TABLE 2. Comparisons of the RMSE (Mean and Standard Deviation) for the Methods With and Without (, Using Only the Short-Axis (SA) or Long-Axis (LA) View) Track-to-Track Fusion. The RMSE is Computed Separately by Projecting the State Estimates Onto Short (First Row) and Long (Second Row) Axis Tagged CMR Images and Comparing them Against Harmonic Phase Estimates. The -Values for Unpaired T-Tests are also Reported. The Proposed Method Yielded Better Estimates Than the Methods Without Fusion in Comparisons to Both Short and Long Axes Tagged CMR Sequences.

	Proposed Method	SA only	LA only	Proposed vs SA	Proposed vs LA
RMSE (evaluated on SA tags)	1.6 ± 0.5 mm	1.8 ± 0.7 mm	2.2 ± 0.6 mm
RMSE (evaluated on LA tags)	2.4 ± 0.6 mm	3.2 ± 0.9 mm	2.5 ± 0.6 mm