Pith. sign in

REVIEW 4 major objections 6 minor 30 references

Adding dynamic structural recovery rates from serial OCT scans improves prediction of visual recovery after macular hole surgery; a multimodal model combining them with clinical data and raw images beats logistic regression by up to 0.12 AU

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

Dynamic recovery-rate features and a cross-attention multimodal network improve classification of postoperative BCVA improvement in macular hole patients, but the gain is partly from same-timepoint structural status rather than purely preoperative data.

T0 review reviewed 2026-08-04 challenge →

load-bearing objection The dynamic recovery-rate idea is genuinely new, but the main claim is undermined by using the same follow-up visit for both the feature and the outcome; this is a solid association study, not a prospective predictor. the 4 major comments →

arxiv 2509.09227 v1 pith:F4QLCXFL submitted 2025-09-11 eess.IV cs.CV

Dynamic Structural Recovery Parameters Enhance Prediction of Visual Outcomes After Macular Hole Surgery

Yinzheng Zhao , Zhihao Zhao , Rundong Jiang , Louisa Sackewitz , Quanmin Liang , Mathias Maier , Daniel Zapp , Peter Charbel Issa
show 1 more author
Mohammad Ali Nasseri
This is my paper
classification eess.IV cs.CV
keywords macular hole surgeryOCT segmentationdynamic recovery parametersvisual acuity predictionmultimodal deep learningBCVAellipsoid zonelongitudinal imaging
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

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 argues that the speed at which retinal structures heal after macular hole surgery carries prognostic information that static measurements miss. It introduces 'dynamic parameters'—recovery rates for the macular hole, pseudocysts, and outer retinal layers—and shows that adding them to logistic regression consistently improves prediction of best-corrected visual acuity improvement, most notably at three months. It then shows that a multimodal deep learning model using clinical variables, extracted features, and raw OCT images outperforms regression at every time point, with the largest AUC difference reaching 0.12. If true, this provides a fully automated, longitudinal framework for personalized postoperative counseling and monitoring.

Core claim

The central claim is that temporal recovery dynamics, not just static morphology, are predictive of functional vision after macular hole surgery. Using a stage-specific segmentation model on longitudinal OCT scans, the authors compute recovery rates for lesion area and outer retinal defect length, then feed these into prediction models for a binary BCVA-improvement outcome. In logistic regression, including dynamic parameters raises accuracy at all four follow-up visits and brings the ellipsoid-zone recovery rate into the significant predictors at three months (OR 1.298 per unit). The multimodal deep learning model, integrating clinical data, extracted parameters, and raw OCT volumes through

What carries the argument

The central new object is the dynamic recovery rate: for each key structure, the ratio of the preoperative lesion extent to the time at which the lesion was observed to have fully resolved, optionally shape-weighted. It turns a sequence of static OCT snapshots into a single speed-of-healing descriptor. The other load-bearing component is the multimodal fusion architecture, which encodes raw OCT images with a pretrained image encoder and fuses them with clinical and feature vectors via cross-attention, letting spatial image information and structured temporal features contribute jointly.

Load-bearing premise

The dynamic recovery rate for a follow-up visit is assumed to be known before the visual acuity measured at that same visit; in the paper it is defined using the time point when the lesion resolved, which is the same examination that supplies the outcome label, so the extra predictive power may come from looking ahead.

What would settle it

Recompute recovery rates using only the preoperative scan and earlier follow-up scans (e.g., predict 3-month BCVA from pre-op and 2-week OCT only) and rerun the logistic regression; if the AUC improvement from dynamic parameters disappears or reverses, the claimed added value is an artifact of temporal leakage rather than genuine prognosis.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Including dynamic recovery rates should become standard in models predicting macular hole surgery outcomes; omitting them costs about 0.03–0.04 AUC.
  • The ellipsoid zone recovery rate at 3 months is a candidate biomarker for photoreceptor reconstitution and intermediate-term visual outcome.
  • Freely combining clinical data, OCT-derived measurements, and raw OCT images yields the strongest predictions; no single modality is sufficient.
  • A fully automated segmentation-to-prediction pipeline is feasible for longitudinal retinal OCT, enabling scalable decision support.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Editorial: The reported 3-month gain for dynamic parameters likely depends on recovery rate being computed from the same visit whose visual outcome is predicted; a prospective formulation using only prior scans may shrink or eliminate the gain.
  • Editorial: Dynamic parameters could be redefined as 'healing velocity over a fixed early window' (e.g., preoperative to 2 weeks) and then tested for predicting later outcomes; this would make the feature truly prospective and clinically actionable.
  • Editorial: A similar recovery-rate construction might transfer to other surgeries with serial OCT follow-up, such as retinal detachment repair or epiretinal membrane peeling, where structural healing speed may track functional recovery.
  • Editorial: Because the outcome threshold was set at 20 ETDRS letters rather than the usual 15, the absolute AUC values may not be comparable to studies using standard definitions; the relative ordering of models is the safer conclusion.
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

