A robust local pulse wave imaging method based on digital image processing techniques.

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Medical Imaging

Background:

• The standard ln(D)U-loop method struggles to accurately measure ascending aortic pulse wave velocity (PWV) from low-quality ultrasound images, limiting clinical application.
• Factors like patient obesity, age, and operator skill can degrade image quality, hindering reliable PWV assessment.

Purpose of the Study:

• To develop and validate an improved diameter waveform extraction method for accurate PWV measurement from poor-quality ultrasound images.
• To enhance the clinical utility of the ln(D)U-loop method for ascending aortic PWV assessment.

Main Methods:

• A novel diameter waveform extraction technique combining thresholding, gradient filtering, and center of gravity was developed.
• The systolic period's rising point identification was refined using an optimal average of two linear regression methods.
• The improved method was validated using in vivo data from elderly patients with cerebral infarction and healthy young adults.

Main Results:

• The improved method successfully extracted diameter waveforms from poor-quality images, yielding more accurate PWV measurements.
• Statistically significant differences in PWV were observed between elderly patients with cerebral infarction and healthy young adults (p < 0.01).
• Bland-Altman analysis showed good agreement between the improved ln(D)U-loop method and the QA-loop method, with a mean deviation of 0.07 m/s.

Conclusions:

• The enhanced ln(D)U-loop method effectively overcomes limitations associated with poor ultrasound image quality for ascending aortic PWV measurement.
• This improved technique holds promise for wider clinical adoption in assessing cardiovascular health via PWV.