Deep Transfer Learning for Pediatric Respiratory Sound Classification
Keywords:
Deep transfer learning, Classifying pediatric respiratory sound, Spectrogram image analysisAbstract
Lower respiratory tract infections (LRTIs) in infants are a major global health problem that can lead to illness and death. The cause of infections can be from a variety of bacteria or viruses, resulting in different breathing sounds. Diagnosis of LRTIs in infants is complex because lungs and heart are close together. This study emphasized on a development of Artificial Intelligence (AI) systems for classifying infant breathing sounds using Deep Transfer Learning (DTL) techniques. Five DTL models: VGG-16, VGG-19, EfficientNet B0, EfficientNet B7, and MobileNet were applied to classify respiratory-sound images in both frequency and time domains for 5 different types of respiratory sounds: Normal, Crackle, Rhonchi, Stridor and Wheezing. MobileNet model achieves the highest accuracy of more than 80%, comparing to the other 4 DTL models. Thus, MobileNet model has a strong potential to be used for assisting medical personnel to accurate diagnosis of LRTIs in infants and enabling appropriate and timely treatment.
References
Tchatchouang S, Nzouankeu A, Kenmoe S, Ngando L, Penlap V, Fonkoua M-C, et al. Bacterial Aetiologies of Lower Respiratory Tract Infections among Adults in Yaoundé, Cameroon. BioMed Research International. 2019; 2019: 4834396; DOI 10.1155/2019/4834396.
Matloobi A, Buday T, Brozmanova M, Konarska M, Poliacek I, Martvon L, et al. The effect of stimulation and unloading of baroreceptors on cough in experimental conditions. Respiratory Physiology & Neurobiology. 2022; 303: 103921; DOI: 10.1016/j.resp.2022.103921.
Musher DM, Abers MS, Bartlett JG. Evolving Understanding of the Causes of Pneumonia in Adults, With Special Attention to the Role of Pneumococcus. Clinical Infectious Diseases. 2017; 65(10): 1736-1744.
Shields MD, Bush A, Everard ML, McKenzie S, Primhak R. BTS guidelines: Recommendations for the assessment and management of cough in children. Thorax. 2008 Apr; 63 Suppl 3: iii1-iii15; DOI: 10.1136/thx.2007.077370.
Jain S, Self WH, Wunderink RG, Fakhran S, Balk R, Bramley AM, et al. Community-Acquired Pneumonia Requiring Hospitalization among U.S. Adults. N Engl J Med. 2015; 373 (5): 415-427; DOI: 10.1056/NEJMoal500245.
Lanata CF, Black RE. Acute Lower Respiratory Infections. In: Semba RD, Bloem MW, Piot P, editors. Nutrition and Health in Developing Countries. Totowa, NJ: Humana Press; 2008; 179-214; DOI: 10.1007/978-1-59745-464-3_7.
Morrow BM. Airway clearance therapy in acute paediatric respiratory illness: A state-of-the-art review. S Afr J Physiother. 2019; 75(1): 1295; DOI: 10.4102/sajp.v75i1.1295.
Mamouris P, Henrard S, Molenberghs G, Verhaegen J, Lin G, Vaes B. Pneumococcal vaccination prevented severe LRTIs in adults: a causal inference framework applied in registry data. J Clin Epidemiol. 2022; 143: 118-127; DOI: 10.1016/j.jclinepi.2021.12.008.
GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020; 396(10258): 1204-1222; DOI: 10.1016/S0140-6736(20)30925-9.
Suwanjutha S, Chantarojsiri T, Prapphal N, Suntornlohanakul S, Laohapand C, Wongpaitoon N. Acute Lower Respiratory Infections. In: Deerojanawong J, Boonjindasap V, editors. Guideline for the management of Acute Respiratory Infections in Children. Nonthaburi: Beyond Enterprise Company LIM; 2019. 72-155.
Schwartz KL, Langford BJ, Daneman N, Chen B, Brown KA, McIsaac W, et al. Unnecessary antibiotic prescribing in a Canadian primary care setting: a descriptive analysis using routinely collected electronic medical record data. CMAJ Open. 2020; 8(2): E360-E369; DOI: 10.9778/cmajo.20190175.
Kumar S, Kumar H. Classification of COVID-19 X-ray images using transfer learning with visual geometrical groups and novel sequential convolutional neural networks. MethodsX. 2023; 11: 102295; DOI: 10.1016/j.mex.2023.102295.