4 major / 6 minor

Summary. The paper proposes a fully automated pipeline for longitudinal OCT segmentation and feature extraction in idiopathic full-thickness macular hole (iFTMH), introduces a category of 'dynamic recovery parameters' derived from serial OCT, and evaluates their contribution to logistic regression and a multimodal deep learning model for predicting postoperative BCVA improvement at 2 weeks, 3, 6, and 12 months. The central claims are: (i) a stage-specific nnUNet segmentation model with mean Dice 0.862; (ii) adding dynamic recovery rates improves logistic regression AUC, especially at 3 months; and (iii) a multimodal DL model combining clinical data, quantitative features, and raw OCT images outperforms logistic regression, with AUC differences up to 0.12. The paper is framed as a clinical decision-support tool for personalized postoperative management.

Significance. If the predictive claims are valid, the paper would make a useful contribution: it applies a fully automated segmentation/extraction pipeline to a public longitudinal dataset, introduces a clinically plausible class of dynamic structural features, and benchmarks a multimodal DL architecture against classical regression. Strengths include the use of a public dataset, explicit comparison of models with and without dynamic parameters, reporting of ORs with confidence intervals in Table 3, and ablation-style comparisons across data modalities. However, the central claim that dynamic parameters 'enhance prediction' depends critically on whether those parameters are available before the outcome they are used to predict. As written, the dynamic recovery rate is defined using the time point at which the lesion is observed to be fully resolved, which is only known at or after the same follow-up visit that supplies the BCVA label. Unless a temporal offset is introduced and documented, the reported AUC gains are concurrent structure-function associations, not prospective predictions. This is the load-bearing issue for the manuscript's stated decision-support contribution.

