Fused Audio Instance and Representation for Respiratory Disease Detection DOI Creative Commons
Tuan Truong, Matthias Lenga, Antoine Serrurier

et al.

Sensors, Journal Year: 2024, Volume and Issue: 24(19), P. 6176 - 6176

Published: Sept. 24, 2024

Audio-based classification techniques for body sounds have long been studied to aid in the diagnosis of respiratory diseases. While most research is centered on use coughs as main acoustic biomarker, other also potential detect Recent studies coronavirus disease 2019 (COVID-19) suggested that breath and speech sounds, addition cough, correlate with disease. Our study proposes fused audio instance representation (FAIR) a method detection. FAIR relies constructing joint feature vector from various represented waveform spectrogram form. We conduct experiments case COVID-19 detection by combining sounds. findings show self-attention combine extracted features breath, leads best performance an area under receiver operating characteristic curve (AUC) score 0.8658, sensitivity 0.8057, specificity 0.7958. Compared models trained solely spectrograms or waveforms, both representations results improved AUC score, demonstrating helps enrich outperforms only one representation. this focuses COVID-19, FAIR’s flexibility allows it multi-modal multi-instance many diagnostic applications, potentially leading more accurate diagnoses across wider range

Language: Английский

Automatic detection and prediction of COVID-19 in cough audio signals using coronavirus herd immunity optimizer algorithm DOI Creative Commons

G. Ayappan,

S. Anila

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 17, 2025

The global spread of COVID-19, particularly through cough symptoms, necessitates efficient diagnostic tools. COVID-19 patients exhibit unique sound patterns distinguishable from other respiratory conditions. This study proposes an advanced framework to detect and predict using deep learning audio signals. Audio data the COUGHVID dataset undergo preprocessing fuzzy gray level difference histogram equalization, followed by segmentation with a U-Net model. Key features are extracted via Zernike Moments (ZM) Gray Level Co-occurrence Matrix (GLCM). Enhanced Deep Neural Network (EDNN), tuned Coronavirus Herd Immunity Optimizer (CHIO), performs final prediction minimizing error metrics. Comparative simulation results reveal that proposed EDNN-CHIO model improves MSE 25.35% SMAPE 42.06% over conventional models like PSO, WOA, LSTM. approach demonstrates superior reduction, highlighting its potential for effective detection.

Language: Английский

Citations

0
COVID-19 Detection from Optimized Features of Breathing Audio Signals Using Explainable Ensemble Machine Learning DOI Creative Commons
Shahnaz Sultana, A. B. M. Aowlad Hossain, Jahangir Alam

et al.

Results in Control and Optimization, Journal Year: 2025, Volume and Issue: unknown, P. 100538 - 100538

Published: Feb. 1, 2025

Language: Английский

Citations

0
EO-LGBM-HAR: A novel meta-heuristic hybrid model for human activity recognition DOI
E. Topuz, Yasin Kaya

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 189, P. 110004 - 110004

Published: March 17, 2025

Language: Английский

Citations

0
Fused Audio Instance and Representation for Respiratory Disease Detection DOI Creative Commons
Tuan Truong, Matthias Lenga, Antoine Serrurier

et al.

Sensors, Journal Year: 2024, Volume and Issue: 24(19), P. 6176 - 6176

Published: Sept. 24, 2024

Audio-based classification techniques for body sounds have long been studied to aid in the diagnosis of respiratory diseases. While most research is centered on use coughs as main acoustic biomarker, other also potential detect Recent studies coronavirus disease 2019 (COVID-19) suggested that breath and speech sounds, addition cough, correlate with disease. Our study proposes fused audio instance representation (FAIR) a method detection. FAIR relies constructing joint feature vector from various represented waveform spectrogram form. We conduct experiments case COVID-19 detection by combining sounds. findings show self-attention combine extracted features breath, leads best performance an area under receiver operating characteristic curve (AUC) score 0.8658, sensitivity 0.8057, specificity 0.7958. Compared models trained solely spectrograms or waveforms, both representations results improved AUC score, demonstrating helps enrich outperforms only one representation. this focuses COVID-19, FAIR’s flexibility allows it multi-modal multi-instance many diagnostic applications, potentially leading more accurate diagnoses across wider range

Language: Английский

Citations

0