To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Harmonizing Chest Imaging and Cough Sound Analysis: A Multi-Modal Approach for Respiratory Disease Detection |
|
Author |
May Rashid, Hamada Nayel, Ahmed Taha |
| Citation |
Vol. 25 No. 1 pp. 231-240 |
|
Abstract |
In response to the challenges of the global respiratory illness specifically COVID-19 pandemic, this research presents a diagnostic model that integrates chest imaging and sound data for enhanced accuracy. Utilizing carefully curated datasets of chest X-rays, CT scans, and cough recordings, our model offers a comprehensive analysis of visual and auditory cues associated with the disease. The imaging dataset captures critical radiological nuances, enriched by detailed clinical features. Simultaneously, the sound dataset, comprising over 25,000 cough recordings, pioneers the integration of acoustic signatures into diagnostic methodologies. Our preprocessing pipeline employs advanced techniques, including image augmentation and Mel spectrogram transformations, ensuring model adaptability. The model architecture synergizes visual and auditory insights, culminating in a unified diagnostic capability that transcends individual modalities. This research contributes to global COVID-19 efforts by providing a nuanced and comprehensive diagnostic approach. By fusing visual and auditory insights, our model addresses the urgency and accuracy required in the face of the pandemic, offering a path towards more effective diagnostics. |
|
Keywords |
Multi-modal COVID-19 detection, cough sound analysis, deep learning |
|
URL |
|
|
Title |
Harmonizing Chest Imaging and Cough Sound Analysis: A Multi-Modal Approach for Respiratory Disease Detection |
|
Author |
May Rashid, Hamada Nayel, Ahmed Taha |
| Citation |
Vol. 25 No. 1 pp. 231-240 |
|
Abstract |
In response to the challenges of the global respiratory illness specifically COVID-19 pandemic, this research presents a diagnostic model that integrates chest imaging and sound data for enhanced accuracy. Utilizing carefully curated datasets of chest X-rays, CT scans, and cough recordings, our model offers a comprehensive analysis of visual and auditory cues associated with the disease. The imaging dataset captures critical radiological nuances, enriched by detailed clinical features. Simultaneously, the sound dataset, comprising over 25,000 cough recordings, pioneers the integration of acoustic signatures into diagnostic methodologies. Our preprocessing pipeline employs advanced techniques, including image augmentation and Mel spectrogram transformations, ensuring model adaptability. The model architecture synergizes visual and auditory insights, culminating in a unified diagnostic capability that transcends individual modalities. This research contributes to global COVID-19 efforts by providing a nuanced and comprehensive diagnostic approach. By fusing visual and auditory insights, our model addresses the urgency and accuracy required in the face of the pandemic, offering a path towards more effective diagnostics. |
|
Keywords |
Multi-modal COVID-19 detection, cough sound analysis, deep learning |
|
URL |
