The aim of this paper was the detection of pathologies through respiratory sounds. The ICBHI (International Conference on Biomedical and Health Informatics) Benchmark was used. This dataset is composed of 920 sounds of which 810 are of chronic diseases, 75 of non-chronic diseases and only 35 of healthy individuals. As more than 88% of the samples of the dataset are from the same class (Chronic), the use of a Variational Convolutional Autoencoder was proposed to generate new labeled data and other well known oversampling techniques after determining that the dataset classes are unbalanced. Once the preprocessing step was carried out, a Convolutional Neural Network (CNN) was used to classify the respiratory sounds into healthy, chronic, and non-chronic disease. In addition, we carried out a more challenging classification trying to distinguish between the different types of pathologies or healthy: URTI, COPD, Bronchiectasis, Pneumonia, and Bronchiolitis. We achieved results up to 0.993 F-Score in the three-label classification and 0.990 F-Score in the more challenging six-class classification.

Detecting Respiratory Pathologies Using Convolutional Neural Networks and Variational Autoencoders for Unbalancing Data.

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.

/pdf/lung-sound-recognition-algorithm-based-on-vggish-bigru-3o2ql1t8q3.pdf

Lung Sound Recognition Algorithm Based on VGGish-BiGRU

The past 5 years have seen an explosion of interest in the use of artificial intelligence (AI) and machine learning techniques in medicine. This has been driven by the development of deep neural networks (DNNs)—complex networks residing in silico but loosely modelled on the human brain—that can process complex input data such as a chest radiograph image and output a classification such as ‘normal’ or ‘abnormal’. DNNs are ‘trained’ using large banks of images or other input data that have been assigned the correct labels. DNNs have shown the potential to equal or even surpass the accuracy of human experts in pattern recognition tasks such as interpreting medical images or biosignals. Within respiratory medicine, the main applications of AI and machine learning thus far have been the interpretation of thoracic imaging, lung pathology slides and physiological data such as pulmonary function tests. This article surveys progress in this area over the past 5 years, as well as highlighting the current limitations of AI and machine learning and the potential for future developments.

https://thorax.bmj.com/content/thoraxjnl/early/2020/05/14/thoraxjnl-2020-214556.full.pdf

Applications of artificial intelligence and machine learning in respiratory medicine

Automatic identification of respiratory diseases from stethoscopic lung sound signals using ensemble classifiers

Characterization and classification of asthmatic wheeze sounds according to severity level using spectral integrated features.

Multichannel lung sound analysis for asthma detection.

Lung sounds carry valuable information regarding the lungs pathology. The collaboration of advanced digital signal processing method and machine learning framework has immense potential to diagnose the lungs status using lung sound signals. In this paper, we have classified normal, asthma, and COPD subjects using their posterior lung sound signals, which was unexplored so far. Asthma and COPD diseases have common manifestations, for which the classification of these diseases in a single platform is like a conundrum. We have collected lung sound signals from 60 subjects (20 normal, 20 asthma, and 20 COPD) using a novel 4-channel data acquisition system. A feature extraction scheme is proposed to extract features from the subbands of the power spectral density (PSD), and fed to the artificial neural network (ANN) classifier for the 3-class classification. The changes in the posterior lung sound signals are targeted for feature extraction, which does not depend on the presence of wheeze or any other marker. The proposed multichannel based multiclass classification system achieves reasonable classification accuracy which is much higher than the theoretical and empirical chance levels, when the information of all the four channels are utilized together.

Classification of Normal, Asthma and COPD Subjects Using Multichannel Lung Sound Signals

Automatic lung sound cycle extraction from single and multichannel acoustic recordings

Automatic lung sound cycle extraction is the pivotal step for automated lung status detection as well as in monitoring the chronic lung diseases. In recent works, an attempt has been made to get rid of the additional airflow sensors due to their inaccuracy, patient's discomfort, and extra cost. In this paper, we have proposed a novel signal processing based approach to automatically extract lung sound cycles. In this framework, spectrograms are used as images to get the trend of the lung sound cycles without the need for identifying the corresponding respiratory phases or use of any reference airflow signal. We have utilized the lung sounds recorded from 8 healthy and 24 diseased subjects (8 subjects each from Asthma, COPD, and DPLD) to develop and evaluate the proposed lung sound cycle extraction algorithm. We employ a 4-fold cross-validation in our study and the average accuracy of 98.62% is found.

Md. Ariful Islam

Papers

Multichannel lung sound analysis for asthma detection.

Classification of Normal, Asthma and COPD Subjects Using Multichannel Lung Sound Signals

Automatic lung sound cycle extraction from single and multichannel acoustic recordings

A novel spectrogram based approach towards automatic lung sound cycle extraction