Publication:
Detection, recognition and transmission of snoring signals by ESP32

Thumbnail Image
Files

Files

Detection, recognition and transmission of snoring signals by ESP32.pdf (4.4 MB)
View Statistics
Reference Managers
Mendeley
Indexers
Google Scholar CORE
QR Code
QR Code
Authors

Authors

Morales Mahecha, Esteban
Melo Camargo, Adriana
Sánchez Jiménez, Edison
Coy Sarmiento, Diego Arturo
Hernández Salazar, Sara Valentina
Paz Penagos, Hernán
Abstract (English)
Obstructive sleep apnea (OSA) is a respiratory disorder that affects a significant percentage of the world’s population, with an estimated prevalence between 9 % and 17 % in adults [1]. Despite its high inci dence, it is estimated that approximately 85 % of patients remain un diagnosed [2]. In Colombia, between 2017 and 2021, 363,204 cases of OSA were diagnosed [3]. Several studies point out that this condition may increase the risk of developing hypertension, cardiovascular dis ease and stroke [4], in addition to causing greater cognitive impairment (mainly in executive function, attention and memory). The most common medical test for the diagnosis of sleep disorders is to perform polysomnography, in which the sleeping patient is observed to determine physiological variables such as oxygen saturation level, amount of oral and nasal airflow, degree of respiratory effort, electro cardiographic measurements, position and general body movement [5]. However, it is a costly, invasive study, which is not available in all health centers and takes a long time to perform [6], so it is relevant to look for alternative ways of diagnosis and follow-up. Snoring is an early sign of this condition and a typical symptom of OSA patients [7], and is caused by vibration of the upper airway soft tissues affecting anatomical structures such as the soft palate, uvula and pharynx [8]. This is why detection and analysis of snoring patterns are essential to identify and follow up sleep disorders such as OSA, which can improve the quality of life of patients by allowing early and accurate diagnosis [9]. For example, a novel method for the detection of obstructive sleep apnea syndrome (OSAHS) through the analysis of snoring sounds captured using microphones is presented in Ref. [10]. Snoring sounds are classified based on their relation to apnea and non-apnea events using deep learning techniques. Feature extraction methods such as MFCC, LPCC and LPMFCC, combined with convolutional neural network (CNN) and long-term memory neural network (LSTM) models were employed. The method based on MFCC and LSTM achieved 87 % accuracy in the binary classification of snoring, allowing also the esti mation of the apnea-hypopnea index (AHI) to determine the severity of OSAHS. On the other hand, in Ref. [11], a classifier based on Long Short-Term Memory (LSTM) was employed to distinguish between snoring related to respiratory events and normal snoring. Sounds were collected from 33 patients and 10 normal people, extracting features such as MFCC, Fbanks, short-term energy, and LPC. The LSTM model, which used multiple features, achieved an accuracy of 95.3 % in snoring classification and 81.6 % in OSAHS severity assessment, providing an effective auxiliary tool for diagnosis. Finally, in Ref. [12], snoring acoustic features were explored to improve noninvasive assessment of OSA. The study introduces new features based on snoring rhythm variability and trends in snoring en ergy throughout the night. These features, along with age and body mass index (BMI), were used to train an extreme boosting algorithm to estimate the apnea-hypopnea index (AHI) and classify the severity of OSA. The results show a significant correlation (R = 0.786) between the estimated AHI and that obtained with polysomnography, outperforming previous models. In this context, this paper focuses on the development of an intelli gent system based on the ESP32 microcontroller for monitoring, pro cessing and transmission of snoring signals. A high-sensitivity MEMS microphone is used to capture snoring sounds accurately. The system was designed to be noninvasive, allowing its use without affecting pa tients’ sleep. One of the most innovative aspects of this work is the implementa tion of data transmission via Wi-Fi, which facilitates real-time moni toring. The intuitive and user-friendly graphical user interface allows visualization of snoring detection, monitoring and continuous evalua tion. This system has great potential for both medical professionals and patients, as it offers an accessible tool for tracking snoring patterns, thus contributing to the diagnosis and treatment of conditions such as OSA.
Extent
7 P.
URI: https://repositorio.escuelaing.edu.co/handle/001/3419
Collections

Collections

AB - Ecitrónica
References

J. Xie, X. Aubert, X. Long, J. van Dijk, B. Arsenali, P. Fonseca, S. Overeem, Audiobased snore detection using deep neural networks, Comput. Methods Progr. Biomed. 200 (2021) 105917, https://doi.org/10.1016/j.cmpb.2020.105917. Epub 2020 Dec 25. PMID: 33434817.

Stelios A. Mitilineos, Nicolas-Alexander Tatlas, Georgia Korompili, Lampros Kokkalas, Stelios M. Potirakis, A real-time snore detector using neural networks and selected sound features, Engineering Proceedings 11 (1) (2021) 8, https://doi.org/10.3390/ASEC2021-11176

S. Bottia, H. Barrios-Bermúdez, D. Rosselli, Prevalence of sleep apnea in Colombia: an analysis of Ministry of Health databases, Iatreia 37 (1) (2023) [Internet], [cited September 12, 2024], https://revistas.udea.edu.co/index.php/iatreia/article/vi ew/351336.

