Atrophy patterns of deep gray matter nuclei in Alzheimer's disease and frontotemporal dementia
Atrophy patterns of deep gray matter nuclei in Alzheimer's disease and frontotemporal dementia
Background
While cortical atrophy has been widely studied in dementia, emerging evidence highlights the role of subcortical degeneration, particularly in deep gray matter structures such as the thalamus, claustrum, and basal nuclei, in both Alzheimer's disease (AD) and frontotemporal dementia (FTD). However, disease-specific subcortical patterns remain undercharacterized.
Objective
To quantify deep gray matter atrophy across the AD continuum (mild cognitive impairment and AD) and three FTD subtypes (bvFTD, svFTD, PNFA), and to assess longitudinal atrophy, cognitive associations, and diagnostic classification.
Methods
We applied a novel segmentation pipeline (sTHOMAS) to T1-weighted MRI data from 380 participants in the ADNI cohort and 274 participants in the FTLDNI cohort, with longitudinal follow-up available for 237 participants. Group differences were assessed using ANCOVA (adjusted for age and sex), followed by post hoc comparisons and effect size estimation (Cohen's d). Neuropsychological associations were examined using partial correlations. A hierarchical Random Forest model was trained to classify diagnostic groups.
Results
Pronounced atrophy was observed in the mediodorsal, anteroventral, pulvinar thalamic nuclei, and nucleus accumbens, and claustrum in both AD and FTD, but was significantly greater in bvFTD. Longitudinal analysis revealed the fastest progression in bvFTD. Classification achieved 96.8% accuracy (AUC = 0.99) for AD versus FTD and 77.7% accuracy (AUC = 0.83) for PNFA versus svFTD. Subcortical atrophy correlated to executive, language, and semantic deficits.
Conclusions
Atrophy in the mediodorsal, pulvinar, anteroventral thalamic nuclei, nucleus accumbens, and claustrum distinguishes AD from FTD and differentiates FTD subtypes. These subcortical structures represent promising biomarkers for diagnosis and monitoring of neurodegeneration.
Banerjee, Arit
9d9819b6-eb8b-4936-8b52-166b5a3d01f8
Yang, Fan
b35ab88a-3424-4643-a14b-c5e47d73d335
Dutta, Joyita
0272704d-c50b-4774-b0ee-743ead7f1702
Cacciola, Alberto
c2539984-5e43-4799-a5cc-03c26c3d7ef6
Hornberger, Michael
a48c1c63-422a-4c11-9a51-c7be0aa3026d
Saranathan, Manojkumar
fd70de06-b9e8-4b06-a24f-7f80b84d348e
Banerjee, Arit
9d9819b6-eb8b-4936-8b52-166b5a3d01f8
Yang, Fan
b35ab88a-3424-4643-a14b-c5e47d73d335
Dutta, Joyita
0272704d-c50b-4774-b0ee-743ead7f1702
Cacciola, Alberto
c2539984-5e43-4799-a5cc-03c26c3d7ef6
Hornberger, Michael
a48c1c63-422a-4c11-9a51-c7be0aa3026d
Saranathan, Manojkumar
fd70de06-b9e8-4b06-a24f-7f80b84d348e
Banerjee, Arit, Yang, Fan, Dutta, Joyita, Cacciola, Alberto, Hornberger, Michael and Saranathan, Manojkumar
(2025)
Atrophy patterns of deep gray matter nuclei in Alzheimer's disease and frontotemporal dementia.
Journal of Alzheimer's Disease.
(doi:10.1177/13872877251390386).
Abstract
Background
While cortical atrophy has been widely studied in dementia, emerging evidence highlights the role of subcortical degeneration, particularly in deep gray matter structures such as the thalamus, claustrum, and basal nuclei, in both Alzheimer's disease (AD) and frontotemporal dementia (FTD). However, disease-specific subcortical patterns remain undercharacterized.
Objective
To quantify deep gray matter atrophy across the AD continuum (mild cognitive impairment and AD) and three FTD subtypes (bvFTD, svFTD, PNFA), and to assess longitudinal atrophy, cognitive associations, and diagnostic classification.
Methods
We applied a novel segmentation pipeline (sTHOMAS) to T1-weighted MRI data from 380 participants in the ADNI cohort and 274 participants in the FTLDNI cohort, with longitudinal follow-up available for 237 participants. Group differences were assessed using ANCOVA (adjusted for age and sex), followed by post hoc comparisons and effect size estimation (Cohen's d). Neuropsychological associations were examined using partial correlations. A hierarchical Random Forest model was trained to classify diagnostic groups.
Results
Pronounced atrophy was observed in the mediodorsal, anteroventral, pulvinar thalamic nuclei, and nucleus accumbens, and claustrum in both AD and FTD, but was significantly greater in bvFTD. Longitudinal analysis revealed the fastest progression in bvFTD. Classification achieved 96.8% accuracy (AUC = 0.99) for AD versus FTD and 77.7% accuracy (AUC = 0.83) for PNFA versus svFTD. Subcortical atrophy correlated to executive, language, and semantic deficits.
Conclusions
Atrophy in the mediodorsal, pulvinar, anteroventral thalamic nuclei, nucleus accumbens, and claustrum distinguishes AD from FTD and differentiates FTD subtypes. These subcortical structures represent promising biomarkers for diagnosis and monitoring of neurodegeneration.
Text
JAD_AB_accepted
- Accepted Manuscript
More information
e-pub ahead of print date: 29 October 2025
Identifiers
Local EPrints ID: 507067
URI: http://eprints.soton.ac.uk/id/eprint/507067
ISSN: 1387-2877
PURE UUID: 4ff91c3f-a0ae-466f-ba78-5d07a868e5c1
Catalogue record
Date deposited: 26 Nov 2025 17:45
Last modified: 27 Nov 2025 03:14
Export record
Altmetrics
Contributors
Author:
Arit Banerjee
Author:
Fan Yang
Author:
Joyita Dutta
Author:
Alberto Cacciola
Author:
Michael Hornberger
Author:
Manojkumar Saranathan
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics