- Rating
- 2
- Confidence
- 3
- Informativeness
- 1
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper proposes IRS-SSL, a self-supervised learning framework for structural brain MRI that learns representations by modeling inter-regional structural similarities at both individual and population levels.
Critical strengths
The paper is well structured and clearly written, with extensive experiments across multiple downstream tasks. Incorporating inter-regional structural similarities into self-supervised learning is an interesting direction that goes beyond conventional voxel-level or instance-level learning.
Critical weaknesses
-
The proposed method relies on a small set of T1 intensity statistics to construct inter-regional structural relationships. Are these features sufficient to capture meaningful anatomical information? Have the authors considered comparing them with more conventional morphological features, such as cortical thickness, regional volume, or surface area?
-
The individual formulation consistently achieves better performance, but the underlying reason is not well explained. Is this because individual-level relationships provide more informative supervision, or because the two formulations differ in the amount and type of structural information they preserve? A more detailed analysis would help clarify this finding.
-
The current comparison between HO and Desikan-Killiany is rather limited. First, the authors could evaluate different parcellation resolutions, such as varying the number of ROIs within the same atlas family. Second, a random parcellation baseline would be particularly informative. If randomly defined brain regions achieve comparable performance, it would raise questions about whether anatomically meaningful structural relationships are actually responsible for the improvements. The existing random region pairing experiment does not fully address this concern.
Questions for authors
See Weakness
8 comments
Comment for Reviewer nZQ8 (2/3)
The subject-identification experiment provides additional support for its retention of subject-discriminative anatomical information, rather than establishing equivalence to conventional morphological features.
Table tc1
| Method | Runtime (1 scan) mean ± SD |
|---|---|
| DeepSCAN segmentation + CortexMorph | 18.7 ± 1.2 sec |
W2: The individual formulation consistently achieves better performance, but the underlying reason is not well explained. Is this because individual-level relationships provide more informative supervision, or because the two formulations differ in the amount and type of structural information they preserve? A more detailed analysis would help clarify this finding.
We thank the reviewer for this insightful comment. We would like to clarify that IRS-SSL-Ind and IRS-SSL-Pop are based on two distinct, well-established neuroscientific constructs, namely individualized structural similarity [c5,c1] and population-level structural covariance [c4, c6]. Both have been associated with clinically relevant variation, but they encode different types of information (See detailed discussion in Sec 2.1).
In the self-supervised task formulations, an important distinction lies in how supervision and gradients are derived. IRS-SSL-Ind constructs a separate structural similarity matrix for each subject, providing subject-specific supervisory signals and gradient contributions. In contrast, IRS-SSL-Pop computes covariance matrices from regional features aggregated across the mini-batch. Consequently, its gradients are coupled across subjects and driven by batch-level structural relationships rather than independent subject-specific targets. The pretraining loss graph (Appendix. I.1 Fig 8) also suggests that IRS-SSL-pop saturates much quicker than IRS-SSL-ind.
IRS-SSL-Ind provides independently defined subject-level supervision, whereas IRS-SSL-Pop provides shared batch-level supervision with cross-subject gradient coupling. This difference in supervisory granularity may help explain the stronger performance of IRS-SSL-Ind on individual-level downstream tasks, where preserving subject-specific anatomical variation is particularly relevant.
W3: The current comparison between HO and Desikan-Killiany is rather limited. First, the authors could evaluate different parcellation resolutions, such as varying the number of ROIs within the same atlas family. Second, a random parcellation baseline would be particularly informative. If randomly defined brain regions achieve comparable performance, it would raise questions about whether anatomically meaningful structural relationships are actually responsible for the improvements. The existing random region pairing experiment does not fully address this concern.
First, the authors could evaluate different parcellation resolutions, such as varying the number of ROIs within the same atlas family
We thank the reviewer for suggesting an evaluation across different parcellation resolutions. While the Harvard-Oxford (HO) anatomical atlas does not offer multi-scale parcellations like functional atlases (e.g., Schaefer or Craddock), we evaluated ROI granularity within the HO family by comparing the coarse 21-ROI HO atlas against a finer 63-ROI parcellation obtained by combining the HO Cortical and Subcortical atlases [c7]. Specifically, whereas the 21-ROI atlas represents the entire cerebral cortex as just two coarse hemisphere-level ROIs alongside subcortical structures, merging it with the HO Cortical atlas subdivides the cortical mantle into 48 finer-grained anatomical parcels alongside the 15 subcortical gray-matter structures (yielding 63 total ROIs after excluding overlapping whole-cortex, white-matter, and ventricular masks).
