- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper proposes Conflict Recalibration and Geometric Aggregation (CRGA) for medical image classification using a pretrained VMamba backbone. The framework contains two components, and both components preserve the original Cartesian selective-scan paths and recover the pretrained backbone mapping at initialization through zero-initialized projections or residual gates.
Critical strengths
CCR intervenes directly in selective state writing and readout rather than simply appending a generic feature-attention block. Comparing directional responses before Cross-Merge and using the resulting information in the next stage is a clear architectural hypothesis. The spatial alignment and directional unfolding operations are explicitly described.
Critical weaknesses
CCR intervenes directly in selective state writing and readout rather than simply appending a generic feature-attention block. Comparing directional responses before Cross-Merge and using the resulting information in the next stage is a clear architectural hypothesis. The spatial alignment and directional unfolding operations are explicitly described.
The central hypothesis is that cross-axis discrepancies provide useful conditioning information beyond the features already available to VMamba. However, the current ablation primarily establishes that adding CCR improves performance; it does not establish why. In particular, comparing a full conflict map with only the eight lowest-scoring positions simultaneously changes spatial coverage, signal density, and score selection. This does not isolate the usefulness of high-discrepancy regions or the specificity of directional conflict.
Ablations of (B)-only, (C)-only, and joint modulation are missing, as are comparisons of same-stage versus next-stage conditioning. These experiments would help establish whether the proposed intervention point is essential.
GCRSA uses the feature-grid center, which the authors explicitly distinguish from an anatomical or acquisition center. This distinction is important: grouping tokens by radius around the image center does not necessarily group them by anatomical layer, lesion boundary, or invasion depth.
The main EUS improvement is 2.2 percentage points on 334 test images. Five development-fold models are evaluated on the same held-out test set, but only mean results are reported. These runs do not constitute five independent test cohorts.
Table 2 does not include the unmodified VMamba-Small backbone or the CCR-only and GCRSA-only variants. Consequently, the public results cannot establish whether the proposed modules improve over their own backbone outside EUS.
Questions for authors
Does CCR outperform parameter-matched conditioning derived from merged features, constant maps, or spatially shuffled conflict maps?
Why is the single-channel, rank-one modulation sufficient? What happens with (B)-only, (C)-only, direction-specific, or direct feature modulation?
Can the authors provide per-run results and patient-clustered confidence intervals for the differences between VMamba, CCR, GCRSA, and the combined model?
What are the exact grouping parameters and aggregation implementation? What are the measured latency, memory, and resolution-scaling costs?
3 comments
R-wgaq–2. Matched selection and interpretation of conflict scores
The original Full-map versus Low-8 ablation changes both coverage and selection. We added Top-8 and Random-8 controls, keeping eight positions at each applicable stage in all sparse variants. Top/Low use each input's largest/smallest discrepancy scores; Random-8 uses predetermined score-independent positions held fixed across training and testing.
Table R-wgaq-2. Full and matched sparse CCR selection on EUS.
| Selection | ACC (%) | Recall (%) | F1 (%) | AUC (%) |
|---|---|---|---|---|
| Full Conflict Map | 81.3 ± 0.52 | 81.5 ± 0.80 | 82.7 ± 0.55 | 93.5 ± 0.57 |
| Top-8 | 80.4 ± 0.25 | 79.2 ± 0.28 | 81.2 ± 0.66 | 93.3 ± 0.26 |
| Random-8 | 78.8 ± 0.24 | 76.9 ± 2.62 | 78.5 ± 1.56 | 93.5 ± 0.17 |
Values are percentages. ± denotes run-level SD.
Full-map conditioning yields the highest observed ACC, although comparison with sparse maps still changes coverage. Under matched budgets, Low-8 is slightly above Top-8, while both exceed Random-8; Random-8 uses fixed score-independent positions. The results support using conflict scores as a spatially distributed relational signal for downstream state computation. This interpretation is reinforced by the Feature-BC, Constant-BC, and Shuffled-BC controls in R-wgaq-1.
R-wgaq–3. GCRSA's center, exact nature of aggregation, and geometric controls
Preprocessing and center definition. EUS images were manually cropped to retain the complete esophageal structure. The crop boundaries were determined by organ completeness and did not use probe location. We did not perform probe detection, probe-centered alignment, or anatomical-center estimation. Each cropped image was directly resized to the target resolution without padding. The GCRSA reference is therefore the center of the resized image grid and is not guaranteed to coincide with the transducer or an anatomical center. We use it as a deterministic computational reference. Circular and radial groups encode complementary geometric token neighborhoods around this reference.
Existing tokens within the inscribed circle are ordered into concentric ring sequences, and grid tokens are organized into angular sectors for center-to-periphery radial sequences. Clockwise/counter-clockwise and inward/outward propagation use learned relative-offset weights shared across channels, followed by zero-initialized residual fusion. The original Cartesian selective-scan paths, spatial grid, and token count are preserved. Mathematically, this is an input-independent offset-weighted aggregation within a learned residual branch, not input-adaptive attention or a claim to have invented linear filtering in polar order.
