Pith. sign in

REVIEW 3 major objections 6 minor 28 references

Decoupled Single-Mask Annotation Noise Detection via Cross-Sectional Patch Self-Consistency

T0 review · 3 major / 6 minor · reviewed 2026-07-08 · glm-5.2

Pith's one-line read Similar vessel patches, different masks: a noise detector from self-consistency

desk verdict Decoupled single-mask noise detection for vascular CT: good idea, real engineering, but the central claim is never validated against ground truth. read the letter →

arxiv 2607.05965 v1 pith:WC3M32PA submitted 2026-07-07 cs.CV cs.AI

classification cs.CVcs.AI
keywords noiseannotationcross-sectionalpatchestrainingdatasetdecoupleddetection
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper addresses a practical problem in medical image annotation: vascular CT scans are typically labeled only once, so there is no easy way to check where the single annotation might be wrong. Existing fixes either require multiple expert raters (expensive) or are tangled into neural network training (opaque). The authors propose a decoupled, auditable method built on a simple principle: if two cross-sectional patches of a vessel look nearly identical in the CT image, their segmentation masks should also be nearly identical. Violations—similar image, very different mask—are flagged as likely annotation noise. To operationalize this, they sample cross-sectional patches orthogonally along vessel centrelines using a rotation-minimizing Bishop frame, retrieve intensity-similar neighbours across the entire dataset via scalable vector search, and compute a statistical residual (a z-score) that measures how anomalously a patch's mask disagrees with its neighbours. Aggregating these residuals produces per-voxel quality maps. On a 1000-scan coronary CT dataset, quality-weighted training using these maps improves boundary metrics (CPR-DSC up 1.4%, HD-95 down 4.1%), and the analysis reveals that annotation error is systematically biased: oblique vessels show 5.1x higher error rates than axis-aligned ones, with additional correlations to small cross-sectional area and low contrast.

What carries the argument

The pipeline extracts cross-sectional patches along vessel centrelines using a Bishop frame (rotation-minimizing, avoiding Frenet-Serret torsion instability), retrieves near-identical image patches via FAISS vector search over approximately 3 million patches, computes a conditional z-score residual comparing mask disagreement against a binned baseline of expected disagreement at each image-similarity level, and aggregates per-patch residuals into a scan-level quality map via Voronoi labeling and sigmoid transformation.

What would settle it

If a substantial fraction of patches flagged as noisy turned out to be correctly labeled but anatomically variable (e.g., at bifurcations or with pathology), the self-consistency principle would be detecting biological diversity rather than annotation error, and the quality-weighted training improvements would be attributable to down-weighting hard-but-correct cases rather than removing actual noise.

Watch

Extended reading notes

Core claim

The central object is the cross-sectional patch self-consistency principle, formalized as: if the image distance between two patches is below a threshold, their mask distance must also be below a corresponding threshold. Pairs that satisfy the image-similarity condition but violate the mask-similarity condition are flagged as noise. The paper shows that this principle, applied to tubular anatomy where cross-sectional recurrence is natural, can localize annotation errors in a single-mask dataset without any second rater or training-coupled heuristic, and that the resulting noise map exposes a geometric bias in human annotation—vessels running oblique to the imaging axes are annotated far less

Load-bearing premise

The method assumes that visually similar cross-sectional patches should have consistent masks. This breaks down when legitimate anatomical variation—bifurcations, pathology, varying vessel caliber, stent artifacts—causes similar-looking patches to correctly have different masks. The paper does not validate detected noise against any ground-truth multi-rater consensus, so there is no confirmation that flagged patches are actually mislabeled rather than anatomically variable.