major comments (4)
  1. [Methods, Automated Feature Quantification; Table 3; Figure 4] The key concern is temporal circularity. The recovery rate is defined as 'the ratio between the initial lesion size on preoperative OCT and the time point at which the lesion was observed to be fully resolved.' For the 3-month model, for example, time-to-resolution is determined by observing visits up to and including the 3-month visit, and the outcome label is BCVA improvement at the same 3-month visit. Thus, recovery rate at month 3 is not known before the month-3 BCVA; the model is classifying the same visit using outcome-derived information, not predicting it. The Methods statement that the model is 'based on preoperative parameters' contradicts the dynamic feature definition. No equations or feature-timing offsets are provided. Please provide the exact definitions of each dynamic parameter, specify the time window used to compute them, and either (a) enforce a strict temporal split
  2. [Methods, Multimodal Deep Learning Prediction Model; Figure 4] The DL model accepts raw OCT images and dynamic parameters from the same follow-up time point at which the BCVA-improvement label is defined. If the input at month t includes images and features from month t, the model is not predicting month-t BCVA from preoperative or earlier data. This is especially relevant because the reported AUC gain from adding dynamic parameters (0.03-0.04) and the headline DL-vs-regression gap (0.12) may largely reflect leakage of the outcome into the feature set. Please clarify, for each evaluation timepoint, exactly which visits' images and features are used as inputs and which visit's BCVA is the label. If no temporal offset is used, the DL comparison is still informative as a concurrent classification benchmark, but the decision-support claim in the Conclusion must be revised.
  3. [Abstract; Methods, Segmentation Model; Table 1] There is a direct numerical inconsistency: the Abstract states 'mean Dice > 0.89', while the Methods text reports 'the average Dice reached 0.862' and Table 1 reports mean Dice 0.8616. The supportive structures achieve Dice above 0.89 individually, but the global mean does not. Please correct the abstract or report a different summary statistic (e.g., structure-specific range).
  4. [Results, Logistic Regression Model and Figure 4; Table 2] AUC values and AUC differences (e.g., 0.94 vs 0.87; exclusion of DP reducing AUC by 0.03-0.04) are presented without confidence intervals or significance tests. Given the moderate sample size and the use of a single public dataset, it is not possible to assess whether the reported improvements are statistically reliable or within sampling variability. Please report CIs for AUCs, p-values for AUC comparisons (e.g., DeLong), and the number of eyes with complete data per timepoint.
minor comments (6)
  1. [Methods, Automated Feature Quantification] The shape-weighted factor is introduced as 'With permitted data, we further incorporated a shape-weighted factor...' but no formula, coefficient, or sensitivity analysis is provided. Please define it precisely or state that it was not used in the reported experiments.
  2. [Methods, Logistic Regression Model and Dataset] The number of eyes and the number of events per outcome class are not reported. This is essential for interpreting the stability of multivariable logistic regression with several predictors. Please add these numbers.
  3. [Results, Logistic Regression Model of Feature Parameters] Typo: 'inclusion of diseases recovery rates' should be 'inclusion of disease recovery rates.' Also, the univariate screening is described as p<0.10 in Methods but the Results text reports P<0.05; please clarify the threshold used.
  4. [Table 3] The 3-month model reports Recovery Rate (Macular Hole) with P=0.050 and Recovery Rate (EZ) with P=0.048. These are borderline and should be described as such; the abstract's statement that dynamic parameters are significant at P<0.05 is only marginally supported. Also, the text 'For each additional 1.0 μm increase in BD' is awkward for an OR of 0.968 per μm; reporting per 100 μm would be more clinically meaningful.
  5. [Methods, Multimodal Deep Learning Prediction Model] The BCVA improvement threshold is set at 20 ETDRS letters, but the distribution of Superior vs. not-Superior outcomes is not reported. Without class counts, accuracy and AUC are difficult to interpret. Please add this information.
  6. [Discussion] The claim that EZ recovery 'may begin earlier than typically appreciated' is speculative: the dynamic parameter as defined is a ratio of initial size to time of resolution, not a trajectory. Please either provide trajectory-level evidence or temper the claim.

Circularity Check

1 steps flagged

Dynamic recovery-rate features are computed from the same postoperative visit (or later) that supplies the BCVA label, so the reported AUC gains measure concurrent structure–function association, not prospective prediction.

specific steps
  1. fitted input called prediction [Methods, 'Automated Feature Quantification and Dynamic Parameter Derivation'; Methods, 'Multimodal Deep Learning Prediction Model'; Results, Table 3]
    "For metrics such as macular hole area and outer retinal defect length, recovery rate was defined as the ratio between the initial lesion size on preoperative OCT and the time point at which the lesion was observed to be fully resolved. ... The purpose of the model is to predict postoperative BCVA improvement based on preoperative parameters."

    The dynamic recovery rate uses 'the time point at which the lesion was observed to be fully resolved' as its denominator. That time point is only known by observing the same postoperative follow-up visits that also define the BCVA-improvement label (e.g., 3 months). If the lesion is resolved by 3 months, the recovery-rate feature is a same-visit status; if it resolves later, the feature encodes information from after the 3-month prediction point. Either way, the feature is not available before the outcome it is said to predict. The paper's own framing says the model predicts 'based on preoperative parameters,' but the dynamic parameters are derived from postoperative follow-up observations. No temporal split or feature lag is provided. Therefore the incremental AUC/OR gains attributed to d

full rationale

The central added-value claim of the paper is that dynamic structural recovery parameters enhance postoperative BCVA prediction. That claim is undermined by the operational definition of recovery rate, which depends on the time point of observed resolution—information obtained during the same follow-up window as the BCVA label. Thus the reported AUC improvement from including dynamic parameters is not evidence of a before-the-fact predictive model. This is a partial circularity: the feature is derived from the very temporal information the model claims to predict. However, the paper also contains independent content: static features (BD, MLD, EZ defect length) and raw OCT images contribute to the models, and the segmentation performance is evaluated against its own annotations. The static-feature associations are not circular, only the dynamic-parameter increment. No load-bearing self-citation was found: the cited prior work (Godbout et al., refs 11 and 26) is the data source and does not overlap with the current authors. The dynamic-parameter problem is the primary reason the central claim cannot be taken at face value as prospective prediction, warranting a score of 6.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 0 invented entities