Comparing our model across both the coarse (21-ROI) and fine-grained cortical (63-ROI) parcels shows that the coarser (HO 21) atlas performs slightly better than the (HO 63) atlas for many tasks. For the brain-tumor grading task, the finer granularity of the (HO 63) atlas would preserve localized lesion cues better than the coarser atlas.
Table tc2: Classification Tasks (Bal. Acc ↑ / AUC ↑ / F1 ↑)
| Task | Setting | IRS-SSL-Ind (HO 63) | IRS-SSL-Ind (HO 21)* |
|---|---|---|---|
| Gender | ID | 82.903 / 0.908 / 0.829 | 85.47 / 0.918 / 0.856 |
| MCI / CN | ID | 70.793 / 0.808 / 0.717 | 75.290 / 0.826 / 0.757 |
| FTD | ID | 79.683 / 0.886 / 0.795 | 80.593 / 0.856 / 0.803 |
| Tumor Grade | ID | 64.21 / 0.788 / 0.66 | 61.763 / 0.692 / 0.625 |
| AD / MCI / CN | OOD | 60.353 / 0.852 / 0.559 | 64.433 / 0.834 / 0.605 |
| MCI / CN | OOD | 73.537 / 0.83 / 0.711 | 75.067 / 0.893 / 0.718 |
Table tc3: Regression & Segmentation
| Model | Age (ID) L1 ↓ | Age (OOD) L1 ↓ | Seg. DSC ↑ |
|---|---|---|---|
| IRS-SSL-Ind (HO 63) | 9.033 | 10.731 | 65.653 |
| IRS-SSL-Ind (HO 21)* | 9.238 | 10.605 | 66.6 |
| *denotes results from the manuscript. | |||
| (to be continued) |
Comment for Reviewer nZQ8 (3/3)
Second, a random parcellation baseline would be particularly informative. If randomly defined brain regions achieve comparable performance, it would raise questions about whether anatomically meaningful structural relationships are actually responsible for the improvements.
To evaluate whether anatomically meaningful boundaries drive our improvements rather than arbitrary spatial partitioning, we generated a spatially contiguous random parcellation (R=21) within the identical Harvard-Oxford foreground mask using Poisson disk sampling [c8] (Appendix H. Figure 6). We use the exact same model architecture, pretraining hyperparameters, and relational supervision loss function as described in the paper, with the sole difference being the atlas.
As shown in the table tc4 and tc5, replacing anatomically defined regions with a random parcellation (IRS-SSL-Ind(HO Random Parcel), R=21) causes a substantial performance drop across every single task and metric compared to both HO 21 (e.g., OOD AD/MCI/CN balanced accuracy drops from 64.43% to 43.28%, OOD MCI vs. CN balanced accuracy drops from 75.07% to 53.96%, Age MAE worsens from 9.24 to 11.13, and Segmentation DSC drops from 66.60 to 56.32). This confirms that our relational self-supervision specifically relies on anatomically meaningful brain structures and their inter-regional relationships rather than arbitrary spatial partitioning.
Table tc4: Classification Tasks (Bal. Acc ↑ / AUC ↑ / F1 ↑)
| Task | Setting | IRS-SSL-Ind (HO Random Parcel) | IRS-SSL-Ind (Original) |
|---|---|---|---|
| Gender | ID | 75.26 / 0.897 / 0.766 | 85.47 / 0.918 / 0.856 |
| MCI / CN | ID | 63.437 / 0.757 / 0.635 | 75.290 / 0.826 / 0.757 |
| FTD | ID | 73.37 / 0.837 / 0.733 | 80.593 / 0.856 / 0.803 |
| Tumor Grade | ID | 53.447 / 0.662 / 0.514 | 61.763 / 0.692 / 0.625 |
| AD / MCI / CN (OASIS-3) | OOD | 43.283 / 0.675 / 0.426 | 64.433 / 0.834 / 0.605 |
| MCI / CN (OASIS-3) | OOD | 53.96 / 0.624 / 0.537 | 75.067 / 0.893 / 0.718 |
| MCI / CN (AIBL) | OOD | 52.357 / 0.618 / 0.526 | 64.336 / 0.72 / 0.625 |
| AD / MCI / CN (AIBL) | OOD | 41.378 / 0.625 / 0.452 | 53.987 / 0.765 / 0.534 |
Table tc5: Regression & Segmentation
| Model | Age (ID) L1 ↓ | Age (OOD) L1 ↓ | Seg. DSC ↑ |
|---|---|---|---|
| IRS-SSL-Ind (HO Random Parcel) | 11.132 | 12.531 | 56.32 |
| IRS-SSL-Ind (Original) | 9.238 | 10.605 | 66.6 |
We thank the reviewer for the constructive feedback.