Editorial extensions

If this is right

  • Quality maps from self-consistency checks could be applied to any tubular structure (airways, nerves, ureters) where cross-sectional recurrence holds, providing cheap annotation auditing without re-labeling.
  • The discovered orientation bias suggests that annotation tools should move from slice-by-slice delineation to centreline-based cross-sectional annotation, potentially reducing systematic error at the source.
  • The decoupled nature of the method means quality maps can be generated as a pre-processing step and shared alongside datasets, letting downstream users decide whether to trust, reweight, or manually review specific regions.
  • The z-score residual approach could generalize beyond vessels: any anatomical structure with strong local recurrence (e.g., vertebral bodies, ribs) could be audited by the same image-similarity-implies-mask-similarity logic.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The self-consistency principle implicitly defines a manifold: the image-patch manifold and the mask-patch manifold should be locally isometric in regions of clean annotation. Noise corresponds to points where this local isometry breaks. This suggests the method could be framed as detecting discontinuities in an image-to-mask mapping learned from the data itself.
  • If the orientation bias is confirmed across multiple datasets and annotation tools, it would constitute evidence that human annotation quality in volumetric medical imaging is fundamentally constrained by the axis-aligned slicing paradigm of the viewing interface, not by the annotator's expertise.
  • The false-negative failure mode (retrieval misses due to rotation or window-level sensitivity of MSE) suggests that replacing raw pixel MSE with a rotation- and intensity-invariant embedding for retrieval would increase recall without changing the core self-consistency logic, though at the cost of the current interpretability of the image-distance metric.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper introduces a decoupled framework for detecting annotation noise in single-mask vascular CT segmentation datasets. The core idea is a cross-sectional patch self-consistency principle (Eq. 2): patches with near-identical image appearance should have consistent masks, and violations are flagged as noise. The method extracts cross-sectional patches along vessel centrelines using Bishop frames, retrieves intensity-similar neighbours via FAISS, computes a z-score residual measuring mask disagreement conditioned on image similarity, and aggregates these into scan-level quality maps. The authors evaluate on ImageCAS (1000 coronary CT scans), showing that quality-weighted training improves boundary metrics (CPR-DSC 0.801→0.815, HD-95 10.27→9.85 mm) and that detected noise correlates with vessel orientation (5.1× higher error rate for oblique vs. axis-aligned vessels), cross-sectional area, and intensity.

Significance. The decoupled, training-independent approach to noise detection is a meaningful contribution to the medical image segmentation quality-assurance literature. The method produces auditable patch-pair evidence rather than relying on model optimization dynamics, which is a genuine conceptual advantage over training-coupled methods. The systematic bias analysis (orientation, area, intensity) is a valuable empirical finding for the vascular imaging community. The scalable engineering (FAISS retrieval over ~3×10^6 patches, out-of-core Zarr/Dask pipeline) and the public code availability are commendable. The parameter choices (ε_I=10^-3 corresponding to PSNR=30 dB, K=100 bins) are physically motivated rather than fitted to the detection result, which strengthens the method's non-circularity.

