Past and Trends in Cough Sound Acquisition, Automatic Detection and Automatic Classification: A Comparative Review

Area of Science:

• Biomedical Engineering
• Digital Health
• Medical Informatics

Background:

• Cough is a prevalent symptom, frequently necessitating medical consultation.
• Effective patient care relies on accurate cough symptom recording and automated analysis.
• Existing research on automated cough analysis requires a comprehensive, quantitative overview.

Purpose of the Study:

• To conduct a quantitative review of cough sound acquisition, detection, and classification technologies.
• To characterize the state-of-the-art and identify emerging trends in automated cough analysis.
• To identify robust cough detection studies suitable for clinical implementation.

Main Methods:

• Systematic literature review and quantitative analysis of 144 selected studies.
• Extraction and processing of performance metrics from relevant research.
• Establishment of objective criteria for selecting high-quality detection studies.

Main Results:

• The field demonstrates significant dynamism, with increasing classification studies and dataset sizes.
• Smartphones and wearable sensors are prevalent for cough sound acquisition.
• Deep learning methods and COVID-19 related research show rapid expansion.
• Twelve detection studies were identified as most comprehensive for clinical deployment.

Conclusions:

• Automated cough analysis is rapidly advancing, driven by mobile health and deep learning.
• The COVID-19 pandemic has accelerated research in automated cough detection and diagnosis.
• The reviewed technologies show strong potential for integration into clinical practice and mobile health solutions.