High-accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural network
IntroductionIn recent years, advancements in machine learning and electronic stethoscope technology have enabled high-precision recording and analysis of lung sounds, significantly enhancing pulmonary disease diagnosis.MethodsThis study presents a comprehensive approach to classify lung sounds into...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-07-01
|
| Series: | Frontiers in Bioengineering and Biotechnology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fbioe.2025.1583416/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849423291153907712 |
|---|---|
| author | Weiwei Zhang Xinyu Li Qiao Liu Xiangyang Zheng Yisu Ge Xiaotian Pan Xiaotian Pan Yu Zhou |
| author_facet | Weiwei Zhang Xinyu Li Qiao Liu Xiangyang Zheng Yisu Ge Xiaotian Pan Xiaotian Pan Yu Zhou |
| author_sort | Weiwei Zhang |
| collection | DOAJ |
| description | IntroductionIn recent years, advancements in machine learning and electronic stethoscope technology have enabled high-precision recording and analysis of lung sounds, significantly enhancing pulmonary disease diagnosis.MethodsThis study presents a comprehensive approach to classify lung sounds into healthy and unhealthy categories using a dataset collected from 112 subjects, comprising 35 healthy individuals and 77 patients with various pulmonary conditions, such as asthma, heart failure, pneumonia, bronchitis, pleural effusion, lung fibrosis, and chronic obstructive pulmonary disease (COPD), grouped as unhealthy. The dataset was obtained using a 3M Littmann® Electronic Stethoscope Model 3,200, employing three types of filters (Bell, Diaphragm, and Extended) to capture sounds across different frequency ranges. We extracted five key audio features—Spectral Centroid, Power, Energy, Zero Crossing Rate, and Mel-Frequency Cepstral Coefficients (MFCCs)—from each recording to form a feature matrix. A Multi-Layer Perceptron (MLP) neural network was trained for binary classification.ResultsThe MLP neural network achieved accuracies of 98%, 100%, and 94% on the training, validation, and testing sets, respectively. This partitioning ensured the model’s robustness and accuracy.DiscussionThe high classification accuracy achieved by the MLP neural network suggests that this approach is a valuable decision-support tool for identifying healthy versus unhealthy lung sounds in clinical settings, facilitating early intervention while maintaining computational efficiency for offline implementation. The combination of detailed feature extraction and an optimized MLP neural network resulted in a reliable method for automated binary classification of lung sounds. |
| format | Article |
| id | doaj-art-90858cb7f13944cd8452b3d2bdf715d1 |
| institution | Kabale University |
| issn | 2296-4185 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Bioengineering and Biotechnology |
| spelling | doaj-art-90858cb7f13944cd8452b3d2bdf715d12025-08-20T03:30:40ZengFrontiers Media S.A.Frontiers in Bioengineering and Biotechnology2296-41852025-07-011310.3389/fbioe.2025.15834161583416High-accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural networkWeiwei Zhang0Xinyu Li1Qiao Liu2Xiangyang Zheng3Yisu Ge4Xiaotian Pan5Xiaotian Pan6Yu Zhou7Infectious Disease Department, The Third Affiliated Hospital of Wenzhou Medical University, Wenzhou, ChinaChina Telecom Corporation Limited Zhejiang Branch, Hangzhou, ChinaInformation Technology Center, Wenzhou Medical University, Wenzhou, ChinaSchool of Data Science and Artificial Intelligence, Wenzhou University of Technology, Wenzhou, ChinaCollege of Computer Science and Artificial Intelligence, Wenzhou University, Wenzhou, ChinaInstitute of Intelligent Media Computing, Hangzhou Dianzi University, Hangzhou, ChinaShangyu Institute of Science and Engineering Co. Ltd., Hangzhou Dianzi University, Shaoxing, ChinaInfectious Disease Department, The Third Affiliated Hospital of Wenzhou Medical University, Wenzhou, ChinaIntroductionIn recent years, advancements in machine learning and electronic stethoscope technology have enabled high-precision recording and analysis of lung sounds, significantly enhancing pulmonary disease diagnosis.MethodsThis study presents a comprehensive approach to classify lung sounds into healthy and unhealthy categories using a dataset collected from 112 subjects, comprising 35 healthy individuals and 77 patients with various pulmonary conditions, such as asthma, heart failure, pneumonia, bronchitis, pleural effusion, lung fibrosis, and chronic obstructive pulmonary disease (COPD), grouped as unhealthy. The dataset was obtained using a 3M Littmann® Electronic Stethoscope Model 3,200, employing three types of filters (Bell, Diaphragm, and Extended) to capture sounds across different frequency ranges. We extracted five key audio features—Spectral Centroid, Power, Energy, Zero Crossing Rate, and Mel-Frequency Cepstral Coefficients (MFCCs)—from each recording to form a feature matrix. A Multi-Layer Perceptron (MLP) neural network was trained for binary classification.ResultsThe MLP neural network achieved accuracies of 98%, 100%, and 94% on the training, validation, and testing sets, respectively. This partitioning ensured the model’s robustness and accuracy.DiscussionThe high classification accuracy achieved by the MLP neural network suggests that this approach is a valuable decision-support tool for identifying healthy versus unhealthy lung sounds in clinical settings, facilitating early intervention while maintaining computational efficiency for offline implementation. The combination of detailed feature extraction and an optimized MLP neural network resulted in a reliable method for automated binary classification of lung sounds.https://www.frontiersin.org/articles/10.3389/fbioe.2025.1583416/fullmachine learningpulmonary disease classificationlung soundselectronic stethoscopemulti-layer perceptronfeature extraction |
| spellingShingle | Weiwei Zhang Xinyu Li Qiao Liu Xiangyang Zheng Yisu Ge Xiaotian Pan Xiaotian Pan Yu Zhou High-accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural network Frontiers in Bioengineering and Biotechnology machine learning pulmonary disease classification lung sounds electronic stethoscope multi-layer perceptron feature extraction |
| title | High-accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural network |
| title_full | High-accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural network |
| title_fullStr | High-accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural network |
| title_full_unstemmed | High-accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural network |
| title_short | High-accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural network |
| title_sort | high accuracy lung sound classification for healthy versus unhealthy diagnosis using artificial neural network |
| topic | machine learning pulmonary disease classification lung sounds electronic stethoscope multi-layer perceptron feature extraction |
| url | https://www.frontiersin.org/articles/10.3389/fbioe.2025.1583416/full |
| work_keys_str_mv | AT weiweizhang highaccuracylungsoundclassificationforhealthyversusunhealthydiagnosisusingartificialneuralnetwork AT xinyuli highaccuracylungsoundclassificationforhealthyversusunhealthydiagnosisusingartificialneuralnetwork AT qiaoliu highaccuracylungsoundclassificationforhealthyversusunhealthydiagnosisusingartificialneuralnetwork AT xiangyangzheng highaccuracylungsoundclassificationforhealthyversusunhealthydiagnosisusingartificialneuralnetwork AT yisuge highaccuracylungsoundclassificationforhealthyversusunhealthydiagnosisusingartificialneuralnetwork AT xiaotianpan highaccuracylungsoundclassificationforhealthyversusunhealthydiagnosisusingartificialneuralnetwork AT xiaotianpan highaccuracylungsoundclassificationforhealthyversusunhealthydiagnosisusingartificialneuralnetwork AT yuzhou highaccuracylungsoundclassificationforhealthyversusunhealthydiagnosisusingartificialneuralnetwork |