A.K. Alimuradov1, B.A. Porezanov2, A.Yu. Tychkov3, A.V. Klimashevich4, M.Yu. Yudin5, А.A. Mamonova6, M.Ya. Karibov7, K.M. Gadzhiev8

1,3,6,8 Penza State University (Penza, Russia)

2 JSC "CeSIS" (Penza, Russia)

4,5 Moscow Multidisciplinary Scientific and Clinical Center n. a. S.P. Botkin (Moscow, Russia)

7 Dagestan State Technical University (Makhachkala, Russia)

1 alansapfir@yandex.ru, 2 tychkov-a@mail.ru, 3 bogdan.porezanov@yandex.ru, 4 a.klimashevich@mknc.ru, 5 doc.judin@icloud.com, 6 mamonova.02@yandex.ru, 7 dstu-u831@mail.ru, 8 gadzhievkkk@gmail.com

Postoperative voice function disorders are a common complication of surgical interventions on the neck organs, however, their objective diagnosis at early stages is complicated due to the non-stationary nature of speech signals and the limited sensitivity of traditional spectral analysis methods. Existing acoustic approaches do not always allow reliable detection of subtle changes in prosodic and amplitude characteristics of the voice that arise due to postoperative edema and changes in the resonant properties of the vocal tract.

Aim – to investigate the diagnostic informativeness of prosodic and amplitude parameters of speech signals computed after empirical mode decomposition and formation of mode component combinations to improve the reliability of voice function changes assessment in patients before and after surgical interventions on the neck organs.

Based on the analysis of more than 200 speech signals from 30 patients, it was established that preliminary decomposition of the signal into empirical modes with subsequent combining of mode components allows extracting frequency ranges most sensitive to postoperative changes. The greatest sensitivity of the mean fundamental frequency was revealed in high-frequency combinations (2nd to 7th), while that of energy characteristics was found in mid- and low-frequency combinations (9th to 16th). The use of improved complete ensemble empirical mode decomposition with adaptive noise provided the smallest parameter variability in repeated measurements compared to classical and ensemble empirical mode decomposition.

The proposed approach allows improving the reliability of objective assessment of postoperative voice function disorders caused by complications after surgical interventions on the thyroid gland and other neck organs. The developed method can serve as a basis for creating automated systems for remote monitoring of voice recovery, which is especially relevant for early diagnosis and control of the effectiveness of rehabilitation measures in clinical practice.

Alimuradov A.K., Porezanov B.A., Tychkov A.Yu., Klimashevich A.V., Yudin M.Yu., Mamonova А.A., Karibov M.Ya., Gadzhiev K.M. A method for improving the reliability of prosodic and amplitude characteristics analysis of speech signals // Biomedical Radioelectronics. 2026. V. 29. № 6. P. 55−62. DOI: https://doi.org/10.18127/j15604136-202606-11

1. Šimić Prgomet I., Frkanec S., Gugić Radojković I., Prgomet D. Prospective Voice Assessment After Thyroidectomy Without Recurrent Laryngeal Nerve Injury. Diagnostics. 2024. V. 15. № 1. P. 37.
2. Singh K., Singh I., Meher R., Gopal A. et al. Incidence of Injury to External Branch of Superior Laryngeal Nerve as Diagnosed by Acoustic Voice Analysis After Thyroidectomy. Indian Journal of Otolaryngology and Head & Neck Surgery. 2025. V. 77. № 1. P. 1–9.
3. Song J.W., Kim J.H., Lee D.Y. Comparison of Voice Changes After Thyroidectomy Between Normal Adult and Elderly Populations. Journal of Korean Society of Laryngology, Phoniatrics and Logopedics. 2025. V. 36. № 3. P. 98–104.
4. Singh K., Singh I., Meher R., Gopal A. et al. Evaluation of Acoustic Voice Parameters After Anterior Cervical Discectomy and Fusion. Cureus. 2022. V. 14. № 1. P. e21489.
5. Zhou Y., Sun Y., Zhang J., Yan Y. Speech emotion recognition using both spectral and prosodic features. 2009 IEEE International Conference on Information and Automation. 2009. P. 625–629.
6. Lotter T., Vary P. Speech Enhancement by MAP Spectral Amplitude Estimation Using a Super-Gaussian Speech Model. EURASIP Journal on Applied Signal Processing. 2005. V. 2005. P. 1110–1126.
7. Alimuradov A.K., Tychkov A.Yu., Simakova O.S. Mamonova A.A., Yuldashev Z.M., Temirova D.A. Tekhnologiya segmentacii «signal/pauza» na osnove analiza urovnya smeshivaniya fragmentov rechevyh signalov. Biomedicinskaya radioelektronika. 2025. T. 28. № 2. S. 38–43 (In Russian).
8. Alimuradov A.K., Tychkov A.Yu., Yuskaev M.I., Dudnikov D.S., Tyurin M.A., Churakov P.P., Yuldashev Z.M. Sposob klassifikacii estestvennyh emocional'nyh sostoyanij cheloveka po rechi na osnove rekurrentnoj nejronnoj seti. Biomedicinskaya radiotekhnika. 2023. T. 26. № 4. S. 73–84 (In Russian).
9. Huang N.E., Shen Z., Long S.R., Wu M.C., Shih H.H., Zheng Q., Yen N.-C., Tung C.C., Liu H.H. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis. Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences. 1998. V. 454. № 1971. P. 903–995.
10. Wu Z., Huang N.E. Ensemble empirical mode decomposition: a noise-assisted data analysis method. Advances in Adaptive Data Analysis. 2009. V. 1. № 1. P. 1–41.
11. Torres M.E., Colominas M.A., Schlotthauer G., Flandrin P. A complete ensemble empirical mode decomposition with adaptive noise. 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE. 2011. P. 4144–4147.
12. Stepanova Yu.E. Akusticheskie metody issledovaniya golosa v sovremennoj foniatrii. Rossijskaya otorinolaringologiya. 2025. T. 24. № 3. S. 126–133 (In Russian).
13. Vashkevich M.I., Lihachyov D.S., Azarov I.S. Sistema analiza i klassifikacii golosovogo signala na osnove pertrubacionnyh parametrov i kepstral'nogo predstavleniya v psihoakusticheskih shkalah. Doklady BGUIR. 2022. T. 20. № 1. S. 73–82 (In Russian).
14. He L., Dellwo V. Amplitude Envelope Kinematics of Speech Signal: Parameter Extraction and Applications. Elektronische Sprachsignalverarbeitung. 2017. P. 107–113.
15. Wang L., Duangpummet S., Unoki M. Blind Estimation of Speech Transmission Index and Room Acoustic Parameters by Using Extended Model of Room Impulse Response Derived From Speech Signals. IEEE Access. 2023. V. 11. P. 49431–49450.
16. Do H.D. Exploiting Signal Linear Trend in the Time Domain to Enhance Speech Feature. IEEE Access. 2022. V. 10. P. 117344–117357.
17. Borovikova D.V., Grishin O.V., Nen'ko A.V., Yupashevskij A.V., Kaz'mina A.S., Markov A.V., Mecler K.A. Razrabotka apparatno-programmnogo kompleksa analiza i korrekcii rechi. Sistemy analiza i obrabotki dannyh. 2021. № 2 (82). S. 135–145 (In Russian).
18. Tseng C.-Y., Chang C.-H. Pause or No Pause? – Prosodic Phrase Boundaries Revisited. Tsinghua Science and Technology. 2008. V. 13. № 4. P. 500–509.