Improved blood velocity measurements with a hybrid image filtering and iterative Radon transform algorithm

Area of Science:

• Neuroimaging and Neuroscience
• Biomedical Engineering
• Physiology

Background:

• Neural activity induces hemodynamic changes detectable by functional magnetic resonance imaging (fMRI).
• Accurate measurement of blood flow, particularly velocity in individual vessels, is crucial for understanding these hemodynamic signals.
• Existing methods for determining blood velocity from space-time images, like the Radon transform, face trade-offs between speed and precision, and are susceptible to artifacts.

Purpose of the Study:

• To develop a more accurate and efficient method for measuring blood velocity from space-time image sequences.
• To overcome limitations of existing Radon transform-based velocity measurement techniques, including precision, speed, and susceptibility to image artifacts.
• To provide a robust algorithm for blood velocity estimation applicable to various imaging modalities and biological preparations.

Main Methods:

• Employed a Sobel filter for pre-processing space-time images to enhance signal quality and reduce noise.
• Utilized an iterative application of the Radon transform to improve the precision of blood velocity measurements.
• Developed an algorithm that does not require a priori angle information, allowing sensitivity to rapid blood flow changes.

Main Results:

• The combined Sobel filtering and iterative Radon transform approach significantly improved the accuracy of blood velocity measurements.
• Iterative Radon transforms provided increased precision while achieving an order of magnitude faster processing compared to traditional methods.
• The enhanced algorithm demonstrated robustness against common imaging artifacts and was effective on various space-time image datasets with red blood cell streaks.

Conclusions:

• Pre-processing with a Sobel filter and iterative Radon transforms offer a superior method for accurate and rapid blood velocity determination.
• This technique enhances the analysis of hemodynamic signals derived from functional magnetic resonance imaging (fMRI) and other line-scan imaging methods.
• The developed algorithm is broadly applicable to the study of blood flow dynamics in vasculature across different imaging conditions and biological systems.