TABLE III.
Assessment of Variability of Tissue Classification From Neonate MRI. The Predictive value for Each Type of Tissue as Determined From the STAPLE Estimates, and the Mean Predictive Value, Are Reported for Each of Eight Segmentations. The Mean Predictive Value May Be Used to Rank the Quality of the Segmentations. Tissue Types Are Background (B), Extra-Cranial Tissue (ECT), Cortical Gray Matter (CGM), Cerebrospinal Fluid (CSF), Myelinated White Matter (MWM), Unmyelinated White Matter (UWM), and Subcortical Gray Matter (SCG)
| Segmentation | B | ECT | COM | CSF | MWM | UWM | SCO | mean PV |
|---|---|---|---|---|---|---|---|---|
| 1 | 0.9999 | 0.6219 | 0.8013 | 0.7035 | 0.3648 | 0.9403 | 0.9097 | 0.7631 |
| 2 | 1.000 | 1.000 | 0.9990 | 1.000 | 0.6336 | 0.9456 | 0.9990 | 0.9396 |
| 3 | 1.000 | 1.000 | 0.9990 | 0.9870 | 0.6339 | 0.9888 | 0.9967 | 0.9436 |
| 4 | 1.000 | 0.8250 | 0.6493 | 0.8853 | 0.4974 | 0.9998 | 0.8076 | 0.8092 |
| 5 | 1.000 | 0.9586 | 0.9816 | 1000 | 1.000 | 0.9967 | 0.9440 | 0.9830 |
| 6 | 1.000 | 1.000 | 0.9997 | 1.000 | 0.9997 | 0.9846 | 0.9692 | 0.9933 |
| 7 | 1.000 | 1.000 | 0.9186 | 1.000 | 0.9997 | 0.9977 | 0.5000 | 0.9166 |
| 8 | 1.000 | 1.000 | 0.7913 | 1.000 | 0.9930 | 0.9863 | 0.9889 | 0.9656 |