The ledger shows the central reliance on domain biomarkers and the temporal assumption that recovery rates are available as pre-outcome predictors. The main free parameters are the arbitrary BCVA threshold and the underspecified shape-weighting coefficient. No new physical entities are introduced.

free parameters (2)
  • BCVA improvement threshold = 20 ETDRS letters
    The binary outcome (Superior vs not) uses an author-selected cutoff of 20 instead of the typical 15; changing it redefines the classification task and all reported AUCs.
  • Shape-weighted factor in dynamic recovery rate = Unspecified coefficient
    The paper mentions a shape-weighted factor and a coefficient-weighted adjustment model for dynamic parameters but gives no formula or fitting procedure, so the numeric definition of the core novelty is unknown.
axioms (4)
  • domain assumption nnU-Net segmentation output can be used as ground-truth-equivalent measurements for feature extraction.
    The full pipeline relies on automatic segmentations (mean Dice 0.862) being accurate enough to compute MLD, BD, area, and defect lengths; no error propagation into downstream predictors is reported.
  • domain assumption OCT morphometric measures (MLD, BD, EZ/ELM defect length) are established biomarkers of visual recovery.
    The paper builds on prior literature for these structure-function associations; this is a reasonable domain assumption but is imported, not derived.
  • ad hoc to paper A recovery rate at follow-up visit t is known before the BCVA outcome at visit t.
    Methods define recovery rate using preoperative size and the time at which the lesion was observed resolved, which uses post-operative information at the same visit as the label. This temporal availability is assumed and is the load-bearing premise for calling the models predictive.
  • ad hoc to paper A shape-weighted coefficient can be included 'with permitted data' without bias.
    The dynamic-parameter paragraph grants an unspecified adjustment model; the conditions under which this factor is included are not defined, making the feature set partially unpredictable.

reviewed 2026-08-04 · how reviews work

0 comments
Cite this review

Pith. "Pith review of Dynamic Structural Recovery Parameters Enhance Prediction of Visual Outcomes After Macular Hole Surgery." pith.science (2026). https://pith.science/paper/F4QLCXFL

@misc{pith2026250909227,
  author       = {Pith},
  title        = {Pith review of: Dynamic Structural Recovery Parameters Enhance Prediction of Visual Outcomes After Macular Hole Surgery},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/F4QLCXFL}},
  note         = {Machine review of arXiv:2509.09227}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Purpose: To introduce novel dynamic structural parameters and evaluate their integration within a multimodal deep learning (DL) framework for predicting postoperative visual recovery in idiopathic full-thickness macular hole (iFTMH) patients. Methods: We utilized a publicly available longitudinal OCT dataset at five stages (preoperative, 2 weeks, 3 months, 6 months, and 12 months). A stage specific segmentation model delineated related structures, and an automated pipeline extracted quantitative, composite, qualitative, and dynamic features. Binary logistic regression models, constructed with and without dynamic parameters, assessed their incremental predictive value for best-corrected visual acuity (BCVA). A multimodal DL model combining clinical variables, OCT-derived features, and raw OCT images was developed and benchmarked against regression models. Results: The segmentation model achieved high accuracy across all timepoints (mean Dice > 0.89). Univariate and multivariate analyses identified base diameter, ellipsoid zone integrity, and macular hole area as significant BCVA predictors (P < 0.05). Incorporating dynamic recovery rates consistently improved logistic regression AUC, especially at the 3-month follow-up. The multimodal DL model outperformed logistic regression, yielding higher AUCs and overall accuracy at each stage. The difference is as high as 0.12, demonstrating the complementary value of raw image volume and dynamic parameters. Conclusions: Integrating dynamic parameters into the multimodal DL model significantly enhances the accuracy of predictions. This fully automated process therefore represents a promising clinical decision support tool for personalized postoperative management in macular hole surgery.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