The exact stage-wise grouping and profile configurations are reported in Table G3.
Table R-wgaq-3. EUS geometric organization and alternative implementations.
| Variant | ACC (%) | Recall (%) | F1 (%) | AUC (%) |
|---|---|---|---|---|
| VMamba-Small | 79.9 ± 0.57 | 76.5 ± 2.30 | 79.9 ± 1.00 | 94.1 ± 0.70 |
| Full GCRSA | 81.1 ± 0.41 | 81.2 ± 0.10 | 81.6 ± 0.85 | 95.0 ± 0.61 |
| Uniform Ring/Ray | 77.7 ± 1.09 | 76.1 ± 1.52 | 78.1 ± 1.72 | 93.3 ± 0.47 |
| Polar-transformed Input | 75.6 ± 1.16 | 74.1 ± 2.11 | 76.8 ± 1.98 | 91.1 ± 0.73 |
| Feature-level Polar-SSM | 73.5 ± 2.27 | 71.1 ± 5.09 | 72.7 ± 3.92 | 90.8 ± 0.76 |
Values are percentages. ± denotes run-level SD.
Uniform Ring/Ray and the polar alternatives perform below Full GCRSA in this comparison. The results favor retaining pretrained Cartesian scanning and adding geometric relationships residually.
The geometry was also perturbed directly: four 5% displacements of the computational center give 81.0% mean ACC (−0.1 pp); four 5% input image translations give 80.8% (−0.3 pp) compared with unperturbed GCRSA (81.1%). The image translation affects the whole backbone; the results reflect bounded average sensitivity, within the evaluated perturbation range.
For the measured inference implementation, Eq. (13) is evaluated using cached geometric indices, reused circular transformation intermediates with separate CW/CCW weights, and parallel masked bidirectional radial accumulation and normalization. Valid offsets and boundary conditions are maintained without polar resampling or extra tokens; the exact stage-specific configurations are given in Table G3.
R-wgaq–4. Public-benchmark competitiveness and component attribution
The two public benchmarks primarily test Full CRGA's competitiveness and applicability to additional medical classification tasks. We also evaluated CCR-only and GCRSA-only alongside the original backbone to answer the reviewer's separate request for component-level attribution.
Table R-wgaq-4. Public VMamba-family comparisons (ACC, %).
| Method | Fetal-Planes-DB | OrganAMNIST-224 |
|---|---|---|
| VMamba-Small | 94.8 | 95.2 |
| CCR-only | 95.4 | 96.1 |
| GCRSA-only | 95.2 | 96.2 |
| Full CRGA | 95.5 | 96.3 |
Full CRGA improves over VMamba-Small by +0.7/+1.1 pp on Fetal-Planes-DB/OrganAMNIST-224. Both modules individually exceed the shared backbone on both datasets: CCR-only by +0.6/+0.9 pp, and GCRSA-only by +0.4/+1.0 pp. Thus, component-level improvements are not restricted to EUS. The complete architecture achieves the highest observed ACC on both public benchmarks. Detailed mechanism tests remain centered on the primary EUS task.
R-wgaq–5. EUS variability, paired patient uncertainty, and computational costs
Per-run test accuracies for all four models are provided in Table R-wgaq-5a.
The main EUS dataset includes 1,658 images from 822 patients, with 1,324 training and 334 test images separated at the patient level. For each training fold, the checkpoint with the best validation performance is selected and evaluated on the same patient-separated test cohort, so the five runs are not five independent test populations. The manuscript ACCs for VMamba-Small, CCR-only, GCRSA-only, and Full CRGA are 79.9%, 81.3%, 81.1%, and 82.1%, with sample SDs across the five runs of 0.57, 0.52, 0.41, and 0.16 pp, respectively.
Table R-wgaq-5a. Per-run EUS test accuracy (ACC, %).
| Model | Run 1 | Run 2 | Run 3 | Run 4 | Run 5 | Mean ± SD |
|---|---|---|---|---|---|---|
| VMamba-S (VSSM) | 79.0 | 79.8 | 79.9 | 80.4 | 80.4 | 79.9 ± 0.57 |
| CCR-only | 82.1 | 81.3 | 81.1 | 81.1 | 80.7 | 81.3 ± 0.52 |
| GCRSA-only | 81.4 | 81.4 | 80.7 | 80.6 | 81.4 | 81.1 ± 0.41 |
| Full CRGA | 82.1 | 82.3 | 82.2 | 81.9 | 82.0 | 82.1 ± 0.16 |
Each run corresponds to a model trained on one development fold and evaluated on the same patient-separated held-out test cohort. SD is the sample standard deviation across the five reported runs (ddof = 1).
Table R-wgaq-5b. Selected patient-level paired EUS differences, exact McNemar p-values, and patient-level 95% CIs.
| Comparison | ΔACC (pp) | Exact McNemar p-value | Patient-level 95% CI (pp) |
|---|---|---|---|
| Full CRGA − VMamba-Small | +2.2 | 0.076813 | [−0.24, +4.63] |
| Full CRGA − CCR-only | +0.8 | 0.035156 | [+0.06, +1.62] |
| Full CRGA − GCRSA-only | +1.00 | 0.026604 | [+0.12, +1.88] |
Full CRGA exceeds both individual components, with exact McNemar p-values of 0.035156 versus CCR-only and 0.026604 versus GCRSA-only. The Full CRGA versus VMamba-Small comparison shows an observed +2.2 pp gain, with p=0.076813, close to the 0.05 threshold.
