Transactions on Biomedical Engineering. 2019. V. 66. № 8. P. 2313–2321.