Factors influencing the accuracy and precision of velocity‐encoded phase imaging

Velocity-encoded phase contrast imaging is being used increasingly in clinical imaging for quantization of blood flow. In this study, the accuracy and precision of ascending aorta flow measurements were found to depend on several subtle aspects of the scan prescription and image analysis. While the usual scan parameters such as TR, TE, and flip angle gave incremental changes in the flow measurements, four additional factors that had a much greater effect on the measurements were identified. These factors were (1) the zero velocity (background) pixel value, (2) the size and shape of the vessel region of interest, (3) the maximum velocity encoded in diastole, and (4) the temporal resolution. Statistical analysis was done on a total of 48 scans on nine normal subjects to confirm the significance of the measured differences using the various choices for each of these factors. These factors must be considered if accurate and precise measurements of blood flow are desired. Estimates of accuracy and precision suggest that quantitative flow measurements from velocity-encoded MR imaging can be clinically useful.