30 extracted references · 2 linked inside Pith

  1. [1]

    Ullrich, C

    S. Ullrich, C. Haritoglou, C. t. Gass, et al. Macular hole size as a prognostic factor in macular hole surgery. British Journal of Ophthalmology. 2002; 86: 390-393

    2002

  2. [2]

    S.-C. Bu, R. Kuijer, X.-R. Li, J. M. Hooymans and L. I. Los Idiopathic epiretinal membrane. Retina. 2014; 34: 2317-2335

    2014

  3. [3]

    J. S. Duker, P . K. Kaiser, S. Binder, et al. The International Vitreomacular Traction Study Group classification of vitreomacular adhesion, traction, and macular hole. Ophthalmology. 2013; 120: 2611-2619

    2013

  4. [4]

    Mitry, J

    D. Mitry, J. Singh, D. Yorston, et al. The predisposing pathology and clinical characteristics in the Scottish retinal detachment study. Ophthalmology. 2011; 118: 1429-1434

    2011

  5. [5]

    Tsuboi, A

    K. Tsuboi, A. Fukutomi, H. Sasajima, et al. Visual acuity recovery after macular hole closure associated with foveal avascular zone change. Translational Vision Science & Technology. 2020; 9: 20-20

    2020

  6. [6]

    H. L. Brooks Jr Macular hole surgery with and without internal limiting membrane peeling. Ophthalmology. 2000; 107: 1939-1948

    2000

  7. [7]

    Kadonosono, F

    K. Kadonosono, F. Yazama, N. Itoh, et al. Treatment of retinal detachment resulting from myopic macular hole with internal limiting membrane removal. American journal of ophthalmology. 2001; 131: 203-207

    2001

  8. [8]

    S. A. Madreperla, G. L. Geiger, M. Funata, Z. de la Cruz and W. R. Green Clinicopathologic correlation of a macular hole treated by cortical vitreous peeling and gas tamponade. Ophthalmology. 1994; 101: 682-686

    1994

  9. [9]

    Kusuhara, M

    S. Kusuhara, M. F. T. Escaño, S. Fujii, et al. Prediction of postoperative visual outcome based on hole configuration by optical coherence tomography in eyes with idiopathic macular holes. American journal of ophthalmology. 2004; 138: 709-716

    2004

  10. [10]

    Ruiz-Moreno, C

    J. Ruiz-Moreno, C. Staicu, D. Piñero, et al. Optical coherence tomography predictive factors for macular hole surgery outcome. British Journal of Ophthalmology. 2008; 92: 640-644

    2008

  11. [11]

    Godbout, A

    M. Godbout, A. Lachance, F. Antaki, A. Dirani and A. Durand Predicting Visual Improvement after Macular Hole Surgery: a Cautionary Tale on Deep Learning with Very Limited Data. arXiv preprint arXiv:2109.09463. 2021

    Pith/arXiv arXiv 2021

  12. [12]

    Steel, P

    D. Steel, P . Donachie, G. Aylward, et al. Factors affecting anatomical and visual outcome after macular hole surgery: findings from a large prospective UK cohort. Eye. 2021; 35: 316-325

    2021

  13. [13]

    Baumann, D

    C. Baumann, D. Iannetta, Z. Sultan, et al. Predictive association of pre-operative defect areas in the outer retinal layers with visual acuity in macular hole surgery. Translational Vision Science & Technology. 2021; 10: 7-7

    2021

  14. [14]

    de Sisternes, J

    L. de Sisternes, J. Hu, D. L. Rubin and T. Leng Visual prognosis of eyes recovering from macular hole surgery through automated quantitative analysis of spectral-domain optical coherence tomography (SD-OCT) scans. Investigative Ophthalmology & Visual Science. 2015; 56: 4631-4643

    2015

  15. [15]

    Ozturk, A

    Y . Ozturk, A. Ağın, N. Kockar, et al. The Importance of Anatomic Configuration and Cystic Changes in Macular Hole: Predicting Surgical Success with a Different Approach. Current Eye Research. 2022; 47: 1436-1443. 24

