{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T18:08:53Z","timestamp":1791482933483,"version":"4.3.4"},"reference-count":25,"publisher":"MDPI AG","issue":"20","license":[{"start":{"date-parts":[[2025,10,21]],"date-time":"2025-10-21T00:00:00Z","timestamp":1761004800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100003141","name":"Secretar\u00eda de Ciencia, Humanidades, Tecnolog\u00eda e Innovaci\u00f3n","doi-asserted-by":"publisher","award":["1234286"],"award-info":[{"award-number":["1234286"]}],"id":[{"id":"10.13039\/501100003141","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JCM"],"abstract":"<jats:p>Background: Automatic adventitious lung sound classification using deep learning is a promising strategy for objective respiratory disease screening. Evaluating model performance is challenging, particularly with imbalanced clinical datasets. This study compares CNN architectures and proposes a dual-stream classification approach. Methods: Using the public ICBHI 2017 dataset, we compared five pre-trained architectures: VGG16, VGG19, InceptionV3, MobileNetV2, and ResNet152V2. To mitigate class imbalance, we implemented pitch shifting, random shifting, and mixup data augmentation. We also developed and evaluated a novel VGGish-dual-stream network. The primary endpoint was the Average Score (AS), the arithmetic mean of Sensitivity and Specificity. Results: Among benchmarked models, ResNet152V2 achieved the highest AS (0.541), approaching the state-of-the-art range (0.56\u20130.58). This performance was characterised by a high Specificity (0.67) but low Sensitivity (0.41). Our proposed dual-stream network yielded a more balanced, albeit slightly lower, performance with an AS of 0.508. Conclusions: Standard CNN architectures like ResNet152V2 can achieve competitive classification performance but may exhibit a clinically significant bias towards high specificity at the expense of sensitivity. This trade-off poses a risk of missing pathological events (false negatives). To ensure clinical safety and utility, future work must prioritise strategies that explicitly improve model sensitivity.<\/jats:p>","DOI":"10.3390\/jcm14207427","type":"journal-article","created":{"date-parts":[[2025,10,21]],"date-time":"2025-10-21T14:53:19Z","timestamp":1761058399000},"page":"7427","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":5,"title":["Comparison of Deep Neural Networks for the Classification of Adventitious Lung Sounds"],"prefix":"10.3390","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8473-0534","authenticated-orcid":false,"given":"Said","family":"Polanco-Martag\u00f3n","sequence":"first","affiliation":[{"name":"Intelligent Systems Department, Universidad Polit\u00e9cnica de Victoria, Ciudad Victoria 87138, Tamaulipas, Mexico"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1044-2710","authenticated-orcid":false,"given":"Yahir","family":"Hern\u00e1ndez-Mier","sequence":"additional","affiliation":[{"name":"Intelligent Systems Department, Universidad Polit\u00e9cnica de Victoria, Ciudad Victoria 87138, Tamaulipas, Mexico"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0102-8227","authenticated-orcid":false,"given":"Marco Aurelio","family":"Nu\u00f1o-Maganda","sequence":"additional","affiliation":[{"name":"Intelligent Systems Department, Universidad Polit\u00e9cnica de Victoria, Ciudad Victoria 87138, Tamaulipas, Mexico"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0965-6714","authenticated-orcid":false,"given":"Jos\u00e9 Hugo","family":"Barr\u00f3n-Zambrano","sequence":"additional","affiliation":[{"name":"Digital Systems Group, Electronics Department, Instituto Nacional de Astrof\u00edsica, \u00d3ptica y Electr\u00f3nica, San Andr\u00e9s Cholula 72840, Puebla, Mexico"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5474-4150","authenticated-orcid":false,"given":"Andrea","family":"Magad\u00e1n-Salazar","sequence":"additional","affiliation":[{"name":"Computer Science Department, Tecnol\u00f3gico Nacional de M\u00e9xico\/CENIDET, Cuernavaca 62493, Morelos, Mexico"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"C\u00e9sar Alejandro","family":"Medell\u00edn-Vergara","sequence":"additional","affiliation":[{"name":"Intelligent Systems Department, Universidad Polit\u00e9cnica de Victoria, Ciudad Victoria 87138, Tamaulipas, Mexico"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2025,10,21]]},"reference":[{"key":"ref_1","unstructured":"Forum of Internation Respiratory Societies (2021). The Global Impact of Respiratory Disease, European Respiratory Society."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"834","DOI":"10.4187\/respcare.01072","article-title":"Technology for Enhancing Chest Auscultation in Clinical Simulation","volume":"56","author":"Ward","year":"2011","journal-title":"Respir. Care"},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Sencar, H.T., and Memon, N. (2013). Digital Image Forensics, Springer.","DOI":"10.1007\/978-1-4614-0757-7"},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Pramono, R.X.A., Bowyer, S., and Rodriguez-Villegas, E. (2017). Automatic adventitious respiratory sound analysis: A systematic review. PLoS ONE, 12.","DOI":"10.1371\/journal.pone.0177926"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"2872","DOI":"10.1109\/TBME.2022.3156293","article-title":"Lung Sound Classification Using Co-Tuning and Stochastic Normalization","volume":"69","author":"Nguyen","year":"2022","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"ref_6","first-page":"70","article-title":"Lung Sound Recognition Based on Pre-Trained Convolutional Neural Network","volume":"5","author":"Bao","year":"2022","journal-title":"Acad. J. Eng. Technol. Sci."