Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Topology-based deep-learning segmentation method for deep anterior lamellar keratoplasty (DALK) surgical guidance using M-mode OCT data

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper's central claim is that adding a star-shape topological loss to a modified U-Net makes M-mode OCT segmentation of the epithelium and Descemet's membrane accurate and fast enough for real-time DALK guidance.

desk verdict A plausible incremental application of a star-shaped topological loss to M-mode OCT corneal segmentation, but the random train/test split undermines the generalization claim. read the letter →

arxiv 2501.04735 v1 pith:QX37HCNZ submitted 2025-01-07 eess.IV cs.CV

J. Yu , H. Yi , Y. Wang , J. D. Opfermann , W. G. Gensheimer , A. Krieger , J. U. Kang This is my paper · ORCID
classification eess.IVcs.CV
keywords deepanteriorlamellarkeratoplastyDALKDescemetmembraneopticalcoherencetomographyM-modeOCTimagesegmentationtopologicallossstarshapeprior
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 is trying to establish that a deep-learning segmentation method built around a topological loss can reliably track two corneal layers—the epithelium and Descemet's membrane—from noisy M-mode OCT signals during robotic DALK surgery. The authors combine a pixel-wise cross-entropy term with a star-shape-prior topological term and a faster modified U-Net, and they report that this hybrid beats the conventional BCE-loss U-Net on every metric and on every data set tested. The motivation is surgical: the Big Bubble technique in DALK requires micron-accurate separation of deep stroma from Descemet's membrane, and real-time guidance needs both accuracy and speed. If the claim holds, the method would make the OCT-guided robot's layer tracking stable enough for clinical use.

What carries the argument

Two pieces carry the argument. The first is the hybrid loss $L_{\mathrm{hybrid}} = \alpha L_{\mathrm{BCE}} + \beta L_T$ with $\alpha/\beta=1/2$: $L_T$ encodes a star shape prior, requiring that along the straight line from any pixel inside the object to the region center, all pixels also belong to the object, which penalizes the holes and rough boundaries that noise produces. The second is a modified U-Net using smaller 3×3 convolution kernels, learnable upsampling with transposed convolutions, and normalization; this speeds up inference without losing spatial resolution. The paper also crops each 512×512 image into 8×512×64 patches for training and reconstructs the patches into the original format during inference.

What would settle it

A leave-one-eye-out test: train on all images from some rabbit eyes, test only on images from a held-out eye, and compare the average Descemet's membrane boundary error; if it exceeds the reported 1.7 µm in vivo by a large margin, the robustness claim as stated for unseen eyes fails.

Watch

Extended reading notes

Core claim

The central claim is that incorporating a star-shape-prior topological loss into a modified U-Net makes corneal layer segmentation from M-mode OCT robust to the noise, signal instability, and motion artifacts that arise when an OCT-fiber needle moves inside the eye. The authors state this as a concrete result: on in vivo, ex vivo, and hybrid rabbit-eye data sets, the proposed method achieved higher SSIM, PSNR, IoU, and Dice coefficients and higher inference frequency (up to 40 Hz) than the conventional method, while reducing average absolute boundary tracking errors—down to 0.54 pixels (1.41 µm) for the epithelium and 0.64 pixels (1.67 µm) for Descemet's membrane in vivo.

Load-bearing premise

The evaluation randomly splits M-mode images into training and test sets without keeping each rabbit eyeball entirely in one side, and the authors acknowledge this can let the network learn from similar frames of the same eye, raising the measured performance beyond what a new eye would give.

Editorial extensions

If this is right

  • The reported inference frequency reaches 35–40 Hz, fast enough to drive real-time robot control during surgery.
  • Boundary tracking errors drop to roughly half a pixel in vivo, about 1.4–1.7 µm, which matches the micron-scale precision the Big Bubble technique needs.
  • The same hybrid-loss design should transfer to other noisy OCT-guided procedures where layer continuity matters more than pixel-level purity.
  • The loss term does not change the annotation protocol, so existing labeled M-mode OCT data can be reused to retrain the network.

