Automated detection of cine-SPAMM magnetic resonance tags at sub-pixel resolutions

Quantitative analysis of cardiac motion using tagged magnetic resonance (MR) images has ultimately been limited by the spatial resolutions of the tagging techniques. A new image processing technique that automatically identifies MR tags to subpixel accuracy in tagged cine-SPAMM magnetic resonance images is presented. The automated subpixel tag detection technique allows regional cardiac wall motion to be noninvasively evaluated in an accurate, fast, and reproducible manner. Simulation results suggest that tag detection accuracy of one-half to one-quarter of a pixel may be possible in typical clinical images