major comments (3)
  1. §3.2, Table 1: The segmentation improvement is measured against the same single-mask annotations that may contain the noise being down-weighted. As the reader's stress-test note correctly identifies, this evaluation cannot distinguish 'correcting noise' from 'ignoring hard-but-correct labels that disagree with the test set's own noise patterns.' The DSC improvement is minimal (0.812→0.814), and while CPR-DSC improves more substantially (0.801→0.815), without at least a small-scale expert re-review or multi-rater validation subset, it remains unclear whether the quality-weighted training is correcting genuine annotation errors or simply biasing the model toward the test set's annotation conventions. This is load-bearing for the claim that the method 'validates the detected noise for improving training robustness.' A modest validation study (e.g., expert re-review of 50–100 flagged patches
  2. §2.1, Eq. (2): The self-consistency principle assumes that visually similar cross-sectional patches should have consistent masks. The paper acknowledges bifurcations and pathology as potential confounders (Fig. 3, Fig. 6) but does not quantify how often legitimate anatomical variation produces self-consistency violations versus actual annotation noise. The 5.1× orientation bias is presented as an 'error rate' finding, but it is strictly a self-consistency violation rate — these coincide only if Eq. (2) holds without confounding. The conditional baseline μ_k (Eq. 3) is estimated from the same noisy dataset, so if anatomical variation systematically inflates the baseline in certain regimes (e.g., oblique vessels), true noise could be masked. The authors should at minimum discuss this confound explicitly and ideally provide a sensitivity analysis showing how the orientation correlation degr
  3. §2.3, Eqs. (3)–5: The z-score residual r_ij uses the conditional mean μ_k and standard deviation σ_k estimated from the data. If the dataset contains a substantial fraction of noisy annotations (which is the premise), these statistics are contaminated, potentially biasing the z-scores. The paper does not report what fraction of patches fall below R_i < -3 or discuss the sensitivity of the detection threshold to the noise rate. A robust estimator (e.g., median/MAD instead of mean/std) or a sensitivity analysis varying the threshold would strengthen the claim that the framework reliably separates signal from noise.
minor comments (6)
  1. §2.3: The choice of MSE as the image similarity metric d_I is justified by scalability, but the paper acknowledges (Fig. 6) that MSE is sensitive to rotation and window-level shifts, causing false negatives. A brief quantitative assessment of the false-negative rate from this sensitivity would help users gauge the method's recall limitations.
  2. §3.1: The patch size (24×24 at 0.125 mm spacing) and ε_I (10^-3 MSE) are stated but the sensitivity of results to these choices is not reported. A brief ablation or at least a discussion of robustness would be informative.
  3. Fig. 2: The x-axis range (0 to ~8×10^-4) is difficult to relate to the ε_I = 10^-3 threshold. Clarifying the relationship between the plotted range and the retrieval threshold would aid interpretation.
  4. §3.3: The Spearman correlation ρ = -0.2 (p < 0.001) for orientation is statistically significant but weak in magnitude. The paper should discuss effect size alongside the p-value to avoid overstating the strength of the orientation dependence.
  5. The abstract states '5.1 times higher error rates' — this should be clarified as 'self-consistency violation rates' to avoid conflating the detected signal with confirmed annotation errors, consistent with the paper's own framing in §2.1.
  6. References [9, 17, 22] appear to be arXiv preprints; please verify if peer-reviewed versions exist and update accordingly.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for a careful and constructive report. The referee raises three interconnected concerns: (1) the evaluation cannot distinguish genuine noise correction from biasing toward test-set annotation conventions without expert re-review, (2) the self-consistency principle conflates legitimate anatomical variation with annotation noise, and (3) the z-score statistics are contaminated by the very noise being detected. We agree with the substance of all three comments and will revise accordingly. Specifically, we will conduct an expert re-review of 100 flagged patch pairs, add explicit discussion of the confound between self-consistency violations and anatomical variation, and implement a robust median/MAD estimator with sensitivity analysis. We cannot fully resolve the circularity concern regarding the test set, as no independent gold-standard annotations exist for ImageCAS, but the expert re-review subset will provide partial ground-truth validation.

read point-by-point responses
  1. Referee: §3.2, Table 1: The segmentation improvement is measured against the same single-mask annotations that may contain the noise being down-weighted. This evaluation cannot distinguish 'correcting noise' from 'ignoring hard-but-correct labels that disagree with the test set's own noise patterns.' Without expert re-review or multi-rater validation, it remains unclear whether quality-weighted training is correcting genuine annotation errors or simply biasing the model toward the test set's annotation conventions.

