Two Alternative Forced Choice Evaluation of Vessel Visibility Increases Due to Zero-filled Interpolation in MR Angiography

The effects of reconstructing low resolution 3D time-of-flight MR images using zero-filled interpolation were examined using observer performance measures. Two alternative forced choice experiments using digital subtraction angiography images as a truth standard were performed. MR angiography data were acquired from six patient volunteers and were reconstructed with and without zero-filled interpolation. The MR angiography images were evaluated by generating divergent maximum intensity projection images with the same view as digital subtraction angiography images obtained on the patients as part of their clinical workup. The projection images were evaluated by three observers. Zerofilled interpolation significantly increased the detectibility of very small vessels (diameters less than 0.7 mm) but had no significant effect on larger vessels.