% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@INPROCEEDINGS{MalekiBalajoo:905315,
author = {Maleki Balajoo, Somayeh and Eickhoff, Simon and Kharabian,
Shahrzad and Plachti, Anna and Waite, Laura and
Hoffstaedter, Felix and Palomero-Gallagher, Nicola and
GENON, Sarah},
title = {{H}ippocampal {M}etabolic {S}ubregions in {H}ealthy {O}lder
and{T}heir {P}rofiles in {N}eurodegeneration},
reportid = {FZJ-2022-00586},
year = {2021},
abstract = {Hippocampus dysfunction is the hallmark of Alzheimer’s
pathology and is frequently investigated with FDG-PET
metabolism measurements. However, while metabolic changes
are a key aspect of Alzheimer’s disease (AD), different
hippocampus’ subregions with their specific metabolic
covariance (MC) networks haven’t been identified in
healthy populations. It is also unclear to what extent these
are affected by AD pathophysiology. As the hippocampus
portrays cytoarchitectural, connectional and functional
heterogeneity, heterogenous patterns of MC could be
expected, leading to hippocampal subregions being
differentially affected by AD pathology. We investigated MC
as correlations in metabolism between hippocampus and brain
voxels in a large cohort of healthy older participants
(n=362). To identify how the pattern of brain MC changes
spatially within the hippocampus, we used a data-driven
approach to cluster hippocampal voxels based on their whole
brain co-metabolism profile (Eickhoff, Yeo, $\&$ Genon,
2018). The stability of different parcellation levels was
measured using split-half cross-validation. We then examined
the whole brain co-metabolism profile of each subregion
using the general linear model. To examine whether the local
metabolism between the metabolically-identified subregions
in healthy older is influenced by AD pathology, we also
performed a two-way ANOVA in the healthy older and in a
cohort of ADNI patients (n=581) with the mean glucose uptake
value as a dependent variable and both the subregions and
diagnostic groups as factors. The ANOVA was followed by
post-hoc analyses to identify which particular group
differences are statistically significant while correcting
for multiple comparisons. The results were compared with
results of the same analysis using the structurally-defined
and widely used FreeSurfer’s subfields.A stable 5-clusters
parcellation could be identified which included an
Anterior-subiculum(Red), an Anterior-CA(Yellow), an
Intermediate-subregion(Pink), a Posterior-subiculum(Blue)
and a Posterior-CA(Green) subregions (Fig. 1-A). As
illustrated in Fig. 1-C, the Anterior-subiculum(Red)
subregion mainly relates to orbito-frontal and temporal
regions while the Intermediate-subregion (Pink) is a
transitional subregion towards the Posterior-subiculum(Blue)
subregion which has a wide pattern of cortical MC. The
Anterior-CA(Yellow) subregion mainly relates to the amygdala
while the Posterior-CA(Green) subregion mainly relates to
other subcortical structures (Fig. 1-C). For both
hippocampal parcellations, the two-way ANOVA revealed both
significant main and interaction effects. Nevertheless,
overall, the differentiation between CA subregions as
provided by FreeSurfer did not exhibit specific group
differences while the anterior-vs-posterior distinction
offered by our parcellation revealed specific group
differences, in particular in the early stages (Table 1).
Overall, our results suggest a MC based differentiation
within the hippocampus that follows the CAs vs Subiculum
differentiation known from local microstructure mapping
(Fig. 1B) and anatomical connectivity (Fig. 1D). The MC
patterns of the identified subregions suggested three main
networks relating the anterior subregions to orbitofrontal
and anterior temporal regions, the Posterior-CA(Green)
subregion to subcortical structures around the ventricles,
and the Posterior-subiculum(Blue) subregion to an extended
cortical pattern (Fig. 1C). These MC patterns converge with
the patterns of structural covariance previously shown in
healthy aging (Plachti et al., 2020)(Fig. 1E), as well as
the different factors of brain atrophy reported in AD (Zhang
et al., 2016)(Fig. 1F) reinforcing the relevance of the
derived differentiation in pathological aging. Finally, our
parcellation allows the identification of specific
diagnostic group differences in regional hippocampus
metabolism.},
month = {Jun},
date = {2021-06-21},
organization = {27th annual meeting of the
Organization for Human Brain Mapping,
Virtual (Germany), 21 Jun 2021 - 25 Jun
2021},
cin = {INM-7},
cid = {I:(DE-Juel1)INM-7-20090406},
pnm = {5252 - Brain Dysfunction and Plasticity (POF4-525) / 5251 -
Multilevel Brain Organization and Variability (POF4-525) /
5253 - Neuroimaging (POF4-525)},
pid = {G:(DE-HGF)POF4-5252 / G:(DE-HGF)POF4-5251 /
G:(DE-HGF)POF4-5253},
typ = {PUB:(DE-HGF)1},
url = {https://juser.fz-juelich.de/record/905315},
}