Exploring the Capacity of 7T MRI to Capture the Claustrum In Vivo at Varying Spatial Resolutions
Poster No:
1754
Submission Type:
Abstract Submission
Authors:
Navona Calarco1, Skerdi Progri1, Sriranga Kashyap2, Kâmil Uludağ2
Institutions:
1University of Toronto, Toronto, Ontario, 2University Health Network, Toronto, ON
First Author:
Co-Author(s):
Introduction:
The human claustrum is exceptionally difficult to study in vivo due to its extreme thinness and proximity to surrounding structures. A growing number of MRI studies report radically varying shape characteristics; for example, manually-derived claustral volumes differ by up to fivefold [FIGURE 1A]. However, recent claustrum mappings derived from ultra-high resolution ex vivo MRI (Coates & Zaretskaya, 2023; Mauri et al., 2024) closely align with one derived from histology (Calarco et al., 2023), suggesting that resolution, not contrast, is the primary limitation. This study systematically evaluates the ability of in vivo ultra-high field MRI to capture the claustrum compared to a histological reference atlas.
Methods:
We assessed claustral shape characteristics, here reporting volume and axis-aligned spatial extent, in three MRI datasets (each n=10). All datasets acquired MP2RAGE at 7-Tesla, at isotropic resolutions of 0.5mm, 0.7mm, and 1.0mm, respectively. The claustrum was annotated by human raters familiar with its anatomical features, but based only on its apparent contrast in MR images [FIGURE 1B]. Shape characteristics were compared to those of a three-dimensional histological "gold standard" model (Calarco et al., 2023, 2024) derived from the BigBrain dataset at 100-micron resolution (Amunts et al., 2013) [FIGURE 1C]. Downsampling simulations of the gold standard at varying resolutions (from 0.4mm to 2.0mm isotropic) and segmentation thresholds (from 0.2 to 0.8), further characterised the limitations of in vivo MRI claustral capture.
Results:
Within the three in vivo MRI datasets, decreasing resolution monotonically decreased claustral volumes and extents [FIGURE 2A]. The most pronounced discrepancy was along the anterior-posterior extent, where an average difference of >5mm was observed between the 0.5mm and 1.0mm datasets, with mostly anterior aspects lost. Variability in volumes was notably high in the 1.0mm dataset [Levene's p<0.05], likely reflecting partial voluming and uncertainty in boundary delineation by human raters. Despite inter-subject variability, differences in extents across all resolutions were statistically significant [all one-way ANOVA p<0.01].
Comparison of the three in vivo MRI datasets to the gold standard [FIGURE 2A, red line] revealed underestimation of maximal extents along all axes. The largest discrepancy of 25mm was observed along the superior-inferior axis, attributable to the failure of the 1.0mm dataset to resolve the ventral claustrum's extension into the temporal lobe. Paradoxically, partial voluming overestimated volumes in the 0.5mm and 0.7mm datasets. This occurred because much of the dorsal claustrum's thin mediolateral extent was inflated, and the ventral claustrum's discontinuous "puddles"-when captured at all-were fused into larger volumes. Dice overlap between the gold standard and the three resolutions was 54% (marginally adequate), 43% (poor), and 15% (poor), respectively [FIGURE 2B].
Downsampling of the gold standard revealed truncated spatial extents and inflated volumes at low thresholds, across all three resolutions [FIGURE 2C]. These results closely matched the characteristics observed in the three in vivo MRI datasets, affirming that discrepancies with the gold standard reflect the inherent limitations of resolution rather than errors in manual segmentation.
Comparison of the three in vivo MRI datasets to the gold standard [FIGURE 2A, red line] revealed underestimation of maximal extents along all axes. The largest discrepancy of 25mm was observed along the superior-inferior axis, attributable to the failure of the 1.0mm dataset to resolve the ventral claustrum's extension into the temporal lobe. Paradoxically, partial voluming overestimated volumes in the 0.5mm and 0.7mm datasets. This occurred because much of the dorsal claustrum's thin mediolateral extent was inflated, and the ventral claustrum's discontinuous "puddles"-when captured at all-were fused into larger volumes. Dice overlap between the gold standard and the three resolutions was 54% (marginally adequate), 43% (poor), and 15% (poor), respectively [FIGURE 2B].
Downsampling of the gold standard revealed truncated spatial extents and inflated volumes at low thresholds, across all three resolutions [FIGURE 2C]. These results closely matched the characteristics observed in the three in vivo MRI datasets, affirming that discrepancies with the gold standard reflect the inherent limitations of resolution rather than errors in manual segmentation.
Conclusions:
Our findings illuminate the inherent challenge of claustral capture via in vivo MRI. Encouragingly, skilled manual segmentation effectively identifies claustral voxels, a critical capability given the current lack of adequate automated segmentation tools. While standard-resolution MRI (≥1 mm) presents substantial underestimation of its extent, submillimeter MRI at ultra-high field can reliably capture a large fraction of claustral anatomy, and should be prioritized in studies of human brain function and disease.
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Subcortical Structures 1
Neuroinformatics and Data Sharing:
Brain Atlases 2
Keywords:
Atlasing
Basal Ganglia
MRI
Segmentation
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.
Calarco, N., Kashyap, S., Uludag, K. (2023), Establishing an MRI reference for the human claustrum. Poster presented at OHBM 2023, Montreal.
Coates, A., & Zaretskaya, N. (2023). High-Resolution Dataset of Manual Claustrum Segmentation. Data in Brief, 54, 110253.
Mauri, C., Fritz, R., Mora, J., Billot, B., Iglesias, J. E., Van Leemput, K., Augustinack, J., & Greve, D. N. (2024). A contrast-agnostic method for ultra-high resolution claustrum segmentation. Preprint available at http://arxiv.org/abs/2411.15388.