Format

Send to

Choose Destination
Neuroimage. 2014 Aug 15;97:245-51. doi: 10.1016/j.neuroimage.2014.04.026. Epub 2014 Apr 13.

Comparison of human septal nuclei MRI measurements using automated segmentation and a new manual protocol based on histology.

Author information

1
Center for Brain Health, Department of Psychiatry, New York University School of Medicine, 145 East 32nd Street, New York, NY 10016, USA; Comprehensive Epilepsy Center, Department of Neurology, New York University School of Medicine, 223 East 34th Street, New York, NY 10016, USA. Electronic address: Tracy.butler@nyumc.org.
2
Center for Molecular and Behavioral Neuroscience, Rutgers, The State University of New Jersey, 197 University Avenue, Newark, NJ 07102, USA.
3
Center for Brain Health, Department of Psychiatry, New York University School of Medicine, 145 East 32nd Street, New York, NY 10016, USA.
4
Comprehensive Epilepsy Center, Department of Neurology, New York University School of Medicine, 223 East 34th Street, New York, NY 10016, USA.
5
Institute for Basic Research in Developmental Disabilities, 1050 Forest Hill Road, Staten Island, NY 10314, USA.

Abstract

Septal nuclei, located in basal forebrain, are strongly connected with hippocampi and important in learning and memory, but have received limited research attention in human MRI studies. While probabilistic maps for estimating septal volume on MRI are now available, they have not been independently validated against manual tracing of MRI, typically considered the gold standard for delineating brain structures. We developed a protocol for manual tracing of the human septal region on MRI based on examination of neuroanatomical specimens. We applied this tracing protocol to T1 MRI scans (n=86) from subjects with temporal epilepsy and healthy controls to measure septal volume. To assess the inter-rater reliability of the protocol, a second tracer used the same protocol on 20 scans that were randomly selected from the 72 healthy controls. In addition to measuring septal volume, maximum septal thickness between the ventricles was measured and recorded. The same scans (n=86) were also analyzed using septal probabilistic maps and DARTEL toolbox in SPM. Results show that our manual tracing algorithm is reliable, and that septal volume measurements obtained via manual and automated methods correlate significantly with each other (p<.001). Both manual and automated methods detected significantly enlarged septal nuclei in patients with temporal lobe epilepsy in accord with a proposed compensatory neuroplastic process related to the strong connections between septal nuclei and hippocampi. Septal thickness, which was simple to measure with excellent inter-rater reliability, correlated well with both manual and automated septal volume, suggesting it could serve as an easy-to-measure surrogate for septal volume in future studies. Our results call attention to the important though understudied human septal region, confirm its enlargement in temporal lobe epilepsy, and provide a reliable new manual delineation protocol that will facilitate continued study of this critical region.

PMID:
24736183
PMCID:
PMC4180657
DOI:
10.1016/j.neuroimage.2014.04.026
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Elsevier Science Icon for PubMed Central
Loading ...
Support Center