Table 2. Accuracy of the automatic atlas-based segmentation.
Region | Automatic-Manual 1 | Automatic-Manual 2 | Interobserver |
Frontal cortex | 0.8946 | 0.9003 | 0.8993 |
Medial frontal cortex | 0.9049 | 0.8754 | 0.8442 |
Cingulate cortex | 0.9494 | 0.8891 | 0.8737 |
Piriform cortex | 0.7576 | 0.7874 | 0.7984 |
Entorhinal cortex | 0.8945 | 0.8098 | 0.8681 |
Parietal cortex | 0.8813 | 0.8847 | 0.8926 |
Occipital cortex | 0.9437 | 0.8175 | 0.8386 |
Insular cortex | 0.7319 | 0.7366 | 0.7528 |
Temporal cortex | 0.9849 | 0.8737 | 0.8797 |
External capsule | 0.9221 | 0.7148 | 0.7051 |
Internal capsule | 0.9522 | 0.8365 | 0.8936 |
Corpus callosum | 0.9780 | 0.8647 | 0.8895 |
Anterior commissure | 0.8469 | 0.7109 | 0.7856 |
Periventricular white matter | 0.9788 | 0.7666 | 0.7735 |
Subcortical white matter | 0.7476 | 0.8359 | 0.8231 |
Corona radiata | 0.8873 | 0.7447 | 0.7547 |
Fimbria of hippocampus | 0.7903 | 0.7452 | 0.7191 |
Fornix | 0.8093 | 0.7286 | 0.7547 |
Claustrum | 0.9564 | 0.8845 | 0.8366 |
Caudate nucleus | 0.9842 | 0.9399 | 0.9513 |
Thalamus | 0.9729 | 0.9012 | 0.8900 |
Hippocampus | 0.9447 | 0.7755 | 0.8575 |
Amygdala | 0.8851 | 0.8364 | 0.8577 |
Hypothalamus | 0.7051 | 0.7695 | 0.7516 |
Lenticular nucleus | 0.9535 | 0.7679 | 0.8513 |
Olfactory bulb | 0.6951 | 0.7652 | 0.8916 |
Cerebellar hemispheres | 0.9458 | 0.7679 | 0.8513 |
Vermis | 0.9580 | 0.9112 | 0.9166 |
Basal forebrain | 0.8664 | 0.7403 | 0.7305 |
Forebrain | 0.7573 | 0.7715 | 0.7481 |
Diencephalon | 0.7741 | 0.8231 | 0.8970 |
Mesencephalon | 0.9636 | 0.9251 | 0.9336 |
Pons | 0.9316 | 0.9255 | 0.9207 |
Medulla oblongata | 0.9995 | 0.9368 | 0.9532 |
Septum | 0.9777 | 0.9296 | 0.9074 |
Dice coefficient between the manually delineated brain regions and the brain regions identified by the automatic atlas-based segmentation, and between the manual delineations performed by two different observers.