Reading between the lines

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

  • The paper does not separate the contributions of the loss and the architecture; an ablation that removes the topological term or reverts to the old U-Net would show which part drives the improvement.
  • Because the evaluation splits images rather than eyeballs, the reported numbers are an upper bound on cross-eye generalization; a leave-one-eye-out evaluation would be a direct test of the generalization claim.
  • The star-shape prior appears naturally suited to other layered or convex anatomical targets in M-mode OCT, such as retinal layers or vessel walls, where the same loss could be reused with minimal changes.
  • If the speed and accuracy numbers hold during needle motion, the robot could plausibly close a control loop on the tracked Descemet's membrane boundary rather than relying on a pre-planned insertion depth.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes a deep-learning segmentation method for M-mode OCT images of the cornea, intended to guide deep anterior lamellar keratoplasty (DALK). The method combines a binary cross-entropy loss with a star-shape-prior topological loss and a modified U-Net architecture. The authors evaluate the method on in vivo, ex vivo, and hybrid rabbit eye data sets, reporting improvements over their previous U-Net with BCE loss in SSIM, PSNR, IoU, Dice, and inference frequency, plus reduced average absolute tracking errors for the epithelium and Descemet's membrane. The paper closes by acknowledging that the random data split may let the network learn from similar samples of the same eyeball.

Significance. If the reported improvements survive a proper held-out-eye evaluation, the method is a useful incremental contribution to real-time OCT-guided surgical robotics: the star-shape topological loss is a principled way to enforce geometric continuity under noise, and the inference speed gain (35–40 Hz vs 25–28 Hz) is practically relevant. The main value is in demonstrating that a topology-aware loss can stabilize layer segmentation in M-mode OCT, a domain where conventional pixel-wise losses produce fragmented boundaries. The principal weakness is that the current evaluation does not demonstrate generalization to unseen eyes, which is the actual clinical requirement. The ablation structure also does not isolate the contribution of the topological loss from the architectural changes.

major comments (3)
  1. [Section 3.1] The train/test split is performed at the level of 512×512 M-mode frames without grouping by individual eyeball, so frames from the same eye appear in both the training and test sets. Since M-mode frames from one eye are highly correlated, the strong metrics in Tables 1 and 2 likely reflect within-eye memorization rather than cross-eye generalization. The authors explicitly concede this in the Discussion: 'the current data set is randomly split, allowing the network to potentially learn from similar samples of the same eyeball.' Because the central claim is robust real-time guidance for DALK, a held-out-eye split (e.g., training on some eyes and testing on entirely different eyes) is a required experiment before the claim can be accepted.
  2. [Section 3.2] The only comparison is against the authors' previous U-Net architecture with BCE loss. Since the proposed method changes both the network architecture and the loss function, the reported improvements cannot be attributed to the topological loss alone. An ablation study is needed: for example, the modified U-Net trained with BCE only, or the previous U-Net trained with the hybrid loss, would isolate the effect of the loss function and the architecture. Without such an ablation, the central claim that the topology-based loss is the source of robustness is not supported.
  3. [Section 3.2, Tables 1 and 2] No error bars, confidence intervals, or statistical significance tests are reported for any metric. The test sets are small (100, 50, and 150 images for in vivo, ex vivo, and hybrid, respectively), and the reported averages could hide substantial run-to-run or sample-to-sample variation. For example, the ex vivo PM and CM IoU values (0.9874 vs 0.9798) differ by less than 0.01, and without variance information it is unclear whether this difference is reproducible. Reporting standard deviations or performing paired significance tests across test images would strengthen the quantitative claims.
minor comments (5)
  1. [Section 2.2, Eq. (3)] The topological loss formula uses the sets O and the center c without explicitly defining them in the text; a precise definition of the star-shape region and how c is determined would improve reproducibility.
  2. [Section 2.2] The paper states α/β = 1/2 but does not specify the actual values of α and β or describe how they were chosen; adding these details would clarify the experimental setup.
  3. [Section 2.3] The statement that the modified U-Net 'adopts a more stable and precise design' is qualitative; quantitative details such as the number of feature channels, depth, and the exact placement of transposed convolutions and normalization are omitted.
  4. [Section 3.2] The video link in Fig. 5 is a placeholder ('http://dx.doi.org/doi.number.goes.here') and should be replaced with the actual DOI or URL.
  5. [References] Reference [1] is cited as the source of the previous U-Net architecture, but the title of [1] ('Comparison of vertical cannula insertion techniques for big bubble deep anterior lamellar keratoplasty') suggests it is a clinical abstract, not an architecture paper; a more appropriate citation for the baseline network should be provided.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the central comparison is an empirical evaluation of a topology-based loss, not a derivation that reintroduces its own fitted inputs.

