Sparse Dictionary-Based Magnetic Resonance Superresolution Imaging with Joint Loss Function Learning

Table 5

The PSNR/SSIM value corresponding to the number of sample blocks of different sampling images (superresolution ratio 1 : 2).

Number of sample image blocks	Sample set 1	Sample set 2	Sample set 3	Sample set 4	Sample set 5
1000	31.98/0.8945	28.64/0.8735	31.74/0.9101	30.45/0.8997	30.10/0.9043
5000	32.15/0.9076	28.79/0.8804	31.94/0.9189	30.52/0.9021	30.23/0.9067
10000	32.20/0.9207	28.81/0.8861	31.96/0.9184	30.56/0.9076	30.28/0.9112
50000	32.29/0.9412	28.84/0.8922	31.99/0.9245	30.57/0.9132	30.31/0.9183
100000	32.24/0.9407	28.87/0.9011	32.03/0.9302	30.59/0.9145	30.34/0.9208
150000	32.31/0.9413	28.91/0.9003	32.04/0.9301	30.60/0.9166	30.37/0.9207
200000	32.26/0.9409	28.84/0.8937	32.04/0.9308	30.58/0.9152	30.35/0.9211
500000	32.32/0.9415	28.86/0.8986	32.01/0.9255	30.61/0.9105	30.35/0.9200