Automatic Classification of Respiratory Sounds Based on Convolutional Neural Network with Multi Images

Minami, Koki; Lu, Huimin; Kamiya, Tohru; Mabu, Shingo; Shoji, K.

doi:10.1145/3436349.3436365

Cited by 5 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The resolution of scalograms comes at the cost of increased feature size, especially for audio data, limiting its applications. To reduce the size of generated scalograms, we used colored (RGB) images as an encoding for scalograms to train TBscreen (64,65), making it easier to use medium-sized deep learning models for training. We could then leverage pretrained image classification models reducing the need for a very large training dataset.…”

Section: Discussionmentioning

confidence: 99%

TBscreen: A passive cough classifier for tuberculosis screening with a controlled dataset

Sharma,

Nduba,

Njagi

et al. 2024

Sci. Adv.

View full text Add to dashboard Cite

Recent respiratory disease screening studies suggest promising performance of cough classifiers, but potential biases in model training and dataset quality preclude robust conclusions. To examine tuberculosis (TB) cough diagnostic features, we enrolled subjects with pulmonary TB ( N = 149) and controls with other respiratory illnesses ( N = 46) in Nairobi. We collected a dataset with 33,000 passive coughs and 1600 forced coughs in a controlled setting with similar demographics. We trained a ResNet18-based cough classifier using images of passive cough scalogram as input and obtained a fivefold cross-validation sensitivity of 0.70 (±0.11 SD). The smartphone-based model had better performance in subjects with higher bacterial load {receiver operating characteristic–area under the curve (ROC-AUC): 0.87 [95% confidence interval (CI): 0.87 to 0.88], P < 0.001} or lung cavities [ROC-AUC: 0.89 (95% CI: 0.88 to 0.89), P < 0.001]. Overall, our data suggest that passive cough features distinguish TB from non-TB subjects and are associated with bacterial burden and disease severity.

show abstract

Section: Discussionmentioning

confidence: 99%

TBscreen: A passive cough classifier for tuberculosis screening with a controlled dataset

Sharma,

Nduba,

Njagi

et al. 2024

Sci. Adv.

View full text Add to dashboard Cite

show abstract

“…In our previous work [14], we combined HMM and a deep neural network architecture to build a classification model for respiratory sounds. Some studies focused on transfer learning strategies by utilizing the pretrained VGG16 model on image datasets [15] [16] [17]. More complex CNN-based systems with a model tuning VOLUME 4, 2016 strategy and a data augmentation technique were proposed to obtain better performance in the respiratory sound classification task [18] [19] [20].…”

Section: Introductionmentioning

confidence: 99%

“…17 database, which are abbreviated as ResNet-backbone, SE-ResNet, FB-SE-ResNet, and FBQ-SE-ResNet. ResNetbackbone is the end-to-end system on the ResNet-9 network with four classes as outputs.…”

mentioning

confidence: 99%

Respiratory Sound Classification: From Fluid-Solid Coupling Analysis to Feature-Band Attention

et al. 2022

View full text Add to dashboard Cite

Based on respiratory sound production mechanisms, we study the relationship between airflow characteristics in the bronchi and the sound pressure spectrum curves to implement an end-toend respiratory sound classification system with a feature-band attention module. First, we analyse fluidsolid coupling simulations of the bronchi and execute acoustic simulations to obtain spectrum curves of the bronchi at the sound pressure level. Then, based on the spectrum characteristics of the bronchi, we propose an attention strategy to refine acoustic features with adaptive weights. In addition, we introduce a featureband attention module to ResNet-based networks with a squeeze-and-excitation block. Finally, we perform experiments on the ICBHI public database to classify respiratory sounds as belonging to one of four classes: normal, wheezes, crackles, and both (wheezes and crackles). The results show that our proposed system produces superior performance compared to the baseline system. This type of feature learning strategy is useful for exploring distinct characteristics of different types of respiratory sounds.INDEX TERMS fluid-solid coupling, attention learning, end-to-end system, respiratory sound classification, squeeze-and-excitation.

show abstract

Lung disease detection using EasyNet

Hassan,

Singhal,

Chaudhary

2024

Biomedical Signal Processing and Control

View full text Add to dashboard Cite

Automatic Classification of Respiratory Sounds Based on Convolutional Neural Network with Multi Images

Cited by 5 publications

References 12 publications

TBscreen: A passive cough classifier for tuberculosis screening with a controlled dataset

TBscreen: A passive cough classifier for tuberculosis screening with a controlled dataset

Respiratory Sound Classification: From Fluid-Solid Coupling Analysis to Feature-Band Attention

Lung disease detection using EasyNet

Contact Info

Product

Resources

About