PCA-based artifact removal algorithm for stroke detection using UWB radar imaging

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Signal Processing

Background:

• Rapid diagnosis and treatment are critical for stroke patients to minimize brain damage.
• Current diagnostic methods often involve delays, impacting patient outcomes.
• Transportable, low-cost diagnostic tools are needed for emergency medical services.

Purpose of the Study:

• To evaluate the effectiveness of a portable Ultra-Wideband (UWB) radar imaging system for rapid detection of acute brain lesions.
• To develop signal processing algorithms that ensure high performance with low computational complexity for a transportable device.
• To explore the potential of UWB radar imaging in improving emergency stroke care.

Main Methods:

• Development and testing of a transportable UWB radar imaging system.
• Application of Principal Component Analysis (PCA)-based preprocessing algorithms.
• Utilizing a computationally simple beamforming algorithm.
• Investigating a multistatic approach to enhance performance.

Main Results:

• PCA-based preprocessing achieved good performance with a simple beamforming algorithm.
• The PCA-based method outperformed existing state-of-the-art preprocessing techniques.
• A multistatic approach further improved performance with only a minor increase in computational complexity.

Conclusions:

• PCA-based preprocessing is a viable solution for UWB radar imaging in emergency stroke diagnostics.
• The developed system shows promise for rapid, on-site detection of acute brain lesions.
• This technology could significantly reduce treatment delays and improve outcomes for stroke patients.