Fig. 4 | Scientific Reports

Fig. 4

From: Performance of a GPU- and time-efficient pseudo-3D network for magnetic resonance image super-resolution and motion artifact reduction

Fig. 4

Comparison of TS-RCAN with other state-of-the-art 3D networks in terms of super resolution reconstruction. (a)–(d) comparison of metrics with two scale factors (\(2\times 2\times 1\) and \(2\times 2\times 2\)), ‘ + ’ represents the SRR with self-ensemble. TS-RCAN achieved comparable performance with MINet and outperformed the other networks. (e)-(j) Comparison of number of operations, GPU consumption and inference time to PSNR and SSIM with scale factor of \(2\times 2\times 2\). TS-RCAN achieved top performance by consuming the minimal computation resources and inference time. * p < 0.05 and ** p < 0.0005.

Back to article page