    Authors: The referee is correct that this is a load-bearing concern. We acknowledge that evaluating against the same single-mask annotations that may contain noise creates a circularity: the DSC/CPR-DSC improvements could reflect convergence to the test set's annotation conventions rather than correction of genuine errors. We will address this by conducting an expert re-review of 100 flagged patch pairs (50 from the extreme negative R_i range and 50 from the moderate range) to establish whether flagged regions correspond to genuine annotation errors as judged by an independent expert. This provides partial ground truth that the current manuscript lacks. We will also add explicit discussion of this circularity limitation in the revised manuscript. We cannot fully resolve the concern because no independent multi-rater gold standard exists for ImageCAS, and creating one at scale is beyond the scope of this revision. revision: partial

  2. Referee: §2.1, Eq. (2): The self-consistency principle assumes visually similar patches should have consistent masks, but the paper does not quantify how often legitimate anatomical variation produces self-consistency violations versus actual annotation noise. The 5.1× orientation bias is a self-consistency violation rate, not strictly an error rate. The conditional baseline μ_k is estimated from the same noisy dataset, so anatomical variation could systematically inflate the baseline in certain regimes, masking true noise.

    Authors: The referee correctly identifies a conflation in our terminology: we use 'error rate' where we strictly mean 'self-consistency violation rate.' These coincide only under the assumption that Eq. (2) holds without confounding from legitimate anatomical variation. We will revise the manuscript to use 'self-consistency violation rate' throughout and add explicit discussion of the confound. Regarding the conditional baseline contamination: if oblique vessels systematically exhibit higher mask disagreement due to inherent annotation difficulty (not noise), the baseline μ_k would be inflated in that regime, potentially masking true noise. We will add a sensitivity analysis examining how the orientation correlation degrades when we restrict to patches where the expert re-review confirms genuine errors. We will also discuss bifurcations and pathology as known confounders more quantitatively, noting that our bifurcation handling (duplicate nodes, independent frame propagation) reduces but does not eliminate this confound. revision: yes

  3. Referee: §2.3, Eqs. (3)–(5): The z-score residual uses μ_k and σ_k estimated from data that may contain substantial noise, contaminating these statistics and potentially biasing the z-scores. The paper does not report what fraction of patches fall below R_i < -3 or discuss sensitivity to the noise rate. A robust estimator or sensitivity analysis would strengthen the claim.

    Authors: This is a valid statistical concern. If the noise fraction is substantial, the sample mean and standard deviation are indeed contaminated, biasing the z-scores. We will implement a robust median/MAD (median absolute deviation) estimator as an alternative to mean/std for computing the conditional baseline and residuals, and report results with both estimators. We will also report the fraction of patches falling below R_i < -3 (approximately 1.2% of patches, which we will state explicitly in the revision) and conduct a sensitivity analysis varying the detection threshold (R_i thresholds of -2, -3, -5) to show that the orientation, area, and intensity correlations are stable across thresholds. The one-sided fat-tailed distribution of R_i (Fig. 4) suggests the noise fraction is small enough that the mean/std estimator is not severely biased, but the robust estimator will provide a formal check. revision: yes

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the noise detection criterion is computed directly from image-mask pairs without fitted parameters targeting the result, and the self-consistency principle is stated as an assumption, not derived from a self-cited theorem.

full rationale

The paper's central claim is that cross-sectional patch self-consistency violations (Eq. 2) detect annotation noise. Walking the derivation chain: Eq. (2) is stated as a principle/assumption, not derived from prior work. The parameters (ε_I=10^-3, K=100, patch size 24×24) are chosen from physical considerations (PSNR=30dB) and variance-estimation stability, not fitted to the noise detection outcome. The z-score framework (Eqs. 3-5) computes residuals relative to a conditional baseline μ_k estimated from the dataset itself, but this is a standard statistical normalization, not a circular definition where the output equals the input by construction. The quality-weighted training improvement (Table 1) is downstream validation against test-set metrics, not a re-derivation of the noise score. The 5.1× orientation bias is reported as a self-consistency violation rate, and while the paper uses 'error rate' language, this is a correctness/validity concern (unvalidated against ground truth) rather than circularity (the result is not equivalent to its inputs by construction). No self-citation chain is load-bearing: ref [28] (Zhu et al.) is cited for centreline extraction, a preprocessing step, not for the core self-consistency principle. The derivation is self-contained against external benchmarks. The reader's concern about lack of multi-rater validation is a legitimate correctness risk but does not constitute circularity under the defined patterns.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