full rationale

The paper's central claim is empirical: a hybrid loss combining BCE with a topological star-shape prior (Eq. 1) and a modified U-Net outperform a conventional BCE-loss U-Net on in vivo, ex vivo, and hybrid M-mode OCT data. The topological loss itself is imported from an external MICCAI 2018 paper (Mirikharaji and Hamarneh), not derived from the authors' own results, and the weight ratio alpha/beta = 1/2 is a fixed hyperparameter rather than a parameter fitted to the test set and then relabeled as a prediction. The conventional baseline is the authors' own prior U-Net, which is a self-citation, but it is not load-bearing: the comparison's validity rests on the reported test-set metrics, and the topological-loss improvement is not defined in terms of that baseline's outputs. The Discussion explicitly concedes that the random train/test split without grouping by eyeball may let the network learn eye-specific patterns, which is a genuine generalization limitation and a threat to the surgical-guidance claim, but it is a data-evaluation weakness, not circularity: the reported numbers are still computed on held-out images and do not reduce by construction to any fitted parameter or self-referential definition. Therefore no derivation step is equivalent to its own input, and the paper earns a low circularity score reflecting only minor, non-load-bearing self-citation.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The method rests on the geometric assumption that corneal layer regions are star-shaped with respect to a center, plus standard deep learning training assumptions. The loss weights α and β are chosen by hand. No new physical entities are introduced.

free parameters (1)
  • α and β loss weights = α/β = 1/2 (e.g., α=1, β=2)
    Chosen by hand without reported tuning or sensitivity analysis; directly controls the balance between pixel loss and topology loss.
assumptions (3)
  • domain assumption Corneal layer regions in M-mode OCT are star-shaped with respect to a fixed center
    The topological loss enforces that every pixel on a line from the center to any region pixel belongs to the region. This is plausible for thin, roughly horizontal layers but may not hold for curved or detached layers.
  • ad hoc to paper BCE loss plus weighted topological loss appropriately balances pixel accuracy and geometric consistency
    The hybrid loss in Eq. 1 with α/β = 1/2 is a design choice; no ablation or sensitivity analysis is given.
  • domain assumption Patches of 8x512x64 preserve the layer structure relevant for tracking
    The input cropping choice is stated but not justified with experiments.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Topology-based deep-learning segmentation method for deep anterior lamellar keratoplasty (DALK) surgical guidance using M-mode OCT data." pith.science (2026). https://pith.science/paper/QX37HCNZ

@misc{pith2026250104735,
  author       = {Pith},
  title        = {Pith review of: Topology-based deep-learning segmentation method for deep anterior lamellar keratoplasty (DALK) surgical guidance using M-mode OCT data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QX37HCNZ}},
  note         = {Machine review of arXiv:2501.04735}
}
read the original abstract

Deep Anterior Lamellar Keratoplasty (DALK) is a partial-thickness corneal transplant procedure used to treat corneal stromal diseases. A crucial step in this procedure is the precise separation of the deep stroma from Descemet's membrane (DM) using the Big Bubble technique. To simplify the tasks of needle insertion and pneumo-dissection in this technique, we previously developed an Optical Coherence Tomography (OCT)-guided, eye-mountable robot that uses real-time tracking of corneal layers from M-mode OCT signals for control. However, signal noise and instability during manipulation of the OCT fiber sensor-integrated needle have hindered the performance of conventional deep-learning segmentation methods, resulting in rough and inaccurate detection of corneal layers. To address these challenges, we have developed a topology-based deep-learning segmentation method that integrates a topological loss function with a modified network architecture. This approach effectively reduces the effects of noise and improves segmentation speed, precision, and stability. Validation using in vivo, ex vivo, and hybrid rabbit eye datasets demonstrates that our method outperforms traditional loss-based techniques, providing fast, accurate, and robust segmentation of the epithelium and DM to guide surgery.

Figures

Figures reproduced from arXiv: 2501.04735 by the authors.

Figure 1
Figure 1. Example of rough and inaccurate segmentation caused by signal noise and fluctuations. Blue Box: Hollow [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. (a) Geometric illustration of the star shape prior in the topological loss. (b) Modified U-Net architecture. The BCE loss, denoted as LBCE, evaluates the pixel-wise discrepancy between predicted result and ground truth (annotated label), ensuring accurate segmentation performance at the pixel level, which is defined as: LBCE = − X N i=1 [yi · log(ˆyi) + (1 − yi) · log(1 − yˆi)] (2) where yi represents the binary gro… view at source ↗
Figure 3
Figure 3. Comparison of segmentation results between the proposed method (PM) and the conventional method (CM). [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Evaluation pipeline for the average absolute pixel error in corneal layer segmentation. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Video 1-Sample video of real-time segmentation and tracking of corneal layers using the proposed method. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Kalman filter/deep-learning hybrid automatic boundary tracking of optical coherence tomography data for deep anterior lamellar keratoplasty (DALK)

    eess.SP 2025-01 conditional novelty 3.0 of 10

    A Kalman filter applied on top of U-Net OCT segmentation reduces reported epithelium and Descemet's membrane boundary errors by about 60% and 48% on ex vivo rabbit eyes.

