Lung Sound Recognition Algorithm Based on VGGish-BiGRU

Abstract

Pulmonary breathing sound plays a key role in the prevention and diagnosis of the lung diseases. Its correlation with pathology and physiology has become an important research topic in the pulmonary acoustics and the clinical medicine. However, it is difficult to fully describe lung sound information with the traditional features because lung sounds are complex and nonstationary signals. And the traditional convolutional neural network cannot also extract the temporal features of the lung sounds. To solve the problem, a lung sound recognition algorithm based on VGGish-BiGRU is proposed on the basis of transfer learning, which combines VGGish network with the bidirectional gated recurrent unit neural network (BiGRU). In the proposed algorithm, VGGish network is pretrained using audio set, and the parameters are transferred to VGGish network layer of the target network. The temporal features of the lung sounds are extracted through retraining BiGRU network with the lung sound data. During retraining BiGRU network, the parameters in VGGish layers are frozen, and the parameters of BiGRU network are fine-tuned. The experimental results show that the proposed algorithm effectively improves the recognition accuracy of the lung sounds in contrast with the state-of-the-art algorithms, especially the recognition accuracy of asthma.INDEX TERMS BiGRU, lung sound recognition, Mel spectrogram, transfer learning, VGGish.

Recognition of pulmonary diseases from lung sounds using convolutional neural networks and long short-term memory

“…( 2018 ) 15 2141 7: Normal, monophonic wheeze polyphonic wheeze, stridor squawk, fine crackle, coarse crackle Spectrograms CNN Accuracy: 95.56 Sensitivity: N/A Specificity: N/A Shi et al. ( 2019 ) 384 1152 3: Normal, asthma, pneumonia Spectrograms VGG-BDGRU Accuracy: 87.41 Sensitivity: N/A Specificity: N/A Demir et al. ( 2020 ) 126 6898 4: Normal, crackles, wheezes crackles+wheezes Spectrograms CNN Accuracy: 71.15 Sensitivity: 61.00 Specificity: 86.00 García-Ordás et al.…”

“…In addition, the diagnosis of pulmonary diseases is usually affected by the quality of the tool, physician experience, and surrounding environment (Shi et al. 2019 ). Therefore, electronic stethoscope has been gradually arising as a replacement to traditional diagnosis tools.…”

“…It has the ability to store lung sounds as signals within a computer; allowing medical doctors to investigate these signals in time-frequency analysis with a better interpretation (Shi et al. 2019 ; Gurung et al. 2011 ).…”

CNN-MoE Based Framework for Classification of Respiratory Anomalies and Lung Disease Detection

“…Feature extraction in state-of-the-art deep learning based systems typically involves generating twodimensional time-frequency spectrograms that are able to capture both fine grained temporal and spectral information as well as present a much wider time context than single frame analysis. While a variety of spectrogram transformations have been utilised, Mel-based methods such as log-Mel spectra [19], [20], [21] and stacked MFCC features [19], [22], [23], [24], [25], [26] are the most popular ones. Some researchers combined different types of spectrogram, e.g.…”

“…A more serious issue with this research field has been the difficulty of comparing between techniques due to the lack of standardised datasets for evaluation. Most publications evaluate on proprietary datasets that are unavailable to others [9], [10], [13], [19], [25].…”