] Young Terry, Paul E. Peppard, Daniel J. Gottlieb, Epidemiology of obstructive sleep apnea: a population health perspective, Am. J. Respir. Crit. Care Med. 165 (9) (2002) 1217–1239.

P. Bernstein, J.H. Ebba, Snoring versus obstructive sleep apnea: a case report, Perm. J. 10 (1) (2006) 21–23, https://doi.org/10.7812/TPP/05-116.

Z. Su, S. Kumar, T.E. Tavolara, M.N. Gurcan, S. Segal, M.K.K. Niazi, Predicting obstructive sleep apnea severity from craniofacial images using ensemble machine learning models, Proc. SPIE-Int. Soc. Opt. Eng. 12465 (2023) 124652P, https://doi. org/10.1117/12.2654353.

K. Qian, C. Janott, V. Pandit, Z. Zhang, C. Heiser, W. Hohenhorst, M. Herzog, W. Hemmert, B. Schuller, Classification of the excitation location of snore sounds in the upper airway by acoustic multifeature analysis, IEEE Trans. Biomed. Eng. 64 (8) (2017) 1731–1741, https://doi.org/10.1109/TBME.2016.2619675. Epub 2016 Oct 21. PMID: 28113249.

] E. Dafna, A. Tarasiuk, Y. Zigel, Automatic detection of whole night snoring events using non-contact microphone, PLoS One 8 (12) (2013) e84139, https://doi.org/ 10.1371/journal.pone.0084139. PMID: 24391903; PMCID: PMC3877189

H. Ankıs¸han, D. Yılmaz, Comparison of SVM and ANFIS for snore related sounds classification by using the largest Lyapunov exponent and entropy, Comput. Math. Methods Med. 2013 (2013) 238937, https://doi.org/10.1155/2013/238937. Epub 2013 Sep 30. PMID: 24194786; PMCID: PMC3806117.

F. Shen, S. Cheng, Z. Li, K. Yue, W. Li, L. Dai, Detection of snore from OSAHS patients based on deep learning, J Healthc Eng 2020 (2020) 8864863, https://doi. org/10.1155/2020/8864863. PMID: 33456742; PMCID: PMC7787852.

Siyi Cheng, Chao Wang, Keqiang Yue, Ruixue Li, Fanlin Shen, Wenjie Shuai, Wenjun Li, Lili Dai, Automated sleep apnea detection in snoring signal using long short-term memory neural networks, Biomed. Signal Process Control 71 (2022) 103238, https://doi.org/10.1016/j.bspc.2021.103238.

] R. Patino J, et al., Dwi simulation-assisted machine learning models for microstructure estimation, Computational Diffusion MRI. Mathematics and Visualization (2020) 125–134.

] R. Twiggs, B. Malphrus, J. Muylaert, SSC10- XII-3 the QB50 program, the first CubeSat constellations doing science, in: 24th AIAA/USU Conference on Small Satellites, 2010, p. 12.

Audio Toolbox - MATLAB & Simulink. MathWorks. [Online]. Dis- ponible at: https ://la.mathworks.com/products/audio.html. [Accessed November 2, 2023].

MathWorks. Parametric Equalizer Design MATLAB Simulink MathWorks. [Online online]. Available At: https://la.mathworks.com/help/audio/ug/parametric-equ alizer-design.html. [Accessed November 2, 2023].

] MathWorks. YAMNet - MATLAB Simulink. [Online]. Available at: https://la.mat hworks.com/help/audio/ref/yamnet.html. [Accessed November 2, 2023]

Python. Python SciPy Butterworth filter. [Online]. Available at: https://python guia.com/filtro-de-python-scipy-butterworth/. [Accessed 2 November. 2023].

Atomic14. Esp32-i2s-mic-test. [Online]. Available at: https://github.com/ato mic14/esp32-i2s-mic-test. [Accessed November 2, 2023.].

] H.P. Maimon N, Does snoring intensity correlate with the severity of obstructive sleep apnea? J. Clin. Sleep Med. (2010)

] J. Fiz, J. Abad, R. Jane, M. Riera, M. Mananas, P. Caminal, J. Morera, Acoustic analysis of snoring sound in patients with simple snoring and obstructive sleep apnoea, Eur. Respir. J. 9 (1996) 2365–2370, https://doi.org/10.1183/ 09031936.96.09112365.

MathWorks. ClassifySound - MATLAB Simulink. [Online]. Dis- ponible at: https ://la.mathworks.com/help/audio/ref/classifysound.html. [Accessed November 2, 2023].

MathWorks. Lowpass - MATLAB Simulink [Online]. Available at: https://la.mathw orks.com/help/signal/ref/lowpass.html. [Accessed November 2, 2023].

MathWorks. Pspectrum - MATLAB Simulink. [Online]. Available at: https://la. mathworks.com/help/signal/ref/pspectrum.html. [Accessed November 2, 2023].

H. Alshaer, R. Hummel, M. Mendelson, T. Mariscal, D. Bradley, Objective relationship between sleep apnea and frequency of snoring assessed by machine learning, J. Clin. Sleep Med. 15 (3) (2019), https://doi.org/10.5664/jcsm.7676. Retrieved from.