Solunum Hastalıkları ile İlişkili Semptom Seslerinin Sınıflandırılması

Abstract

Covid-19 gibi solunum yolu enfeksiyonlarının erken tespiti, hastalığın daha kolay tedavisine ve hastanın daha rahat bir süre geçirmesine yol açarak ciddi komplikasyon olasılığını azaltabilir. Öksürme ve hapşırma gibi solunum seslerinin sıklığı, şiddeti ve türü (kuru veya balgamlı), hastalığın teşhisi, tedavisi ve davranışlarının tespitinde tıp uzmanları için çıkarılabilen zengin bilgiler taşımaktadır. Bunun için, makine veya derin öğrenimine dayalı otomatik yaklaşımların geliştirilmesi oldukça önemlidir. Center for Open Science (OSFHOME), 2020 yılında güncellediği veri küme üzerine, bu alanda çalışan araştırmacıları, ses kayıtlarını kullanarak hastalık seslerinin otomatik algılanması için makine öğrenimi modelleri oluşturmaya davet etti. Veri seti, “Pfizer Digital Medicine Challenge” için oluşturulmuştur ve amacı öksürme ve hapşırma gibi seslerinin tespiti için makine öğrenimi modellerinin geliştirilmesidir. Veri seti üç parçaya ayrılmıştır; eğitim, doğrulama ve test kümeleri. Sunulan çalışmada, bu veri seti üzerine yeni bir makine öğrenimi sistemi önerildi. Eğitim, doğrulama ve test örneklerinden öznitelikler elde edildikten sonra, dört farklı sınıflandırıcının parametrelerini hesaplamak için doğrulama veri kümesi kullanıldı ve son aşamada test veri kümesi üzerine sınıflandırma gerçekleştirildi. Elde edilen sonuçlara göre, radyal tabanlı çekirdek fonksiyonlu destek vektör makine (DVM) sınıflandırıcısı solunum seslerini diğer seslere karşı, %76 civarında bir doğruluk oranıyla diğer sınıflandırıcılara göre daha başarılı sınıflandırdı.

Keywords

• Makine öğrenimi
• Solunum sesleri
• Sınıflandırma
• Hapşırık
• COVID-19
• Öksürük

Classification of Symptom Sounds Associated with Respiratory Disease

Abstract

Early detection of respiratory infections such as Covid-19 can lead to the easier treatment of the disease and a more comfortable time for the patient, reducing the likelihood of serious complications. The frequency, severity, and type (dry or phlegm) of respiratory sounds such as coughing and sneezing carry a wealth of information that can be extracted for medical professionals in diagnosing the disease, treating it, and determining its behavior. For this, it is very important to develop automated approaches based on machine or deep learning. Center for Open Science (OSFHOME) invited researchers working in this field to create machine learning models for automatic detection of the disease sounds using sound recordings, based on the dataset it updated in the 2020 year. The dataset was created for the "Pfizer Digital Medicine Challenge" and its purpose is to develop machine learning models for detecting sounds such as coughing and sneezing. The dataset is divided into three parts; training, validation, and test sets. In the presented study, a new machine learning system is proposed on this dataset. After the features were obtained from the training, validation, and test samples, the validation dataset was used to calculate the parameters of the four different classifiers, and in the final stage, the classification was performed on the test set. According to the results, the radial-based kernel function support vector machine (RBF-SVM) classifier classified respiratory sounds against other sounds more successfully than other classifiers with an accuracy rate of around 76%.

Keywords

• Respiratory sounds
• Classification
• Machine learning
• COVID-19
• Cough
• Sneeze

