Estimation of Brain Atrophy from Baseline Cerebral Microbleeds
Poster No:
168
Submission Type:
Abstract Submission
Authors:
Mohammad-Reza Nazem-Zadeh1, Catherine Robb1, Ben Sinclair1, Amy Brodtmann1, Stephanie Ward1, Terence O’Brien1, John McNeil1, Meng Law1
Institutions:
1Monash University, Melbourne, Victoria
First Author:
Co-Author(s):
Introduction:
Cerebral microbleeds (CMBs) are small chronic brain hemorrhages, present in approximately 20-30% of the elderly population and up to 50% in individuals with hypertension, diabetes, or cerebrovascular conditions (Greenberg et al., 2009). CMBs are associated with an increased risk of ischemic and hemorrhagic strokes, cognitive impairment, and neurodegenerative diseases such as Alzheimer's disease and vascular dementia, often progressing with brain atrophy (Cordonnier, Al-Shahi Salman, & Wardlaw, 2007; Cordonnier & van der Flier, 2011). CMBs appear as hypointense lesions on gradient echo magnetic resonance imaging (GE-MRI), including T2* and susceptibility-weighted imaging (SWI) (Camblor et al., 2023; Shams et al., 2015). However, manual labelling of CMBs is labor-intensive, subjective, and potentially inaccurate. We hypothesized that automated detection of CMBs using AI could achieve reliable results comparable to manual labelling. Furthermore, we hypothesized that the number of baseline CMBs could predict brain atrophy over a three-year period.
Methods:
We retrospectively analyzed 573 participants from the ASPREE Neuro Study (McNeil et al., 2018; McNeil et al., 2017; McNeil et al., 2019), of which 465 had imaging data at all three time points (baseline, year 1, and year 3). Freesurfer (Fischl, 2012) was used to extract subcortical volumes and cortical measures, including grey matter volume, number of vertices, surface area, and cortical thickness, for whole-brain and lobe-specific regions (frontal, parietal, temporal, occipital lobes, cingulate, and insula). Brain atrophy was measured as the change in these metrics from baseline to year 3. Pairwise t-tests with Bonferroni correction were used to evaluate atrophy patterns across time points and brain regions, controlling for intracranial volume (ICV). The SHIVA-CMB pipeline (Tsuchida et al., 2024), a deep-learning-based automatic detection algorithm, was fine-tuned using local T2* (SWI-magnitude) and phase-corrected SWI images from 26 participants with expert-labelled CMBs at all three time points. Cluster analysis was used to count CMBs in local regions corresponding to atrophy measures for all 465 participants. Global CMB counts were validated by a neuroradiologist.
Results:
A strong correlation was observed between CMB counts from the neuroradiologist and the SHIVA-CMB algorithm applied on both T2* and SWI images (Pearson coefficient r = 0.77 and r = 0.72, respectively; p ≈ 0; Fig. 1). Significant global brain atrophy was detected from baseline to year 3 for most cortical and subcortical measures (Bonferroni-corrected p < 0.05). Baseline CMB counts on T2* images correlated significantly with cortical thickness changes in the frontal and occipital lobes (Spearman coefficient r = -0.1 to -0.11; p < 0.05; Fig. 2 (a) and (b)). Similarly, baseline CMB counts on SWI images correlated with changes in the cortical number of vertices, grey matter volume, and surface area in the insula (Spearman coefficient r = -0.1 to -0.13, p < 0.05; Fig. 2 (c), (d), and (e)).
·Figure 1: Comparison of AI-based CMB detection versus manual labelling after SHIVA-CMB retraining using T2* (a) and SWI (b).
·Figure 2: Correlation of baseline CMBs with brain atrophy metrics: (a, b) cortical thickness in frontal and occipital lobes (T2*); (c-e) cortical vertices, grey matter volume, and surface area in the
Conclusions:
This study validated the SHIVA-CMB algorithm after fine-tuning with local data, for accurately detecting and quantifying CMBs, and found promising association between regional CMBs at baseline and later brain atrophy. Early detection of CMBs and monitoring their progression could improve the diagnosis and management of cerebrovascular and neurodegenerative diseases, ultimately aiding in the prevention of cognitive decline.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 1
Lifespan Development:
Lifespan Development Other 2
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Cortical Anatomy and Brain Mapping
Novel Imaging Acquisition Methods:
Anatomical MRI
Multi-Modal Imaging
Keywords:
Aging
Cerebrovascular Disease
Data analysis
Degenerative Disease
MRI
Structures
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?
Are you Internal Review Board (IRB) certified? Please note: Failure to have IRB, if applicable will lead to automatic rejection of abstract.
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.
Cordonnier, C., Al-Shahi Salman, R., & Wardlaw, J. (2007). Spontaneous brain microbleeds: systematic review, subgroup analyses and standards for study design and reporting. Brain, 130(8), 1988-2003.
Cordonnier, C., & van der Flier, W. M. (2011). Brain microbleeds and Alzheimer’s disease: innocent observation or key player? Brain, 134(2), 335-344.
Fischl, B. (2012). FreeSurfer. Neuroimage, 62(2), 774-781.
Greenberg, S. M., Vernooij, M. W., Cordonnier, C., Viswanathan, A., Salman, R. A.-S., Warach, S., . . . Breteler, M. M. (2009). Cerebral microbleeds: a guide to detection and interpretation. The Lancet Neurology, 8(2), 165-174.
McNeil, J. J., Nelson, M. R., Woods, R. L., Lockery, J. E., Wolfe, R., Reid, C. M., . . . Storey, E. (2018). Effect of aspirin on all-cause mortality in the healthy elderly. New England Journal of Medicine, 379(16), 1519-1528.
McNeil, J. J., Woods, R. L., Nelson, M. R., Murray, A. M., Reid, C. M., Kirpach, B., . . . Tonkin, A. M. (2017). Baseline characteristics of participants in the ASPREE (ASPirin in Reducing Events in the Elderly) study. The Journals of Gerontology: Series A, 72(11), 1586-1593.
McNeil, J. J., Woods, R. L., Ward, S. A., Britt, C. J., Lockery, J. E., Beilin, L. J., & Owen, A. J. (2019). Cohort profile: The ASPREE longitudinal study of older persons (ALSOP). International journal of epidemiology, 48(4), 1048-1049h.
Shams, S., Martola, J., Cavallin, L., Granberg, T., Shams, M., Aspelin, P., . . . Kristoffersen-Wiberg, M. (2015). SWI or T2*: which MRI sequence to use in the detection of cerebral microbleeds? The Karolinska Imaging Dementia Study. American Journal of Neuroradiology, 36(6), 1089-1095.
Tsuchida, A., Goubet, M., Boutinaud, P., Astafeva, I., Nozais, V., Hervé, P.-Y., . . . Joliot, M. (2024). SHIVA-CMB: A Deep-Learning-based Robust Cerebral Microbleed Segmentation Tool Trained on Multi-Source T2* GRE-and Susceptibility-weighted MRI.