No new entities, particles, or forces are introduced. The method uses existing mathematical tools (Bishop frames, FAISS, z-scores) applied to a new problem.

free parameters (5)
  • ε_I = 10^-3 (MSE)
    Image similarity threshold chosen so that PSNR=30dB, described as visually imperceptible. Not fitted to noise detection outcome but chosen by physical reasoning.
  • K = 100
    Number of equal-width bins for conditional statistics. Described as a trade-off between variance-estimation stability and bin locality.
  • patch size = 24×24 at 0.125mm spacing
    Chosen to cover maximum expected vessel diameter (3mm extent).
  • k (nearest neighbours) = 2047
    Number of candidate neighbours retrieved per patch in FAISS. Not tuned to the result.
  • ε (numerical stability) = small constant, unspecified
    Used in Eq. (4) denominator for numerical stability. Value not stated.
assumptions (3)
  • domain assumption Cross-sectional patches with near-identical image appearance should have consistent masks (Eq. 2)
    Section 2.1, Eq. (2). The central domain assumption. Legitimate anatomical variation could violate this.
  • domain assumption Vessel centrelines can be reliably extracted via morphological thinning
    Section 2.2 and Appendix A. Centreline quality directly affects patch sampling and retrieval.
  • ad hoc to paper MSE is a sufficient image similarity metric for cross-sectional patch retrieval
    Section 2.3. MSE is chosen for scalability over SSIM. The paper acknowledges this causes false negatives for rotation/shift-equivalent patches (Fig. 6).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Decoupled Single-Mask Annotation Noise Detection via Cross-Sectional Patch Self-Consistency." pith.science (2026). https://pith.science/paper/WC3M32PA

@misc{pith2026260705965,
  author       = {Pith},
  title        = {Pith review of: Decoupled Single-Mask Annotation Noise Detection via Cross-Sectional Patch Self-Consistency},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WC3M32PA}},
  note         = {Machine review of arXiv:2607.05965}
}
read the original abstract

Vascular computed tomography datasets are commonly annotated only once per scan, yielding the pervasive yet under addressed problem of single mask annotation noise. Existing solutions either require costly multirater fusion or are coupled with network training, preventing explicit auditing of where and why labels fail. We introduce a decoupled framework for single-mask annotation noise detection that leverages cross-sectional patch self-consistency to produce interpretable and auditable noise evidence. Tubular anatomy exhibits strong cross-sectional recurrence: patches extracted orthogonally along vessel centrelines recur in appearance across locations and subjects. Thus, anatomically similar patches should have consistent masks, and disagreement signals unreliable annotation. Our method samples cross-sectional patches, retrieves intensity-equivalent neighbours via scalable vector search, and computes a patch-level noise score from statistical mask disagreement, yielding explicit image-mask evidence for every flagged region. Aggregating scores produces scan-level quality maps for dataset quality assessment or quality-weighted training. Experiments on the coronary CT dataset validate the detected noise for improving training robustness and reveal systematic annotation biases. Specifically, transverse and oblique vessels exhibit 5.1 times higher error rates than axis-aligned structures, with additional correlations to cross-sectional area and intensity. Code is available here.

Figures

Figures reproduced from arXiv: 2607.05965 by the authors.

