JMIR Publications

ABSTRACT

Background:

Coughing is a common clinical symptom and a protective respiratory reflex closely associated with various respiratory system diseases. The acoustic characteristics of cough sounds are influenced by underlying pathological factors, with distinct acoustic signatures corresponding to different etiologies. Through rigorously analysis of these sounds, rapid identification and preliminary diagnosis of related conditions may be achieved. diseases. This approach holds great potential for broad application in mobile health (mHealth) and ubiquitous health (uHealth) platforms.

Objective:

Exploring the application of acoustic analysis of cough sounds in the diagnosis of respiratory diseases to enhance the diagnostic efficiency of healthcare professionals.

Methods:

In this study, we conducted extensive data collection, including voluntary cough audio recordings from patients diagnosed with respiratory diseases (such as chronic obstructive pulmonary disease, lung cancer, COVID-19, and pneumonia), as well as from healthy participants. A total of 2,610 audio samples were collected. We incorporated a Channel Attention Mechanism (CAM) into the final convolutional block of each residual block in the ResNet18 neural network, thereby constructing the CAM-ResNet18 neural network model. The recorded cough audio samples were converted into spectrograms to form the input dataset for model training. The CAM-ResNet18 model was trained on the training set of this dataset, with iterative parameter adjustments until convergence was achieved. Finally, spectrograms from the test set were fed into the pre-trained model for accurate classification of the cough-related conditions.

Results:

Experiments results on the collected audio dataset demonnstrate that the proposed CAM-ResNet18 model achieves an accuracy of 83.9% and an average F1 score of 82.52% in classifying these five types of cough sounds. In comparison, the traditional ResNet18 model achieves an accuracy of 78.16% and an average F1 score of 78.29%, indicating a clear performance improvement with the integration of the Channel Attention Mechanism.

Conclusions:

The experimental results validate the effectiveness of the proposed method, highlighting its significant potential for application in clinical diagnosis.