Longitudinal FreeSurfer processing of UK Biobank two-timepoint data
Poster No:
1492
Submission Type:
Abstract Submission
Authors:
Rach Dawson1, Frederik Lange2, Fidel Alfaro Almagro3, Jesper Andersson4, Stephen Smith4
Institutions:
1University of Oxford, Oxford, Oxfordshire, UK, 2WiN FMRIB - University of Oxford, Oxford, Oxfordshire, 3WiN FMRIB - University of Oxford, Oxford, State/Province, 4University of Oxford, Oxford, Oxfordshire
First Author:
Co-Author(s):
Introduction:
Longitudinal analysis of brain morphology provides critical insights into structural changes associated with aging, neurodegeneration, and therapeutic interventions. Traditional single-timepoint (cross-sectional) analyses are prone to variability and bias, often limiting their sensitivity to subtle within-subject changes. FreeSurfer, a surface-based analysis tool (Dale, 1999) provides an unbiased (Reuter, 2011) longitudinal processing pipeline (Reuter, 2012) that aims to reduce variability by leveraging within-subject templates for longitudinal analyses. Utilising a subset of two-timepoint UK Biobank (UKB) data (Miller, 2016), here we aim to evaluate the performance of longitudinal FreeSurfer processing against multiple single-timepoint processing for use with all two-timepoint UKB data.
Methods:
The FreeSurfer longitudinal pipeline was applied to 500 UKB participants with two-timepoint structural MRI data. Each subject and timepoint was first processed independently, [CROSS], where full image segmentation and surface reconstruction were performed. Next, a within-subject template, [BASE], was created per subject by generating an unbiased image from both timepoints to represent the subject's average anatomy. Finally, each timepoint was longitudinal reprocessed, [LONG], using the [BASE] template and [CROSS] outputs.
This longitudinal approach aims to minimise non-aging-related, within-subject variability and reduce biases introduced by asymmetries in single-timepoint processing. Methods were compared using Absolute Symmetrised Percent Change (ASPC) (Reuter, 2012), with respect to volumetric measurements at the first and second timepoints, V1 and V2 respectively.
ASPC = 100 |V₂ - V₁| / (0.5 (V₁ + V₂)) (1)
We utilise ASPC to quantify the change in volumetric estimates across subcortical, cortical grey matter, and white matter structures. Lower ASPC values indicate greater reliability and reduced variability in volumetric measurements, however in longitudinal studies where true anatomical changes occur, ASPC reflects a combination of biological change and measurement error.
This longitudinal approach aims to minimise non-aging-related, within-subject variability and reduce biases introduced by asymmetries in single-timepoint processing. Methods were compared using Absolute Symmetrised Percent Change (ASPC) (Reuter, 2012), with respect to volumetric measurements at the first and second timepoints, V1 and V2 respectively.
ASPC = 100 |V₂ - V₁| / (0.5 (V₁ + V₂)) (1)
We utilise ASPC to quantify the change in volumetric estimates across subcortical, cortical grey matter, and white matter structures. Lower ASPC values indicate greater reliability and reduced variability in volumetric measurements, however in longitudinal studies where true anatomical changes occur, ASPC reflects a combination of biological change and measurement error.
Results:
We compare the reliability of volumetric measures for 500 UKB subjects using independently processed timepoints [CROSS] and longitudinal processed [LONG] methods, shown in Fig. 1, for various left hemisphere subcortical, cortical grey and white matter volume segmentations as indicated. Fig. 1 illustrates that across all structures the measured ASPC through [LONG] processing consistently reduces variability compared to [CROSS] processing. Substantial reductions in variability in subcortical structures, such as the amygdala, by over 5% through implementation of [LONG] processing, suggest that longitudinal processing can significantly enhance the reliability of volumetric measurements. This reliability improvement aligns with the reported findings of the FreeSurfer longitudinal pipeline study (Reuter, 2012) and is further demonstrated by similar (>5%) improvements in cortical grey matter and white matter variability measurements, such as the inferior parietal and parahippocampal areas respectively. This reduction in variance translates into increased precision for both volumetric and surface-based analyses (Reuter, 2012), improving statistical power to detect subtle longitudinal changes and enabling potential reductions in sample size requirements for future studies.
Conclusions:
This study validates the longitudinal FreeSurfer processing pipeline for two-timepoint UK Biobank data using ASPC as a robust metric for assessing volumetric variability. By reducing variability across cortical and subcortical structures, longitudinal processing enhances measurement reliability and reduces the errors inherent in single-timepoint methods. These findings provide confidence that longitudinal FreeSurfer will substantially benefit two-timepoint UKB subjects as we begin to process that data.
Modeling and Analysis Methods:
Image Registration and Computational Anatomy 1
Novel Imaging Acquisition Methods:
Anatomical MRI 2
Keywords:
Computational Neuroscience
MRI
Open-Source Software
Sub-Cortical
White Matter
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.
Miller, K.L. (2016). Multimodal population brain imaging in the UK Biobank prospective epidemiological study. Nature Neuroscience, 19(11).
Reuter, M. (2011). Avoiding asymmetry-induced bias in longitudinal image processing. In NeuroImage, 57(1), 19-21.
Reuter, M. (2012). Within-Subject Template Estimation for Unbiased Longitudinal Image Analysis. Neuroimage, 61(4), 1402-1418.