Table R-wgaq-5c. Inference costs at 224 × 224, batch size 1.
| Method | Parameters (M) | Latency (ms) | Peak allocated memory (MB) |
|---|---|---|---|
| VMamba-Small | 49.4 | 9.1 | 223.0 |
| CCR-only | 49.4 | 9.7 | 223.1 |
| GCRSA-only | 49.5 | 13.2 | 225.1 |
| Full CRGA | 49.5 | 13.7 | 225.1 |
Full CRGA adds approximately 0.1 M parameters, 4.6 ms (+50.5%) model-forward latency, and 2.1 MB peak allocated GPU memory at 224 × 224. At 384 × 384, the added latency is 9.4 ms (+74.6%) in the evaluated accelerated inference setup. These measured costs remain within a practical range for the intended computer-aided clinical workflow, where the model is designed to assist physicians.
Response to Reviewer R-wgaq
We appreciate the reviewer's recognition of the architectural hypothesis behind CCR and the preservation of pretrained VMamba computation. The concerns focus on whether directional discrepancy contributes beyond generic modulation, the role and capacity of the one-channel B/C correction, the interpretation of GCRSA's geometric reference, public-data attribution, and reliability and computational cost. We address these issues with the completed targeted controls and explicit implementation details.
R-wgaq–1. Why a single-channel CCR map, and how are direction and state parameters handled?
The single-channel map is a low-capacity spatial conditioning signal, not a replacement for VMamba's original input- and direction-dependent state parameters. At each spatial location u, its learned additive corrections can be viewed schematically as ΔB(u)=WBM(u) and ΔC(u)=WCM(u). The original state parameters and recurrences remain intact; the low-rank correction limits the additional capacity and uses zero-initialized projections to recover pretrained computation at initialization. We do not claim one channel is the globally optimal capacity.
With aligned horizontal and vertical responses H and V, standard additive Cross-Merge F=H+V does not uniquely determine the cross-directional relation: (1,1) and (2,0) have the same sum and different discrepancies. CCR therefore extracts absolute, normalized, and cosine differences before fusion, rather than claiming to recover diagnostic saliency from the merged representation. The map describes difference between directional contextual responses, not which scanning direction is clinically superior or which pixels belong to a lesion.
For a directional scan d, the implemented conditioning is
\widehat B^{\\,d}&=B^d+\mathcal T_d(P_B(M)),\\\\ \widehat C^{\\,d}&=C^d+\mathcal T_d(P_C(M)). \end{aligned}$$ Here $\mathcal T_d$ aligns the common spatial conditioning map with each directional Cross-Scan sequence. The additional projections $P_B$ and $P_C$ are **shared**, while VMamba's original directional state parameters and selective recurrences remain **direction-specific**. This is **direction-aligned conditioning**, while the original directional state parameters and selective recurrences remain direction-specific. **Table R-wgaq-1. EUS conditioning-source, target, and stage controls.** | Variant | ACC (%) | Recall (%) | F1 (%) | AUC (%) | |---|---|---|---|---| | **Original CCR: next-stage B+C** | 81.3 ± 0.52 | 81.5 ± 0.80 | 82.7 ± 0.55 | 93.5 ± 0.57 | | Feature-BC | 79.0 ± 0.44 | 78.0 ± 3.14 | 79.6 ± 1.82 | 94.0 ± 0.63 | | Constant-BC | 80.0 ± 0.64 | 79.3 ± 1.10 | 80.2 ± 1.35 | 93.8 ± 0.69 | | Shuffled-BC | 79.3 ± 0.45 | 76.3 ± 1.45 | 79.0 ± 1.01 | 92.8 ± 0.52 | | Conflict-Feature | 79.9 ± 0.31 | 76.9 ± 0.58 | 79.3 ± 0.40 | 93.9 ± 0.25 | | B-only | 80.1 ± 0.43 | 78.1 ± 1.72 | 80.3 ± 1.38 | 93.3 ± 0.51 | | C-only | 80.3 ± 0.90 | 79.4 ± 1.29 | 80.7 ± 1.27 | 94.0 ± 0.64 | | Same-stage (subsequent block; no feedback) | 78.9 ± 0.68 | 77.8 ± 1.23 | 79.3 ± 0.81 | 93.3 ± 0.40 | *Values are percentages. ± denotes run-level SD.* Feature-BC versus original CCR tests **conditioning information with the same B/C target**; Conflict-Feature tests **direct feature versus B/C intervention with the same discrepancy source**. Constant and Shuffled maps test spatially varying content and alignment. B-only, C-only, and Same-stage further isolate state writing, readout, and intervention location. Original CCR gives the highest observed ACC among these controls. The experiments support the particular combination of **cross-directional information and next-stage B/C conditioning**.