. Author manuscript; available in PMC: 2012 Jun 5.

Published in final edited form as: IEEE Trans Med Imaging. 2011 Oct 20;31(4):907–923. doi: 10.1109/TMI.2011.2172951

TABLE I.

Reconstruction Accuracy, Sharpness, and Noise Level Measurements for the Numerical Phantom Study

Amplitude fraction irregularity (%)	Motion rate^† (bpm)	View-angle sampling (view angles)	rRMSE (%)	Pseudo PSF width (mm)		Pseu do PSF shift (mm)		Noise standard deviation (10^–3mm^–1)
				Parallel to motion	Perpendicular to motion	Parallel to motion	Perpendicular to motion
0	64	15	5.64	1.21	0.39	0.45	0.21	8.3
10	64	15	5.94	1.26	0.33	0.63	0.24	8.4
20	64	15	6.90	1.51	0.17	0.57	0.23	8.4
30	64	15	7.64	1.70	0.48	0.77	0.08	8.4
40	64	15	9.29	2.37	0.94	0.23	0.03	8.3
50	64	15	10.5	2.64	0.77	1.54	0.84	8.4

^†

Assuming an acquisition time of 14 seconds.

Both the prior image and the reference images have noise standard deviations of 8.5 × 10^–3 mm^–1.

The ROI used for the rRMSE measurements and the line segments used to evaluate the pseudo PSF are shown in Fig. 2.