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Petmezas, G., Cheimariotis, G.A., Stefanopoulos, L., Rocha, B., Paiva, R.P., Katsaggelos, A.K., and Maglaveras, N. (2022). Automated lung sound classification using a hybrid CNN-LSTM network and focal loss function. Sensors, 22.","DOI":"10.3390\/s22031232"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"3201","DOI":"10.1007\/s42835-023-01425-y","article-title":"Ensemble learning model for classification of respiratory anomalies","volume":"18","author":"Kim","year":"2023","journal-title":"J. Electr. Eng. Technol."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"383","DOI":"10.1109\/OJEMB.2024.3402139","article-title":"Attention Feature Fusion Network via Knowledge Propagation for Automated Respiratory Sound Classification","volume":"5","author":"Crisdayanti","year":"2024","journal-title":"IEEE Open J. Eng. Med. Biol."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"2595","DOI":"10.1109\/JBHI.2020.3048006","article-title":"A lightweight CNN model for detecting respiratory diseases from lung auscultation sounds using EMD-CWT-based hybrid scalogram","volume":"25","author":"Shuvo","year":"2020","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"035001","DOI":"10.1088\/1361-6579\/ab03ea","article-title":"An open access database for the evaluation of respiratory sound classification algorithms","volume":"40","author":"Rocha","year":"2019","journal-title":"Physiol. Meas."},{"key":"ref_12","unstructured":"Simonyan, K., and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Leibe, B., Matas, J., Sebe, N., and Welling, M. (2016, January 11\u201314). Identity Mappings in Deep Residual Networks. Proceedings of the Computer Vision\u2014ECCV 2016, Amsterdam, The Netherlands.","DOI":"10.1007\/978-3-319-46478-7"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., and Wojna, Z. (2016, January 27\u201330). Rethinking the inception architecture for computer vision. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA.","DOI":"10.1109\/CVPR.2016.308"},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Wang, M., Mei, Q., Song, X., Liu, X., Kan, R., Yao, F., Xiong, J., and Qiu, H. (2023). A Machine Anomalous Sound Detection Method Using the lMS Spectrogram and ES-MobileNetV3 Network. Appl. Sci., 13.","DOI":"10.3390\/app132312912"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Gairola, S., Tom, F., Kwatra, N., and Jain, M. (2021, January 1\u20135). Respirenet: A deep neural network for accurately detecting abnormal lung sounds in limited data setting. Proceedings of the 2021 43rd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), Virtual.","DOI":"10.1109\/EMBC46164.2021.9630091"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Yang, Z., Liu, S., Song, M., Parada-Cabaleiro, E., and Schuller, B. (2020, January 25\u201329). Adventitious Respiratory Classification Using Attentive Residual Neural Networks. Proceedings of the Interspeech, Shanghai, China.","DOI":"10.21437\/Interspeech.2020-2790"},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Ma, Y., Xu, X., and Li, Y. (2020, January 25\u201329). LungRN+ NL: An improved adventitious lung sound classification using non-local block resnet neural network with mixup data augmentation. Proceedings of the Interspeech, Shanghai, China.","DOI":"10.21437\/Interspeech.2020-2487"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"2938","DOI":"10.1109\/JBHI.2021.3064237","article-title":"CNN-MoE based framework for classification of respiratory anomalies and lung disease detection","volume":"25","author":"Pham","year":"2021","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Ma, Y., Xu, X., Yu, Q., Zhang, Y., Li, Y., Zhao, J., and Wang, G. (2019, January 17\u201319). Lungbrn: A smart digital stethoscope for detecting respiratory disease using bi-resnet deep learning algorithm. Proceedings of the 2019 IEEE Biomedical Circuits and Systems Conference (BioCAS), Nara, Japan.","DOI":"10.1109\/BIOCAS.2019.8919021"},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Chambres, G., Hanna, P., and Desainte-Catherine, M. (2018, January 4\u20136). Automatic detection of patient with respiratory diseases using lung sound analysis. Proceedings of the 2018 International Conference on Content-Based Multimedia Indexing (CBMI), La Rochelle, France.","DOI":"10.1109\/CBMI.2018.8516489"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Maglaveras, N., Chouvarda, I., and de Carvalho, P. (2017, January 18\u201321). Hidden markov model based respiratory sound classification. Proceedings of the Precision Medicine Powered by pHealth and Connected Health: ICBHI 2017, Thessaloniki, Greece.","DOI":"10.1007\/978-981-10-7419-6"},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Yang, R., Lv, K., Huang, Y., Sun, M., Li, J., and Yang, J. (2023). Respiratory Sound Classification by Applying Deep Neural Network with a Blocking Variable. Appl. Sci., 13.","DOI":"10.3390\/app13126956"},{"key":"ref_24","first-page":"1","article-title":"RDLINet: A novel lightweight inception network for respiratory disease classification using lung sounds","volume":"72","author":"Roy","year":"2023","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Rocha, B.M., Pessoa, D., Marques, A., Carvalho, P., and Paiva, R.P. (2021). Automatic Classification of Adventitious Respiratory Sounds: A (Un)Solved Problem?. Sensors, 21.","DOI":"10.3390\/s21010057"}],"container-title":["Journal of Clinical Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2077-0383\/14\/20\/7427\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,21]],"date-time":"2025-10-21T15:16:28Z","timestamp":1761059788000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2077-0383\/14\/20\/7427"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,21]]},"references-count":25,"journal-issue":{"issue":"20","published-online":{"date-parts":[[2025,10]]}},"alternative-id":["jcm14207427"],"URL":"https:\/\/doi.org\/10.3390\/jcm14207427","relation":{},"ISSN":["2077-0383"],"issn-type":[{"value":"2077-0383","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,10,21]]}}}