Regional deep atrophy : Using temporal information to automatically identify regions associated with Alzheimer’s disease progression from longitudinal MRI
Dong, Mengjin ; Xie, Long ; Das, Sandhitsu R. ; Wang, Jiancong ; Wisse, Laura E.M. LU
; deFlores, Robin
; Wolk, David A.
and Yushkevich, Paul A.
(2024)
In Imaging Neuroscience
2.
p.1-23
- Abstract
Longitudinal assessment of brain atrophy, particularly in the hippocampus, is a well-studied biomarker for neurodegenerative diseases, such as Alzheimer’s disease (AD). Estimating brain progression patterns can be applied to understanding the therapeutic effects of amyloid-clearing drugs in research and detecting the earliest sign of accelerated atrophy in clinical settings. However, most state-of-the-art measurements calculate changes directly by segmentation and/or deformable registration of MRI images, and may misreport head motion or MRI artifacts as neurodegeneration, impacting their accuracy. In our previous study, we developed a deep learning method DeepAtrophy that uses a convolutional neural network to quantify differences... (More)
Longitudinal assessment of brain atrophy, particularly in the hippocampus, is a well-studied biomarker for neurodegenerative diseases, such as Alzheimer’s disease (AD). Estimating brain progression patterns can be applied to understanding the therapeutic effects of amyloid-clearing drugs in research and detecting the earliest sign of accelerated atrophy in clinical settings. However, most state-of-the-art measurements calculate changes directly by segmentation and/or deformable registration of MRI images, and may misreport head motion or MRI artifacts as neurodegeneration, impacting their accuracy. In our previous study, we developed a deep learning method DeepAtrophy that uses a convolutional neural network to quantify differences between longitudinal MRI scan pairs that are associated with time. DeepAtrophy has high accuracy in inferring temporal information from longitudinal MRI scans, such as temporal order or relative interscan interval. DeepAtrophy also provides an overall atrophy score that was shown to perform well as a potential biomarker of disease progression and treatment efficacy. However, DeepAtrophy is not interpretable, and it is unclear what changes in the MRI contribute to progression measurements. In this paper, we propose Regional Deep Atrophy (RDA), which combines the temporal inference approach from DeepAtrophy with a deformable registration neural network and attention mechanism that highlights regions in the MRI image where longitudinal changes are contributing to temporal inference. RDA has similar prediction accuracy as DeepAtrophy, but its additional interpretability makes it more acceptable for use in clinical settings, and may lead to more sensitive biomarkers for disease monitoring and progression understanding in preclinical AD.
(Less)
- author
- Dong, Mengjin
; Xie, Long
; Das, Sandhitsu R.
; Wang, Jiancong
; Wisse, Laura E.M.
LU
; deFlores, Robin
; Wolk, David A.
and Yushkevich, Paul A.
- organization
- publishing date
- 2024-09
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Alzheimer’s disease, attention maps, longitudinal analysis, medial temporal lobe, sensitive biomarkers, structural MRI
- in
- Imaging Neuroscience
- volume
- 2
- pages
- 23 pages
- publisher
- MIT Press
- external identifiers
-
- scopus:105007063071
- ISSN
- 2837-6056
- DOI
- 10.1162/imag_a_00294
- language
- English
- LU publication?
- yes
- id
- 0f69e8d0-1f8c-4882-86f0-2df345c667a2
- date added to LUP
- 2025-09-26 13:31:28
- date last changed
- 2025-09-26 13:31:48
@article{0f69e8d0-1f8c-4882-86f0-2df345c667a2,
abstract = {{<p>Longitudinal assessment of brain atrophy, particularly in the hippocampus, is a well-studied biomarker for neurodegenerative diseases, such as Alzheimer’s disease (AD). Estimating brain progression patterns can be applied to understanding the therapeutic effects of amyloid-clearing drugs in research and detecting the earliest sign of accelerated atrophy in clinical settings. However, most state-of-the-art measurements calculate changes directly by segmentation and/or deformable registration of MRI images, and may misreport head motion or MRI artifacts as neurodegeneration, impacting their accuracy. In our previous study, we developed a deep learning method DeepAtrophy that uses a convolutional neural network to quantify differences between longitudinal MRI scan pairs that are associated with time. DeepAtrophy has high accuracy in inferring temporal information from longitudinal MRI scans, such as temporal order or relative interscan interval. DeepAtrophy also provides an overall atrophy score that was shown to perform well as a potential biomarker of disease progression and treatment efficacy. However, DeepAtrophy is not interpretable, and it is unclear what changes in the MRI contribute to progression measurements. In this paper, we propose Regional Deep Atrophy (RDA), which combines the temporal inference approach from DeepAtrophy with a deformable registration neural network and attention mechanism that highlights regions in the MRI image where longitudinal changes are contributing to temporal inference. RDA has similar prediction accuracy as DeepAtrophy, but its additional interpretability makes it more acceptable for use in clinical settings, and may lead to more sensitive biomarkers for disease monitoring and progression understanding in preclinical AD.</p>}},
author = {{Dong, Mengjin and Xie, Long and Das, Sandhitsu R. and Wang, Jiancong and Wisse, Laura E.M. and deFlores, Robin and Wolk, David A. and Yushkevich, Paul A.}},
issn = {{2837-6056}},
keywords = {{Alzheimer’s disease; attention maps; longitudinal analysis; medial temporal lobe; sensitive biomarkers; structural MRI}},
language = {{eng}},
pages = {{1--23}},
publisher = {{MIT Press}},
series = {{Imaging Neuroscience}},
title = {{Regional deep atrophy : Using temporal information to automatically identify regions associated with Alzheimer’s disease progression from longitudinal MRI}},
url = {{http://dx.doi.org/10.1162/imag_a_00294}},
doi = {{10.1162/imag_a_00294}},
volume = {{2}},
year = {{2024}},
}