An Approach to Detect Chronic Obstructive Pulmonary Disease Using UWB Radar-Based Temporal and Spectral Features

Area of Science:

• Medical Technology
• Artificial Intelligence
• Pulmonology

Background:

• Chronic Obstructive Pulmonary Disease (COPD) is a leading global cause of mortality.
• Delayed diagnosis of COPD leads to irreversible lung damage and reduced quality of life.
• Current diagnostic methods have limitations in accuracy and accessibility.

Purpose of the Study:

• To develop an early and accurate detection method for COPD.
• To leverage artificial intelligence (AI) for improved COPD diagnosis.
• To explore the utility of respiration rate features for COPD detection.

Main Methods:

• A novel framework utilizing ultra-wideband (UWB) radar for non-invasive respiration data collection.
• Extraction of novel temporal and spectral features from UWB radar signals.
• Development and testing of machine learning and deep learning models using these features.

Main Results:

• Achieved 100% accuracy in detecting COPD patients.
• Demonstrated the effectiveness of UWB radar-based features for COPD identification.
• Validated performance using k-fold cross-validation and comparison with existing studies.

Conclusions:

• The proposed AI framework shows high potential for early and efficient COPD detection.
• Non-invasive UWB radar technology can significantly aid in identifying COPD patients.
• This approach may improve patient outcomes by enabling timely intervention.