Table 2.
Performance ranking of four sequences acquired using a birdcage coil with N3 correction
| Permutation test (P values) Birdcage with N3 correction
| ||||
|---|---|---|---|---|
| SPGR | MP-RAGE | IR-SPGR | SYN | |
| SPGR | 0.001 (SPGR) | <0.0001 (SPGR) | <0.0001 (SPGR) | |
| MP-RAGE | 0.264 (MP-RAGE) | 0.0081 (MP-RAGE) | ||
| 0.443 (IR-SPGR) | ||||
| IR-SPGR | 0.016 (IR-SPGR) | |||
| SYN | ||||
For all tables, sequence types are listed in the order of performance; the P values are the significance levels in favor of the alternative hypothesis that the sequence in parenthesis has a lower deviation (for example, IR-SPGR has a lower deviation than Synthetic T1 with a significant P value of 0.016). In other words, when the P value is significant, the sequence in parentheses is the one that performs best. Both P values are reported if two pulse sequences are statistically indistinguishable: so, for example, two sequences appear in the MP-RAGE vs. IR-SPGR box. This is because we tested the hypotheses both ways (that is, H1 = MP-RAGE shows less variance than IR-SPGR; and then again, H1 = IR-SPGR shows less variance than MP-RAGE). Note that SPGR in the tables actually refers to SPGR for GE acquisitions and FLASH for Siemens acquisitions.