Structural MRI-based premature brain ageing in epilepsy is associated with reduced processing speed
Poster No:
739
Submission Type:
Abstract Submission
Authors:
Heath Pardoe1, Chris Tailby1, Jodie Chapman1, Molly Ireland1, David Vaughan1, David Abbott1, Andrei Irimia2, Graeme Jackson1, for the Australian Epilepsy Project Investigators1
Institutions:
1The Florey Institute of Neuroscience and Mental Health, Melbourne, Australia, 2University of Southern California, Los Angeles, CA
First Author:
Co-Author(s):
for the Australian Epilepsy Project Investigators
The Florey Institute of Neuroscience and Mental Health
Melbourne, Australia
The Florey Institute of Neuroscience and Mental Health
Melbourne, Australia
Introduction:
Machine learning-based age prediction methods applied to structural MRI have shown increased brain predicted age difference (BrainPAD) in individuals with medically refractory epilepsy (Pardoe 2017). BrainPAD, which reflects structural MRI-based age relative to chronological age, is a marker for brain health. Processing speed and memory decline with increasing age, and deficits in these domains are commonly observed in individuals with epilepsy. In this study we investigated the association between BrainPAD and processing speed and memory performance in participants from the Australian Epilepsy Project (AEP).
Methods:
Participants aged 18-67 from four AEP cohorts were included in the study. These included individuals with (i) a 1st unprovoked seizure (N = 38, mean age 36 ± 14 years, 63% female), (ii) newly diagnosed epilepsy (N = 99, 36 ± 14 years, 53% female), (iii) drug resistant epilepsy (N = 156, 37 ± 11 years, 55% female) and (iv) healthy controls without epilepsy (N = 182, 40 ± 13 years, 66% female). Participants underwent structural T1w MRI scanning and cognitive assessments administered using AEP teleneuropsychology software. Processing speed and memory scores were derived by averaging z-scores across related cognitive tests, adjusted for age, sex and years of education.
Whole brain T1w MRI scans were analysed using the USC brain age prediction model (https://github.com/irimia-laboratory/USC_BA_estimator, Yin 2023), a deep learning approach utilizing the skull-stripped, spatially normalized brain.mgz file from the Freesurfer image processing pipeline. The age prediction model was trained on 16,213 cognitively normal participants. BrainPAD was calculated as the difference between the MRI-based age estimate and chronological age.
We modelled the relationship between BrainPAD (outcome variable) and mean z-scores for processing speed and memory using separate multiple linear regression models with cohort, age and sex as explanatory variables.
Whole brain T1w MRI scans were analysed using the USC brain age prediction model (https://github.com/irimia-laboratory/USC_BA_estimator, Yin 2023), a deep learning approach utilizing the skull-stripped, spatially normalized brain.mgz file from the Freesurfer image processing pipeline. The age prediction model was trained on 16,213 cognitively normal participants. BrainPAD was calculated as the difference between the MRI-based age estimate and chronological age.
We modelled the relationship between BrainPAD (outcome variable) and mean z-scores for processing speed and memory using separate multiple linear regression models with cohort, age and sex as explanatory variables.
Results:
Reduced processing speed was associated with increased BrainPAD (-0.97 years/unit change in z-score, p = 4.9 × 10-3), such that a one standard deviation decline in processing speed was associated with 1 year increase in brain age. No significant association was observed between BrainPAD and memory (p = 0.2). Increased BrainPAD was observed in medically refractory epilepsy cases relative to healthy controls (3.34 years, p = 3 × 10-7). Participants with a first seizure and newly diagnosed epilepsy also showed smaller but significant brain age increases (2.8 years, p = 8.7 × 10-3; 2.4 years, p = 6.4 × 10-3).
Conclusions:
Our findings demonstrate that increased BrainPAD, a proxy for accelerated ageing, is associated with reduced processing speed but not memory in individuals with epilepsy. This may be due to differing age-related trajectories of these cognitive domains; processing speed peaks between 20-30 years of age and steadily declines thereafter, whereas the most substantial decline in memory occur beyond around the age of 60, which is towards the upper limit of our cohort. Differential findings may also be due to differences in the neurobiological substrates underpinning these cognitive functions; processing speed is a diffuse process which may make it more sensitive to the global brain changes indexed by BrainPAD. Conversely, memory has classically been associated with medial temporal structures, and such local specificity might be diluted across the broader BrainPAD metric.
The USC age prediction model utilises a deep learning approach applied to minimally preprocessed MRI data, which contrasts with earlier machine-learning approaches that rely on volumetric feature extraction steps. The processing speed related differences observed using the USC model may therefore be driven by both gray and white matter changes.
Our study supports BrainPAD's utility as a brain health marker, linking neuroanatomical changes to cognitive decline in epilepsy.
The USC age prediction model utilises a deep learning approach applied to minimally preprocessed MRI data, which contrasts with earlier machine-learning approaches that rely on volumetric feature extraction steps. The processing speed related differences observed using the USC model may therefore be driven by both gray and white matter changes.
Our study supports BrainPAD's utility as a brain health marker, linking neuroanatomical changes to cognitive decline in epilepsy.
Higher Cognitive Functions:
Executive Function, Cognitive Control and Decision Making 1
Lifespan Development:
Aging 2
Keywords:
Aging
Cognition
Epilepsy
Machine Learning
Memory
STRUCTURAL MRI
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
Yin, C., et al., Anatomically interpretable deep learning of brain age captures domain-specific cognitive impairment. Proceedings of the National Academy of Sciences USA, 2023. 120(2): p. e2214634120.