LSTM-Based RNN Framework to Remove Motion Artifacts in Dynamic Multicontrast MR Images with Registration Model

Table 1

Effectiveness of proposed methods is compared with existing methods after taking averaged value from the ATDC dataset.

Schemes	Encoder-decoder	Conv LSTM	Multiscale	St	SSIM	PSNR	Accuracy
DC-CNN				0.7983	0.7956	26.967	0.9345
SMSR-CNN				0.8726	0.9011	27.754	0.9453
FMSI-CNN				0.9052	0.9415	26.863	0.9467
DRCA-CNN				0.9078	0.9465	27.543	0.9671
LSTM-based RNN				0.9756	0.9523	28.671	0.9824