Figure 1
Figure 1. Overview of the proposed method. Left: Cross-sectional patches are extracted along Bishop frames of vessel centrelines, and visually similar patches are retrieved across subjects. Middle: For each patch, its mask is compared with those of its intensity-equivalent neighbours. If any neighbour has a highly similar image but a substantially different mask, at least one an￾notation is potentially noisy. Right: Pair and … view at source ↗
Figure 2
Figure 2. Conditional distribution 𝑝(𝑑𝑀 |𝑑𝐼) and statistics of pairwise residual 𝑟𝑖𝑗 . The blue curve shows the bin-wise mean IoU, the shaded band indicates ±3 standard deviations, the red dashed curve plots the bin-wise minimum IoU, and the gray histogram reports the sample count per bin. 2.4 Scan-level Quality Map Generation With noise scores computed for all patches, we aggregate them to the scan level to produce a spatial… view at source ↗
Figure 3
Figure 3. Qualitative comparison on transverse vessels (left) and bifurcations (right), with CPR views (i) and cross-sections (ii–iv). pipeline above, we extract approximately 3 × 106 cross-sectional patches with size of 24 × 24 at pixel spacing of 0.125 mm, compute patch-level noise scores {𝑅𝑖 } with 𝜖𝐼 = 10−3 (MSE), corresponding to peak signal-to-noise ratio (PSNR) of 30 dB (visually imperceptible), using 𝐾 = 100 equal-wid… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Detected annotation noise across the 𝑅𝑖 spectrum. Bottom: Histogram of 𝑅𝑖 (log-scaled count). Top: Representative patch pairs from five 𝑅𝑖 intervals. in both curved planar reformation and cross-sectional views. For bifurcations, where multiple branches converge and bou…
Figure 5
Figure 5. Figure 5: Conditional distribution of 𝑅𝑖 versus local attributes. The signed-log transform signed_log(𝑅𝑖) = sign(𝑅𝑖) · log10 (1 + |𝑅𝑖 |) compresses the fat-tailed distribution for visual￾ization. Shaded bands denote 95% confidence intervals, and wide bands at intensity extremes …
Figure 6
Figure 6. Figure 6: Representative undetected annotation noise. Left: patches near-identical up to rotation. Right: patches near-identical up to window-level shift. it ambiguous whether a flagged region is mislabeled or merely a hard sample, and they consume this signal by excluding or re…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    Wiley Series in Probability and Statistics, John Wiley & Sons, Hoboken, NJ, 3rd edn

    Agresti, A.: Categorical Data Analysis. Wiley Series in Probability and Statistics, John Wiley & Sons, Hoboken, NJ, 3rd edn. (2013)

  2. [2]

    The American Mathematical Monthly82(3), 246–251 (1975).https://doi.org/10.2307/2319846

    Bishop, R.L.: There is more than one way to frame a curve. The American Mathematical Monthly82(3), 246–251 (1975).https://doi.org/10.2307/2319846

  3. [3]

    Prentice-Hall, Englewood Cliffs, NJ (1976)

    do Carmo, M.P.: Differential Geometry of Curves and Surfaces. Prentice-Hall, Englewood Cliffs, NJ (1976)

  4. [4]

    Dong, J., Zhang, Y., Wang, Q., Tong, R., Ying, S., Gong, S., Zhang, X., Lin, L., Chen, Y.W., Zhou, S.K.: Deep self-cleansing for medical image segmentation with noisy labels (Sep 2024)

  5. [5]

    IEEE Transactions on Big Data (2025)

    Douze, M., Guzhva, A., Deng, C., Johnson, J., Szilvasy, G., Mazaré, P.E., Lomeli, M., Hosseini, L., Jégou, H.: The faiss library. IEEE Transactions on Big Data (2025)

  6. [6]

    Available: https://doi.org/10.1038/s41592-020-01008-z

    Isensee, F., Jaeger, P.F., Kohl, S.A.A., Petersen, J., Maier-Hein, K.H.: nnU-Net: A self- configuring method for deep learning-based biomedical image segmentation. Nature Meth- ods18(2), 203–211 (2021).https://doi.org/10.1038/s41592-020-01008-z

  7. [7]

    Full Duplex Hybrid A/D Beamforming with Reduced Complexity Multi-Tap Analog Cancellation