    2022

  16. [16]

    Kanno, T

    J. Kanno, T. Shoji, H. Ishii, et al. Idiopathic Macular Hole Area to Foveal Avascular Zone Ratio and Its Effects on Visual Acuity Before and After Surgery. Translational Vision Science & Technology. 2025; 14: 22-22

    2025

  17. [17]

    J. Yang, H. Xia, Y . Liu, et al. Ellipsoid zone and external limiting membrane-related parameters on spectral domain-optical coherence tomography and their relationships with visual prognosis after successful macular hole surgery. Frontiers in Medicine. 2021; 8: 779602

    2021

  18. [18]

    D. Xu, A. Yuan, P . K. Kaiser, et al. A novel segmentation algorithm for volumetric analysis of macular hole boundaries identified with optical coherence tomography. Investigative ophthalmology & visual science. 2013; 54: 163-169

    2013

  19. [19]

    Zgolli, C

    H. Zgolli, C. Abdelhedi, S. Mabrouk, et al. Prognostic factors for visual recovery after successful large macular hole surgery using the inverted flap technique. Journal Français d'Ophtalmologie. 2023; 46: 1069-1078

    2023

  20. [20]

    M. Roth, N. Schön, L. Jürgens, et al. Frequently assessed and used prognostic factors for outcome after macular hole surgery: which is better? BMC ophthalmology. 2021; 21: 1-11

    2021

  21. [21]

    Feroz, S

    L. Feroz, S. F. Rizvi, S. Naz and M. T. H. Khan Pre-operative predicting factor in visual outcome after macular hole surgery. Pakistan Journal of Medical Sciences. 2020; 36: 1053

    2020

  22. [22]

    Isensee, P

    F. Isensee, P . F. Jaeger, S. A. Kohl, J. Petersen and K. H. Maier-Hein nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nature methods. 2021; 18: 203-211

    2021

  23. [23]

    Wakely, R

    L. Wakely, R. Rahman and J. Stephenson A comparison of several methods of macular hole measurement using optical coherence tomography, and their value in predicting anatomical and visual outcomes. British Journal of Ophthalmology. 2012; 96: 1003-1007

    2012

  24. [24]

    K. Han, Y . Wang, H. Chen, et al. A survey on visual transformer. arXiv preprint arXiv:2012.12556. 2020

    Pith/arXiv arXiv 2012

  25. [25]

    B. Qi, Y . Yu, X. Yang, et al. The Evolution and Visual Prognosis of Glial Proliferation with Different Grades after Macular Hole Surgery: An Optical Coherence Tomography-Based Study. Ophthalmologica. 2023; 246: 314-323

    2023

  26. [26]

    Lachance, M

    A. Lachance, M. Godbout, F. Antaki, et al. Predicting visual improvement after macular hole surgery: a combined model using deep learning and clinical features. Translational Vision Science & Technology. 2022; 11: 6-6

    2022

  27. [27]

    Baumann, S

    C. Baumann, S. Hoffmann, A. Almarzooqi, et al. Defining a cutoff for progression of macular holes. Translational vision science & technology. 2021; 10: 2-2

    2021

  28. [28]

    Geenen, D

    C. Geenen, D. C. Murphy, M. T. Sandinha, J. Rees and D. H. Steel Significance of preoperative external limiting membrane height on visual prognosis in patients undergoing macular hole surgery. Retina. 2019; 39: 1392-1398

    2019

  29. [29]

    Landa, R

    G. Landa, R. Gentile, P . Garcia, T. Muldoon and R. Rosen External limiting membrane and visual outcome in macular hole repair: spectral domain OCT analysis. Eye. 2012; 26: 61-69

    2012

  30. [30]

    Bottoni, S

    F. Bottoni, S. De Angelis, S. Luccarelli, M. Cigada and G. Staurenghi The dynamic healing process of idiopathic macular holes after surgical repair: a spectral-domain optical coherence tomography study. Investigative ophthalmology & visual science. 2011; 52: 4439-4446

    2011

This paper was first reviewed by deepseek-v4-flash on August 4, 2026.