Reference graph

Works this paper leans on

10 extracted references · 8 canonical work pages · cited by 1 Pith paper

  1. [1]

    G., Opfermann, J., Wang, Y., Kaluna, J., Krieger, A., and Kang, J

    Gensheimer, W. G., Opfermann, J., Wang, Y., Kaluna, J., Krieger, A., and Kang, J. U., ``Comparison of vertical cannula insertion techniques for big bubble deep anterior lamellar keratoplasty,'' Investigative Ophthalmology & Visual Science 65 (7), 3700--3700 (2024)

    work page 2024

  2. [2]

    U., and Gensheimer, W

    Opfermann, J., Wang, Y., Kaluna, J., Krieger, A., Kang, J. U., and Gensheimer, W. G., ``Novel vertical big bubble needle approach vs. the standard approach for deep anterior lamellar keratoplasty (dalk),'' Investigative Ophthalmology & Visual Science 65 (7), 3684--3684 (2024)

    work page 2024

  3. [3]

    S., Li, K., Yu, J., Liu, X., Kang, J

    Singh, M. S., Li, K., Yu, J., Liu, X., Kang, J. U., et al., ``Live porcine eye model studies of subretinal injection using handheld endoscopy oct integrated injector,'' Investigative Ophthalmology & Visual Science 65 (7), 5499--5499 (2024)

    work page 2024

  4. [4]

    Xu, J., Yu, J., Yao, J., and Zhang, R., ``The neural networks-based needle detection for medical retinal surgery,'' in [ International Conference on Computer Graphics, Artificial Intelligence, and Data Processing (ICCAID 2022) 0.1em ] , 12604 , 674--678, SPIE (2023)

    work page 2023

  5. [5]

    D., Kaluna, J., Gensheimer, B

    Wang, Y., Guo, S., Opfermann, J. D., Kaluna, J., Gensheimer, B. G., Krieger, A., and Kang, J. U., ``Common-path optical coherence tomography guided vertical pneumodissection for dalk,'' in [ Optical Fibers and Sensors for Medical Diagnostics, Treatment and Environmental Applications XXIII 0.1em ] , 12372 , 15--19, SPIE (2023)

    work page 2023

  6. [6]

    Wang, Y., OPTICAL COHERENCE TOMOGRAPHY BASED OPHTHALMIC AND GASTROINTESTINAL SURGICAL GUIDANCE USING DEEP LEARNING , PhD thesis, Johns Hopkins University (2024)

    work page 2024

  7. [7]

    D., Wang, Y., Kang, J

    Kaluna, J., Opfermann, J. D., Wang, Y., Kang, J. U., Gensheimer, W., and Krieger, A., ``A robotic injection system for consistent pneumo-dissection of the deep stroma in big bubble dalk surgery,'' in [ 2024 46th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) 0.1em ] , 1--7, IEEE (2024)

    work page 2024

  8. [8]

    Wang, Y., Opfermann, J., Yu, J., Yi, H., Kaluna, J., Biswas, R., Zuo, R., Gensheimer, W., Krieger, A., and Kang, J., ``Reimagining partial thickness keratoplasty: An eye mountable robot for autonomous big bubble needle insertion,'' arXiv preprint arXiv:2410.14577 (2024)

    arXiv 2024

Show all 10 references
  1. [9]

    Mirikharaji, Z. and Hamarneh, G., ``Star shape prior in fully convolutional networks for skin lesion segmentation,'' in [ Medical Image Computing and Computer Assisted Intervention--MICCAI 2018: 21st International Conference, Granada, Spain, September 16-20, 2018, Proceedings,...

    2018

  2. [10]

    write newline

    " write newline "" before.all 'output.state := FUNCTION blank.sep after.quote 'output.state := FUNCTION fin.entry output.state after.quoted.block = 'skip 'add.period if write newline FUNCTION new.block output.state before.all = 'skip output.state after.quote = after.quoted.blo...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.