The Lung Sound Classification Using Mel Frequency Cepstral Coefficients and Convolutional Neural Networks
Abstract
Background: Challenges in diagnosing respiratory disorders are often caused by the lack of technological tools capable of accurately recognizing lung sound patterns, thereby reducing the potential for subjective misdiagnosis by medical personnel.
Objective: This study aims to develop a lung sound classification model that is able to detect respiratory disorders early and accurately.
Methods: The method used includes a combination of data augmentation techniques and Mel Frequency Cepstral Coefficient (MFCC) feature extraction to improve the performance of Convolutional Neural Network (CNN) in classifying lung sounds. A total of 1,350 lung audio recordings were categorized into nine classes, including normal and abnormal sounds. The augmentation techniques applied include the addition of white noise, pitch scaling, time stretching, and random gain to enrich the variety of training data.
Result: The results show that the E-CNN2D model is able to achieve an accuracy of up to 95%, surpassing the previous model, which had an accuracy range of 83-93%.
Conclusion: With these results, this study has the potential to be a fast and accurate diagnostic tool solution so that it can support medical personnel in reducing the risk of subjective misdiagnosis in respiratory disorders.
References
[1] W. Sun, Y. Zhang, and F. Chen, “Research on heart and lung sound separation method based on DAE–NMF–VMD,” EURASIP J. Adv. Signal Process., vol. 2024, no. 1, 2024, doi: 10.1186/s13634-024-01152-0.
[2] O. Distler et al., “Early View Predictors of progression in systemic sclerosis patients with interstitial lung disease,” 2020.
[3] X. Li, Y. Shang, J. Wei, and Y. Zhou, “Research on electronic stethoscope system and signal processing algorithm,” J. Phys. Conf. Ser., vol. 2634, no. 1, 2023, doi: 10.1088/1742-6596/2634/1/012037.
[4] M. Fraiwan, L. Fraiwan, B. Khassawneh, and A. Ibnian, “A dataset of lung sounds recorded from the chest wall using an electronic stethoscope,” Data Br., vol. 35, p. 106913, 2021, doi: 10.1016/j.dib.2021.106913.
[5] Y. Kim, Y. K. Hyon, S. Lee, S. D. Woo, T. Ha, and C. Chung, “The coming era of a new auscultation system for analyzing respiratory sounds,” BMC Pulm. Med., vol. 22, no. 1, pp. 1–11, 2022, doi: 10.1186/s12890-022-01896-1.
[6] L. Arts, E. H. T. Lim, P. M. van de Ven, L. Heunks, and P. R. Tuinman, “The diagnostic accuracy of lung auscultation in adult patients with acute pulmonary pathologies: a meta-analysis,” Sci. Rep., vol. 10, no. 1, pp. 1–11, 2020, doi: 10.1038/s41598-020-64405-6.
[7] J. S. Park, K. Kim, J. H. Kim, Y. J. Choi, and K. Kim, “A machine learning approach to the development and prospective evaluation of a pediatric lung sound classification model,” Sci. Rep., no. 0123456789, pp. 1–10, 2023, doi: 10.1038/s41598-023-27399-5.
[8] H. Hafe-Dys, A. Bręborowicz, P. Kleka, J. Kociński, and A. Biniakowski, “The Accuracy of Lung Auscultation in the Practice of Physicians and Medical Students,” Adv. Med. Imaging, Detect. Diagnosis, pp. 853–871, 2023, doi: 10.1201/9781003298038-28.
[9] B. Z. Francisca, S. K. Dafne, S. D. Ignacio, and B. N. Pablo, “Lung auscultation in the 21th century,” Rev. Chil. Pediatr., vol. 91, no. 4, pp. 500–506, 2020, doi: 10.32641/rchped.v91i4.1465.
[10] G. Petmezas et al., “Automated Lung Sound Classification Using a Hybrid CNN-LSTM Network and Focal Loss Function,” Sensors, vol. 22, no. 3, 2022, doi: 10.3390/s22031232.
[11] D. Rafiqo, Y. Suyanto, and C. Atmaji, “Klasifikasi Suara Paru-Paru Berdasarkan Ciri MFCC,” IJEIS (Indonesian J. Electron. Instrum. Syst., vol. 12, no. 1, p. 1, 2022, doi: 10.22146/ijeis.70813.
[12] H. Gulzar, J. Li, A. Manzoor, S. Rehmat, U. Amjad, and H. J. Khan, “Detection of Crackles and Wheezes in Lung Sound Using Transfer Learning,” Heal. Informatics - An Int. J., vol. 12, no. 2, pp. 01–14, 2023, doi: 10.5121/hiij.2023.12201.
[13] L. Nanni, G. Maguolo, and M. Paci, “Data augmentation approaches for improving animal audio classification,” Ecol. Inform., vol. 57, no. December 2019, p. 101084, 2020, doi: 10.1016/j.ecoinf.2020.101084.
[14] Y. Kim et al., “Respiratory sound classification for crackles, wheezes, and rhonchi in the clinical field using deep learning,” Sci. Rep., vol. 11, no. 1, pp. 1–11, 2021, doi: 10.1038/s41598-021-96724-7.
[15] K. L. Ang, S. Member, J. Kah, and P. Seng, “Big Data and Machine Learning With Hyperspectral Information in Agriculture,” vol. 9, pp. 36699–36718, 2021, doi: 10.1109/ACCESS.2021.3051196.
[16] A. H. Al-badri, N. O. R. A. Ismail, K. Al-dulaimi, A. Rehman, I. Abunadi, and S. A. L. I. Bahaj, “Hybrid CNN Model for Classification of Rumex Obtusifolius in Grassland,” IEEE Access, vol. 10, no. August, pp. 90940–90957, 2022, doi: 10.1109/ACCESS.2022.3200603.
[17] A. Agustina, F. Yanto, E. Budianita, I. Iskandar, and F. Syafria, “Klasifikasi Penyakit Tanaman Padi Menggunakan Metode Cnn Arsitektur Densenet-121 Dan Augmentasi Data,” J. Inf. Syst. Informatics Eng., vol. 8, no. 1, pp. 124–134, 2024.
[18] M. Muadz, E. Yahya, and A. Rizal, “Lung Sound Classification Using Wavelet Transformation and Recurrent Neural Networks,” e-Proceeding Eng., vol. 8, no. 2, pp. 3209–3217, 2021.
[19] Z. K. H. Abdul, “Mel Frequency Cepstral Coefficient and Its Applications : A Review,” vol. 10, no. October, 2022.
[20] J. Amose and M. Vairavan, “Optimal chest position of auscultation for chronic obstructive pulmonary disease diagnosis using machine learning,” vol. 32, no. 3, pp. 1417–1425, 2023, doi: 10.11591/ijeecs.v32.i3.pp1417-1425.
[21] A. R. Adriansyah, K. D. Prasetyo, H. A. A. Faruqi, “Pengenalan pola fonem vokal menggunakan short time fourier transform (STFT) dan fitur mel frequency cepstral coefficient (MFCC),” vol. 7, no. 1, pp. 1–6, 2021.
[22] R. Hazra and S. Majhi, “Detecting respiratory diseases from recorded lung sounds by 2D CNN,” Proc. 2020 Int. Conf. Comput. Commun. Secur. ICCCS 2020, 2020, doi: 10.1109/ICCCS49678.2020.9277101.
This work is licensed under a Creative Commons Attribution 4.0 International License.