References :
[c1] Zhao, K., et al. (2021). Regional radiomics similarity networks (R2SNs) in the human brain: Reproducibility, small-world properties and a biological basis. Network neuroscience (Cambridge, Mass.), 5(3), 783–797.
[c2] Rebsamen, M., et al. (2020). Direct cortical thickness estimation using deep learning-based anatomy segmentation and cortex parcellation. Human brain mapping, 41(17), 4804–4814.
[c3] McKinley, R., & Rummel, C. (2023). CortexMorph: fast cortical thickness estimation via diffeomorphic registration using VoxelMorph. Medical image computing and computer-assisted intervention : MICCAI 2023 (pp. 730–739.)
[c4] Dong, Chao, et al. "Brain structural covariances in the ageing brain in the UK Biobank." Brain Structure and Function 229.5 (2024).
[c5] Liu, Han, et al. "A radiomics-based brain network in T1 images: construction, attributes, and applications." Cerebral Cortex 34.2 (2024)
[c6] Mongay-Ochoa, Neus, et al. "Structural covariance analysis for neurodegenerative and neuroinflammatory brain disorders." Brain 148.9 (2025)
[c7] Crofts, Jonathan J., et al. "Network analysis detects changes in the contralesional hemisphere following stroke." NeuroImage 54.1 (2011): 161-169.
[c8] Arslan, Salim, et al. "Human brain mapping: A systematic comparison of parcellation methods for the human cerebral cortex." NeuroImage 170 (2018): 5-30.
Comment for Reviewer nZQ8 (1/3)
We thank the reviewer for the positive comment. Indeed our study conceptualizes inter-regional structural organization as the primary self-supervision principle, going beyond conventional voxel-level or instance-level learning.
We thank the reviewer for this comment. Our set of T1-intensity statistics provide a compact, computationally tractable description of regional intensity patterns, suitable for large-scale pre-training. IRS-SSL uses these statistics to derive inter-regional relationships. In addition to the strong downstream performance and radiologist evaluation reported in the paper, here we additionally conducted a cross-session individual-identification experiment (see below). This result indicates that the descriptors contain reproducible, subject-specific structural information and supports their utility as a self-supervisory signal. We do not claim that they provide a complete characterization of brain morphology or replace conventional morphological features (Note that conventional morphological features could be impractical for our large-scale pretraining. See below.)
Individual identification experiment. To provide additional support that our set of T1 intensity statistics retain meaningful information for inter-regional structural relationships, we conducted an individual identification experiment on the HCP test–retest dataset N=45, following the R2SN protocol[c1]. We use A(k) as described in our paper as the input for this identification experiment.
The results support that our set of intensity statistics preserve meaningful, subject-specific anatomical information, as evidenced by significant test-retest identification performance. We evaluated all (n5) combinations of our five intensity statistics for n=3,4, and 5, and report the identification accuracy averaged across combinations for each value of n. We also evaluated the set of 25 statistics described in R2SN [1] on the Harvard-Oxford Atlas as a reference.
As shown in the results, using our set of intensity statistics achieved accuracy comparable to that obtained with the 25 R2SN statistics (77.8%). These results support that our set of intensity statistics retains reproducible, subject-specific information consistent with anatomical variation. We have added this analysis and clarified its interpretation in Appendix G of the revised manuscript.
We agree that conventional morphological descriptors such as cortical thickness, regional volume, and surface area are important anatomical baselines. However, our objective is not to propose the selected T1 intensity statistics as substitutes for these descriptors, or to claim that they contain more anatomical information. Rather, we seek a compact structural representation that is suitable for large-scale pre-training.
As discussed in Sec 2.3, conventional morphological features could be impractical for our pretraining task, because of the heavy computational overhead due to repeated anatomical segmentation and regional cortical thickness estimation in every pretraining iteration. To assess the potential computational overhead of using conventional descriptors during pretraining, we extracted cortical thickness and regional volumes from 10 subjects from the A4 dataset using the standard two-stage voxel-based pipeline comprising DeepSCAN [c2] for tissue and anatomical segmentation and CortexMorph [c3] for registration-based regional cortical thickness estimation. Assuming one extraction per scan, the mean runtime of 18.7 seconds per scan on our pretraining dataset of 12541 scans for 200 epochs would be prohibitive for pretraining. In contrast, as discussed in Sec 2.3, A(k) based on intensity (radiomics) statistics would not require repeated anatomical segmentation and regional cortical thickness estimation, and is suitable for large-scale pretraining. ( to be continued )