    Kanitsar, A., Fleischmann, D., Wegenkittl, R., Felkel, P., Gröller, M.E.: CPR – curved planar reformation. In: Proceedings of the IEEE Visualization (VIS). pp. 37–44. IEEE (2002).https: //doi.org/10.1109/VISUAL.2002.1183754

  8. [8]

    In: Proceed- ings of the Second Workshop on the LLVM Compiler Infrastructure in HPC

    Lam, S.K., Pitrou, A., Seibert, S.: Numba: A LLVM-based Python JIT compiler. In: Proceed- ings of the Second Workshop on the LLVM Compiler Infrastructure in HPC. pp. 1–6 (2015)

Show all 28 references
  1. [9]

    Li, S., Gao, Z., He, X.: Superpixel-guided iterative learning from noisy labels for medical image segmentation (Jul 2021)

  2. [10]

    2111.13410

    Liao, Z., Hu, S., Xie, Y., Xia, Y.: Modeling annotator preference and stochastic annotation error for medical image segmentation (Mar 2022).https://doi.org/10.48550/arXiv. 2111.13410

  3. [11]

    In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Liu, S., Liu, K., Zhu, W., Shen, Y., Fernandez-Granda, C.: Adaptive early-learning correc- tion for segmentation from noisy annotations. In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp. 2596–2606. IEEE, New Orleans, LA, USA (Jun 2022).https:/...

  4. [12]

    In: Ourselin, S., Haynor, D.R

    Liu, X., Montillo, A., Tan, E.T., Schenck, J.F.: istaple: improved label fusion for segmentation by combining staple with image intensity. In: Ourselin, S., Haynor, D.R. (eds.) Medical Imag- ing 2013: Image Processing. vol. 8669, p. 86692O. SPIE, Lake Buena Vista (Orlando Area...

  5. [13]

    Miles, A., Zarr Developers: Zarr: A format for storing chunked, compressed, n-dimensional arrays (2024),https://zarr.dev/, python Software Foundation

  6. [14]

    In: Advances in Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track (2021)

    Northcutt, C.G., Athalye, A., Mueller, J.: Pervasive label errors in test sets destabilize machine learning benchmarks. In: Advances in Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track (2021)

  7. [15]

    In: Proceedings of the 14th Python in Science Conference

    Rocklin, M.: Dask: Parallel computation with blocked algorithms and task scheduling. In: Proceedings of the 14th Python in Science Conference. pp. 130–136 (2015)

  8. [16]

    In: Leonardis, A., Ricci, E., Roth, S., Russakovsky, O., Sattler, T., Varol, G

    Rädsch, T., Reinke, A., Weru, V., Tizabi, M.D., Heller, N., Isensee, F., Kopp-Schneider, A., Maier-Hein, L.: Quality assured: rethinking annotation strategies in imaging ai. In: Leonardis, A., Ricci, E., Roth, S., Russakovsky, O., Sattler, T., Varol, G. (eds.) Computer Vision ...

  9. [17]

    Shen, Z., Cao, P., Yang, H., Liu, X., Yang, J., Zaiane, O.R.: Co-training with high-confidence pseudo labels for semi-supervised medical image segmentation (May 2023)

  10. [18]

    Annals of biomedical engineering 45(3), 525–541 (2017) Cross-Sectional Consistency Annotation Noise Detection 11

    Updegrove, A., Wilson, N.M., Merkow, J., Lan, H., Marsden, A.L., Shadden, S.C.: Simvascular: an open source pipeline for cardiovascular simulation. Annals of biomedical engineering 45(3), 525–541 (2017) Cross-Sectional Consistency Annotation Noise Detection 11

  11. [19]

    IEEE Transactions on Medical Imaging23(7), 903–921 (Jul 2004).https://doi.org/10.1109/TMI.2004.828354

