Efficient ResNet-based Deep Learning Model forAsthma and COPD Detection Using Short-TimeFourier Transform Spectrograms

Abstract

This study proposes a deep learning-based framework for the automated detection of asthma and chronicobstructive pulmonary disease (COPD) using respiratory sound analysis. Breath sound recordings are preprocessedthrough resampling, silence trimming, and Wiener filtering to enhance signal quality. Short-Time Fourier Transform(STFT) is employed to convert audio signals into spectrogram representations, which are further augmented using timemasking, frequency masking, and time warping to improve model generalization. The proposed model utilizesEfficientNet with multi-scale feature fusion to capture both local and global patterns in respiratory sounds.Experimental results demonstrate that the proposed approach achieves superior performance, with an accuracy of99.21%, sensitivity of 99.56%, and specificity of 100%, outperforming existing CNN, ResNet, and SVM-basedmethods. The findings indicate that the proposed method is a reliable and efficient tool for non-invasive respiratorydisease diagnosis.