References

1. A. A. Saraiva et al., “Classification of respiratory sounds with convolutional neural network,” Bioinforma. 2020 - 11th Int. Conf. Bioinforma. Model. Methods Algorithms, Proceedings; Part 13th Int. Jt. Conf. Biomed. Eng. Syst. Technol. BIOSTEC 2020, pp. 138–144, 2020, doi: 10.5220/0008965101380144.
2. D. Sánchez Morillo, S. Astorga Moreno, M. Á. Fernández Granero, and A. León Jiménez, “Computerized analysis of respiratory sounds during COPD exacerbations,” Comput. Biol. Med., vol. 43, no. 7, pp. 914–921, Aug. 2013, doi: 10.1016/J.COMPBIOMED.2013.03.011.
3. M. Melek, “Diagnosis of COVID-19 and non-COVID-19 patients by classifying only a single cough sound,” Neural Comput. Appl. 2021, pp. 1–12, Jul. 2021, doi: 10.1007/S00521-021-06346-3.
4. U. R. Abeyratne, V. Swarnkar, A. Setyati, and R. Triasih, “Cough Sound Analysis Can Rapidly Diagnose Childhood Pneumonia,” Ann. Biomed. Eng. 2013 4111, vol. 41, no. 11, pp. 2448–2462, Jun. 2013, doi: 10.1007/S10439-013-0836-0.
5. V. Swarnkar, U. R. Abeyratne, A. B. Chang, Y. A. Amrulloh, A. Setyati, and R. Triasih, “Automatic Identification of Wet and Dry Cough in Pediatric Patients with Respiratory Diseases,” Ann. Biomed. Eng. 2013 415, vol. 41, no. 5, pp. 1016–1028, Jan. 2013, doi: 10.1007/S10439-013-0741-6.
6. M. Al-khassaweneh and R. B. Abdelrahman, “A signal processing approach for the diagnosis of asthma from cough sounds,” http://dx.doi.org/10.3109/03091902.2012.758322, vol. 37, no. 3, pp. 165–171, Apr. 2013, doi: 10.3109/03091902.2012.758322.
7. H. Chatrzarrin, A. Arcelus, R. Goubran, and F. Knoefel, “Feature extraction for the differentiation of dry and wet cough sounds,” in MeMeA 2011 - 2011 IEEE International Symposium on Medical Measurements and Applications, Proceedings, 2011, pp. 162–166, doi: 10.1109/MeMeA.2011.5966670.
8. E. Nemati, M. M. Rahman, V. Nathan, K. Vatanparvar, and J. Kuang, “A Comprehensive Approach for Classification of the Cough Type,” in Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, Jul. 2020, vol. 2020-July, pp. 208–212, doi: 10.1109/EMBC44109.2020.9175345.

9. V. Bhateja, A. Taquee, and D. K. Sharma, “Pre-Processing and Classification of Cough Sounds in Noisy Environment using SVM,” in 2019 4th International Conference on Information Systems and Computer Networks, ISCON 2019, Nov. 2019, pp. 822–826, doi: 10.1109/ISCON47742.2019.9036277.
10. N. Simou, N. Stefanakis, and P. Zervas, “A universal system for cough detection in domestic acoustic environments,” in European Signal Processing Conference, Jan. 2021, vol. 2021-January, pp. 111–115, doi: 10.23919/Eusipco47968.2020.9287659.
11. F. Barata, K. Kipfer, M. Weber, P. Tinschert, E. Fleisch, and T. Kowatsch, “Towards device-agnostic mobile cough detection with convolutional neural networks,” in 2019 IEEE International Conference on Healthcare Informatics, ICHI 2019, Jun. 2019, doi: 10.1109/ICHI.2019.8904554.
12. “Center for Open Science.” https://www.cos.io/?_ga=2.107126978.1325905396.1631883086-1696097544.1629751041 (accessed Sep. 17, 2021).
13. “OSF | Dataset of sounds of symptoms associated with respiratory sickness Wiki.” https://osf.io/tmkud/wiki/home/ (accessed Sep. 17, 2021).
14. R. Gonzalez, “Better Than MFCC Audio Classification Features,” Era Interact. Media, vol. 9781461435013, pp. 291–301, Oct. 2013, doi: 10.1007/978-1-4614-3501-3_24.
15. M. A. Hossan, S. Memon, and M. A. Gregory, “A novel approach for MFCC feature extraction,” 4th Int. Conf. Signal Process. Commun. Syst. ICSPCS’2010 - Proc., 2010, doi: 10.1109/ICSPCS.2010.5709752.
16. Y. Wang and B. Lawlor, “Speaker recognition based on MFCC and BP neural networks,” 2017 28th Irish Signals Syst. Conf. ISSC 2017, Jul. 2017, doi: 10.1109/ISSC.2017.7983644.
17. A. Winursito, R. Hidayat, and A. Bejo, “Improvement of MFCC feature extraction accuracy using PCA in Indonesian speech recognition,” 2018 Int. Conf. Inf. Commun. Technol. ICOIACT 2018, vol. 2018-January, pp. 379–383, Apr. 2018, doi: 10.1109/ICOIACT.2018.8350748.
18. N. Melek Manshouri, “Identifying COVID-19 by using spectral analysis of cough recordings: a distinctive classification study,” Cogn. Neurodynamics 2021, pp. 1–15, Jul. 2021, doi: 10.1007/S11571-021-09695-W.
19. K. S. (Krothapalli S. Rao and Manjunath K. E., “Speech recognition using articulatory and excitation source features,” p. 92.
20. M. Melek, N. Manshouri, and T. Kayikcioglu, “Low-Cost Brain-Computer Interface Using the Emotiv Epoc Headset Based on Rotating Vanes,” Trait. du Signal, vol. 37, no. 5, pp. 831–837, Nov. 2020, doi: 10.18280/ts.370516.
21. U. Ozkaya, F. Melgani, M. B. Bejiga, L. Seyfi, and M. Donelli, “GPR B scan image analysis with deep learning methods,” Measurement 2020, 165, 107770. doi: 10.1016/j.measurement.2020.107770.