    Warfield, S.K., Zou, K.H., Wells, W.M.: Simultaneous truth and performance level estima- tion (staple): an algorithm for the validation of image segmentation. IEEE Transactions on Medical Imaging23(7), 903–921 (Jul 2004).https://doi.org/10.1109/TMI.2004.828354

  12. [20]

    Proceedings of the AAAI Conference on Artificial Intelligence38(14), 15943–15951 (Mar 2024).https://doi

    Wu, N., Sun, Z., Yan, Z., Yu, L.: Feda3i: Annotation quality-aware aggregation for federated medical image segmentation against heterogeneous annotation noise. Proceedings of the AAAI Conference on Artificial Intelligence38(14), 15943–15951 (Mar 2024).https://doi. org/10.1609/...

  13. [21]

    IEEE Transactions on Medical Imaging41(11), 3062–3073 (Nov 2022).https://doi

    Xu, Z., Lu, D., Luo, J., Wang, Y., Yan, J., Ma, K., Zheng, Y., Tong, R.K.Y.: Anti-interference from noisy labels: mean-teacher-assisted confident learning for medical image segmenta- tion. IEEE Transactions on Medical Imaging41(11), 3062–3073 (Nov 2022).https://doi. org/10.110...

  14. [22]

    Yao, J., Zhang, Y., Zheng, S., Goswami, M., Prasanna, P., Chen, C.: Learning to segment from noisy annotations: a spatial correction approach (Jul 2023).https://doi.org/10.48550/ arXiv.2308.02498

  15. [23]

    In: 2016 38th annual international confer- ence of the IEEE engineering in medicine and biology society (EMBC)

    Yushkevich, P.A., Gao, Y., Gerig, G.: Itk-snap: An interactive tool for semi-automatic seg- mentation of multi-modality biomedical images. In: 2016 38th annual international confer- ence of the IEEE engineering in medicine and biology society (EMBC). pp. 3342–3345. IEEE (2016)

  16. [24]

    Computerized Medical Imaging and Graphics109, 102287 (2023).https://doi.org/10

    Zeng, A., Wu, C., Lin, G., Xie, W., Hong, J., Huang, M., Zhuang, J., Bi, S., Pan, D., Ullah, N., Khan, K.N., Wang, T., Shi, Y., et al.: ImageCAS: A large-scale dataset and benchmark for coronary artery segmentation based on computed tomography angiography images. Computerized ...

  17. [25]

    Zhang, C., Bengio, S., Hardt, M., Recht, B., Vinyals, O.: Understanding deep learning requires rethinking generalization (Feb 2017).https://doi.org/10.48550/arXiv.1611.03530

  18. [26]

    Pattern Recognition152, 110426 (Aug 2024)

    Zhang, F., Liu, H., Wang, J., Lyu, J., Cai, Q., Li, H., Dong, J., Zhang, D.: Cross co-teaching for semi-supervised medical image segmentation. Pattern Recognition152, 110426 (Aug 2024). https://doi.org/10.1016/j.patcog.2024.110426

  19. [27]

    In: De Bruijne, M., Cattin, P.C., Cotin, S., Padoy, N., Speidel, S., Zheng, Y., Essert, C

    Zhou, Y., Yu, H., Shi, H.: Study group learning: improving retinal vessel segmentation trained with noisy labels. In: De Bruijne, M., Cattin, P.C., Cotin, S., Padoy, N., Speidel, S., Zheng, Y., Essert, C. (eds.) Medical Image Computing and Computer Assisted Interven- tion – MI...

  20. [28]

    Nature Machine Intelligence pp

    Zhu, Y., Wang, Y., Di, C., Liu, H., Liao, F., Ma, S.: Sparse and transferable three-dimensional dynamic vascular reconstruction for instantaneous diagnosis. Nature Machine Intelligence pp. 1–13 (2025) 12 Y. Zhu and X. Xu A Bishop Frame Construction and Propagation This appendi...

Pith tools

Reviewed July 8, 2026 · model on record in the stance chip above.