Simonyan K, Zisserman A. Very Deep Convolutional Networks for Large-Scale Image Recognition. 3rd International Conference on Learning Representations (ICLR 2015), Computational and Biological Learning Society. 2014; pp. 1–14; San Diego, CA, USA; DOI: 10.48550/arXiv.1409.1556.
Porter P, Brisbane J, Tan J, Bear N, Choveaux J, Della P, et al. Diagnostic Errors Are Common in Acute Pediatric Respiratory Disease: A Prospective, Single-Blinded Multicenter Diagnostic Accuracy Study in Australian Emergency Departments. Front Pediatr. 2021; 9: 736018; DOI: 10.3389/fped.2021.736018.
Serbes G, Sakar CO, Kahya Y, Aydin N. Pulmonary crackle detection using time-frequency analysis. Digital Signal Processing. 2013; 23(3): 1012-1021; DOI: 10.1016/j.dsp.2012.12.009.
McGee S. Auscultation of the Lungs. In: McGee S, editor. Evidence-Based Physical Diagnosis (Fourth Edition). Philadelphia: Elsevier; 2018; p: 261-274. e4.
Zulfiqar R, Majeed F, Irfan R, Rauf HT, Benkhelifa E, Belkacem AN. Abnormal Respiratory Sounds Classification Using Deep CNN Through Artificial Noise Addition. Front Med (Lausanne). 2021; 8:714811; DOI: 10.3389/fmed.2021.714811.
Reichert S, Gass R, Brandt C, Andrès E. Analysis of respiratory sounds: state of the art. Clin Med Circ Respirat Pulm Med. 2008; 2: 45-58; DOI: 10.4137/ccrpm.s530.
Andrès E, Gass R, Charloux A, Brandt C, Hentzler A. Respiratory sound analysis in the era of evidence-based medicine and the world of medicine 2.0. J Med Life. 2018; 11(2): 89-106.
Kim Y, Hyon Y, Jung SS, Lee S, Yoo G, Chung C, et al. Respiratory sound classification for crackles, wheezes, and rhonchi in the clinical field using deep learning. Scientific reports. 2021;11(1):1-11; DOI: 10.1038/s41598-021-96724-7.
Yosinski J, Clune J, Bengio Y, Lipson H. How transferable are features in deep neural networks? Advances in neural information processing systems (NIPS). 2014; 27: 3320-3328; DOI: 10.48550/arXv.1411.1792.
Huh M, Agrawal P, Efros AA. What makes ImageNet good for transfer learning? arXiv: 1608.08614v2. 2016; DOI: 10.48550/arXiv.1608.08614.
Faiza MS, Moftah RAM, Bozed KA. Applying Convolutional Neural Network and Transfer Learning to Mel-Spectrograms for COVID-19 Diagnosis via Cough Sounds. Int J Sci Res Publ. 2023; 13(06): 32-3713805; DOI: 10.29322/IJSRP.13.06.2023.
Ali AA, Al-Marzouqi H. Melanoma detection using regular convolutional neural networks.2017 International Conference on Electrical and Computing Technologies and Applications (ICECTA), United Arab Emirates, 2017: 1-5; DOI: 10.1109/ICECTA.2017.8252041.
Bansal M, Kumar M, Sachdeva M, Mittal A. Transfer learning for image classification using VGG19: Caltech-101 image data set. J Ambient Intell and Humaniz Comput. 2023;14(4): 3609–3620; DOI: 10.1007/s12652-021-03488-z.
Hoang VT, Jo KH. Practical Analysis on Architecture of EfficientNet. 2021 14th International Conference on Human System Interaction (HSI); 2021 July 8-10; pp. 1-4; Gdansk, Poland: Institute of Electrical and Electronics Engineers; 2021; DOI: 10.1109/HSI52170.2021.9538782.
Howard AG, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T, et al. Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv: 1704.04861v1. 2017; DOI: 10.48550/arXiv.1704.04861.
Srivastava H, Sarawadekar K. A Depthwise Separable Convolution Architecture for CNN Accelerator. 2020 IEEE Applied Signal Processing Conference (ASPCON); 2020 Oct 7-9; pp. 1-5; Kolkata, India: Institute of Electrical and Electronics Engineers; 2020; DOI: 10.1109/ASPCON49795.2020.9276672.
Li H, Wang J, Xiong N, Zhang Y, Vasilakos AV, Luo X. A Siamese Inverted Residuals Network Image Steganalysis Scheme based on Deep Learning. ACM Trans Multimedia Comput Commun Appl. 2023;19(6):214. DOI: 10.1145/3579166.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 KKU Research Journal (Graduate Studies)
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
