Table 3.
Differences between the training and test dataset.
| Training T2w | Test T2w | Training DWI | Test DWI | |
|---|---|---|---|---|
| Slice thickness (mm) | 3.3 ± 0.5 | 3.1 ± 0.2 | 5.1 ± 1.1 | 5.0 ± 0.0 |
| Repetition time (ms) | 4994.9 ± 1913.5 | 3971.5 ± 708.1 | 6463.1 ± 1539.1 | 4121.9 ± 562.7 |
| Pixel bandwidth (Hz) | 215.9 ± 48.9 | 201.6 ± 8.8 | 1792.5 ± 351.6 | 1735.0 ± 8.3 |
| Flip angle (°) | 137.0 ± 16.5 | 148.7 ± 5.6 | 90.0 ± 0.0 | 90.0 ± 0.0 |
| Echo time (ms) | 97.8 ± 14.5 | 101.9 ± 4.7 | 67.2 ± 13.4 | 73.2 ± 2.2 |
| Field strength (T) | 2.5 ± 0.7 | 3.0 ± 0.0 | 2.5 ± 0.7 | 3.0 ± 0.0 |
| In-plane resolution (mm) | 0.6 ± 0.2 | 0.6 ± 0.0 | 1.6 ± 0.4 | 2.0 ± 0.1 |
Each column shows the mean and standard deviation of the different acquisition parameters of the T2w and DW sequences for the training and test cohort, respectively. Only the flip angle (90°) and pixel bandwidth of DWI were not significantly different between cohorts.