Enlarged Perivascular Spaces and Vascular Risk Factors: Cross-Sectional and Longitudinal Analysis
Poster No:
1617
Submission Type:
Late-Breaking Abstract Submission
Authors:
Shizuka Hayashi1, Jiyang Jiang2, Yang Song3, Matthew Paradise4, David Chan1, Dadong Wang1, Perminder Sachdev1, wei wen5
Institutions:
1University of New South Wales, Sydney, NSW, 2Centre for Healthy Brain Aging (CHeBA), School of Psychiatry, University of New South Wales (UNSW), Sydney, New South Wales, 3School of Computer Science and Engineering, University of New South Wales (UNSW), Sydney, New South Wales, 4University of New South Wales, UNSW Sydney, N/A, 5University of New South Wales, UNSW Sydney, NSW
First Author:
Co-Author(s):
Jiyang Jiang, PhD
Centre for Healthy Brain Aging (CHeBA), School of Psychiatry, University of New South Wales (UNSW)
Sydney, New South Wales
Centre for Healthy Brain Aging (CHeBA), School of Psychiatry, University of New South Wales (UNSW)
Sydney, New South Wales
Yang Song, PhD
School of Computer Science and Engineering, University of New South Wales (UNSW)
Sydney, New South Wales
School of Computer Science and Engineering, University of New South Wales (UNSW)
Sydney, New South Wales
Late Breaking Reviewer(s):
Rosanna Olsen
Rotman Research Institute, Baycrest Academy for Research and Education
Toronto, Ontario
Rotman Research Institute, Baycrest Academy for Research and Education
Toronto, Ontario
Introduction:
The perivascular space (PVS), a fluid-filled region surrounding small brain vessels, is crucial for interstitial fluid drainage, toxin clearance, and immune function via the glymphatic system (Nedergaard M, 2013). PVS are commonly observed in the basal ganglia (BG) and centrum semiovale (CSO) (Doubal FN, 2010) and are clinically relevant to vascular cognitive impairment and Alzheimer's disease (Banerjee G, 2017). While PVS enlargement mechanisms remain under investigation, both genetic and modifiable vascular risk factors have been extensively studied (Francis, 2016). Hypertension is consistently linked to PVS burden, but associations with smoking, diabetes, BMI, and alcohol remain mixed. Similarly, the APOE-ɛ4 allele's role is debated (Lynch M, 2022), highlighting the need for refined research methodologies.
Visual rating remains the clinical gold standard for PVS assessment, though it is labour-intensive and rater-dependent, particularly for CSO-PVS (Potter GM, 2015). While deep learning methods improve diagnostic accuracy, they require extensive labelled data and are often validated against single raters. This study introduced a weakly supervised DL model for PVS detection, using non-inferiority analysis to compare its performance to multiple manual ratings. This enables large-scale clinical dataset processing.
This study investigated vascular risk factors-including hypertension, hypercholesterolemia, diabetes, obesity, alcohol consumption, smoking, and a composite vascular risk score in relation to PVS burden in BG and CSO using UK Biobank data (Sudlow C, 2015). Both cross-sectional and longitudinal analyses were performed, with additional sex-stratified analysis to explore sex-specific associations.
Visual rating remains the clinical gold standard for PVS assessment, though it is labour-intensive and rater-dependent, particularly for CSO-PVS (Potter GM, 2015). While deep learning methods improve diagnostic accuracy, they require extensive labelled data and are often validated against single raters. This study introduced a weakly supervised DL model for PVS detection, using non-inferiority analysis to compare its performance to multiple manual ratings. This enables large-scale clinical dataset processing.
This study investigated vascular risk factors-including hypertension, hypercholesterolemia, diabetes, obesity, alcohol consumption, smoking, and a composite vascular risk score in relation to PVS burden in BG and CSO using UK Biobank data (Sudlow C, 2015). Both cross-sectional and longitudinal analyses were performed, with additional sex-stratified analysis to explore sex-specific associations.
Methods:
We applied a machine learning-based deep learning model to segment PVS in MRI scans from the UK Biobank. PVS burden was assessed in two brain regions, the basal ganglia (BG) and centrum semiovale (CSO), in both cross-sectional (38,121 subjects) and longitudinal (4,225 subjects) analyses of community-dwelling individuals aged 47 to 90 years. We evaluated key vascular risk factors (hypertension, hypercholesterolemia, obesity, diabetes, alcohol consumption, and smoking), APOE4 genotype, and their associations with PVS burden using robust statistical methods.
·Study workflow from data acquisition and PVS segmentation to cross-sectional and longitudinal analyses of vascular risk factors and PVS burden.
Results:
In the cross-sectional analysis, vascular risk factors were associated with increased BG-PVS, including hypertension (b = 0.089, 95% CI: 0.069 - 0.108), hypercholesterolemia (b = 0.043, 95% CI: 0.017 - 0.064), obesity (b = 0.043, 95% CI: 0.016 - 0.064), and smoking (b = 0.056, 95% CI: 0.037 - 0.074). In contrast, APOE‐ɛ4 carriers (b = 0.039, 95% CI: 0.0015 - 0.076) and individuals with hypertension (b = 0.093, 95% CI: 0.056 - 0.13) had increased CSO-PVS burden. Alcohol consumption showed a sex-specific effect, with moderate intake reducing BG-PVS burden in males but increasing it in females with moderate-to-heavy consumption. Longitudinally, ageing was associated with a greater increase in CSO-PVS over time (b = 0.007, 95% CI: 0.000 - 0.015), while baseline vascular risk factors were not significantly associated with PVS progression after adjusting for age and sex.
·Cross-sectional and stratified analyses of vascular risk factors on PVS. Panels A and B show main and sex interactions, C compares effects. Error bars show 95% CI, with * (p < 0.05) and ** (corrected)
Conclusions:
Our findings highlight the region-specific and sex-dependent effects of vascular risk factors on PVS burden, reinforcing their relevance in cerebrovascular disease. Ageing was the primary determinant of CSO-PVS progression, while vascular risk factors predominantly influenced BG-PVS. Additionally, the machine learning approach provided an efficient method for large-scale MRI analysis, demonstrating its potential for future automated assessments of cerebrovascular disease. These results emphasize the importance of considering sex, vascular health, and statistical methods when evaluating perivascular spaces in ageing populations.
Lifespan Development:
Aging 2
Modeling and Analysis Methods:
Multivariate Approaches 1
Segmentation and Parcellation
Other Methods
Keywords:
Aging
Cerebrovascular Disease
Machine Learning
MRI
Statistical Methods
Other - perivascular space
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.
Doubal FN. (Mar 2010). Enlarged perivascular spaces on MRI are a feature of cerebral small vessel disease. Stroke, 41(3), 450-4.
Francis F. (Jun 2019). Perivascular spaces and their associations with risk factors, clinical disorders and neuroimaging features: A systematic review and meta-analysis. Int J Stroke, 14(4), 359-371.
Lynch M. (2022). Perivascular spaces as a potential biomarker of Alzheimer's disease. Front Neurosci. 16, 1021131.
Nedergaard M. (Jun 28 2013). Neuroscience. Garbage truck of the brain. Science. 340(6140), 1529-30.
Potter GM. (2015). Cerebral perivascular spaces visible on magnetic resonance imaging: development of a qualitative rating scale and its observer reliability. Cerebrovasc Dis. 39(3-4), 224-31.
Sudlow C. (Mar 2015). UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 12(3), e1001779.