Quantitative assessment of flow velocity-estimation algorithms for optical Doppler tomography imaging

Area of Science:

• Medical imaging
• Biomedical engineering
• Fluid dynamics

Background:

• Accurate blood flow velocity estimation is crucial for diagnosing and monitoring various cardiovascular conditions.
• Existing velocity estimation algorithms, including centroid and correlation techniques, have limitations in accuracy and noise sensitivity.

Purpose of the Study:

• To quantitatively compare the performance of three categories of velocity estimation algorithms: centroid techniques, sliding-window filtering, and correlation techniques.
• To introduce and evaluate two novel algorithms: weighted centroid and sliding-window filtering.
• To assess the accuracy and robustness of these algorithms using both simulated and in vivo data.

Main Methods:

• Quantitative comparison of five velocity estimation algorithms: adaptive centroid, weighted centroid, sliding-window filtering, autocorrelation, and cross-correlation.
• Development and implementation of weighted centroid and sliding-window filtering techniques.
• Validation using simulated data and in vivo blood flow measurements.

Main Results:

• The sliding-window filtering technique exhibited higher velocity estimation accuracy compared to centroid and correlation methods.
• The sliding-window filtering technique demonstrated superior robustness to noise.
• Weighted centroid and sliding-window filtering were introduced as new algorithms.

Conclusions:

• The sliding-window filtering technique is the most accurate and robust method for blood flow velocity estimation among the evaluated algorithms.
• This finding has implications for improving the diagnostic capabilities of ultrasound and other medical imaging modalities.
• Further research may focus on optimizing the sliding-window filtering technique for specific clinical applications.