Evaluation of methods for bolus arrival time determination using a four-dimensional MRA flow phantom

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Radiology

Background:

• Accurate determination of bolus arrival time (BAT) is crucial for analyzing hemodynamic characteristics in 4D MR angiography (MRA).
• Existing methods for BAT extraction often lack a gold standard for evaluation, necessitating the creation of datasets with known BAT values.
• Spatiotemporal 4D MRA is used to analyze cerebral vessel anomalies.

Purpose of the Study:

• To evaluate and compare different methods for extracting bolus arrival time (BAT).
• To introduce a novel 4D flow phantom for simulating 4D MRA datasets with known BAT values.
• To assess the performance of model-dependent versus model-independent BAT extraction techniques.

Main Methods:

• A 4D flow phantom was generated using a synthetic 3D MRA dataset with defined BAT values.
• Voxel-by-voxel concentration-time curves were computed using the gamma-variate function to create simulated 4D MRA data.
• Partial volume effects and Gaussian noise were integrated into the simulated dataset. Three state-of-the-art BAT extraction methods were applied and compared against known values.

Main Results:

• The simulated 4D MRA dataset was visually comparable to clinical data.
• Model-dependent methods for BAT extraction demonstrated higher accuracy compared to the model-independent method.
• Performance was evaluated across phantom datasets with varying vessel diameters and signal-to-noise ratios.

Conclusions:

• The developed 4D flow phantom provides a valuable tool for evaluating BAT extraction methods in 4D MRA.
• Model-dependent approaches are recommended for more accurate bolus arrival time determination in 4D MRA.
• Further research can utilize this phantom to refine and validate other quantitative imaging analysis techniques.