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

TitleComparison of human septal nuclei MRI measurements using automated segmentation and a new manual protocol based on histology.
Publication TypeJournal Article
Year of Publication2014
AuthorsButler T, Zaborszky L, Pirraglia E, Li J, Wang XHugh, Li Y, Tsui W, Talos D, Devinsky O, Kuchna I, Nowicki K, French J, Kuzniecky R, Wegiel J, Glodzik L, Rusinek H, deLeon MJ, Thesen T
JournalNeuroimage
Volume97
Pagination245-51
Date Published2014 Aug 15
ISSN1095-9572
KeywordsAdolescent, Adult, Automation, Brain Mapping, Epilepsy, Temporal Lobe, Female, Humans, Image Processing, Computer-Assisted, Magnetic Resonance Imaging, Male, Middle Aged, Observer Variation, Septal Nuclei, Young Adult
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.

DOI10.1016/j.neuroimage.2014.04.026
Alternate JournalNeuroimage
PubMed ID24736183
PubMed Central IDPMC4180657
Grant ListR01 AG022374 / AG / NIA NIH HHS / United States
M01 RR000096 / RR / NCRR NIH HHS / United States
K23 NS057579 / NS / NINDS NIH HHS / United States
NS023945 / NS / NINDS NIH HHS / United States
M01RR0096 / RR / NCRR NIH HHS / United States
UL1 TR000038 / TR / NCATS NIH HHS / United States
R01AG022374 / AG / NIA NIH HHS / United States
R01AG12101 / AG / NIA NIH HHS / United States
P30 AG008051 / AG / NIA NIH HHS / United States
UL1TR000038 / TR / NCATS NIH HHS / United States
R01 AG012101 / AG / NIA NIH HHS / United States
P30AG008051 / AG / NIA NIH HHS / United States
R01 NS023945 / NS / NINDS NIH HHS / United States
NS057579 / NS / NINDS NIH HHS / United States
R24-MH 068855 / MH / NIMH NIH HHS / United States
R24 MH068855 / MH / NIMH NIH HHS / United States
Related Institute: 
Brain Health Imaging Institute (BHII)

Weill Cornell Medicine
Department of Radiology
525 East 68th Street New York, NY 10065