{"id":"bc4a893d-0b14-4d7a-a8dd-bf0454e6e76a","arxiv_id":"2504.21765","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Subject-specific eye models from two healthy subjects show that sclera stiffness dominates optic nerve head mechanics, while Bruch's membrane can be omitted without meaningful changes in predicted strains.","lead":"Computer models of the back of the eye, built from scans of two healthy people, show which tissue layers really matter for predicting how the optic nerve's support mesh deforms under pressure. The outer sclera dominates the response, while the thin Bruch's membrane is unimportant, offering a simpler template for future glaucoma research models.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The recommendation to exclude Bruch's membrane is based on a single pressure condition; its effect across the IOP/ICP validation grid is never tested, so the 'accurate representation' claim is not fully supported.","rationale":"The reader's weakest assumption concerned unrepresentative literature-based material properties and estimated dimensions of non-visible tissues. My concern is related but distinct: even if those property and geometry inputs are correct, the tissue-necessity ranking is established at a single pressure pair. The binary removal test for Bruch's membrane was performed only at IOP=20/ICP=10, and the reduced model was then used to compute the pressure-response validation without ever being compared against the full model across the IOP/ICP grid. This leaves open the possibility that Bruch's membrane has a negligible effect at baseline but a non-negligible effect at other pressures, which would undermine the central claim that the reduced model provides accurate representation across the tested range. The proposed test is purely computational and directly settles this concern by quantifying the removal effect across all nine pressure combinations. If the effect remains below 5% everywhere, the recommendation is robust for the stated loading range; if not, the conditional verdict should be strengthened to require either inclusion of Bruch's membrane or an explicit pressure-range restriction on the claim. The paper's overall contribution as a sensitivity framework remains valuable, so the verdict should stay conditional rather than move to accept or reject.","tokens_in":19862,"tokens_out":7067,"duration_ms":73646,"concrete_test":"Re-run the binary removal test for Bruch's membrane across the full pressure grid used in Section 4 (IOP = 10, 20, 30 mmHg; ICP = 5, 10, 15 mmHg; nine combinations) and compute the relative difference in average and peak LC subregion strains between the full model and the reduced model without Bruch's membrane. If any difference exceeds 5%, or if the difference varies substantially from the 1.2% baseline value, then the recommendation to exclude Bruch's membrane is not robust across the loading conditions of interest.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central recommendation to omit Bruch's membrane rests on one binary removal test performed at a single loading condition (IOP=20 mmHg, ICP=10 mmHg; Section 1, Figure 6), which showed an average LC strain change of about 1.2%. Section 4 then uses the reduced model (without Bruch's membrane) to evaluate LC strains across IOP 10-30 mmHg and ICP 5-15 mmHg, but never re-tests the effect of Bruch's membrane under these varying pressures. The negligibility of Bruch's membrane could be load-dependent: its mechanical role in transmitting IOP and ICP may change as the pressure differential across the eye changes, and the binary removal test gives no information about behavior outside the baseline condition. Without demonstrating that the 1.2% effect is representative across the pressure range used for validation, the claim that 'accurate ONH mechanical representation can be obtained' without Bruch's membrane is not established. The validation in Section 4 compares trends only qualitatively against literature and does not compare the reduced model against the full model, so it cannot detect whether omitting Bruch's membrane introduces pressure-dependent errors.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper builds subject-specific finite element models of the optic nerve head from SS-OCT imaging of two healthy subjects, assigning distinct, literature-based material properties to all major ocular tissues, including several non-visible regions whose geometry is estimated from published data. The authors perform local sensitivity analyses (±25% modulus variations) and binary tissue-removal tests to rank the influence of each tissue on strains in the lamina cribrosa, retina, and optic nerve. They report that scleral stiffness has the largest overall impact and that Bruch's membrane has a negligible effect, then construct a reduced model without Bruch's membrane and evaluate LC strains under an IOP/ICP grid (IOP 10-30 mmHg, ICP 5-15 mmHg), comparing the resulting trends qualitatively with literature. The central claim is that accurate ONH mechanical representation can be obtained by including only the tissues identified as necessary, with scleral stiffness as the most influential input.","tokens_in":19996,"tokens_out":3410,"duration_ms":37334,"significance":"If the claim is established, the paper offers practical guidance for ONH modeling: future in vivo subject-specific models could omit Bruch's membrane and prioritize accurate scleral properties without materially changing predicted LC strains. The study is clearly described, uses clinically plausible imaging data, and does not fit any parameter to the target strain outcomes, avoiding circularity in the sensitivity ranking. The pressure-response trends are consistent with several published studies. However, the significance is tempered by the small sample (two subjects, one shown), the reliance on literature-based geometry and material properties for non-visible tissues, the qualitative nature of the validation, and the fact that the key binary-removal test for Bruch's membrane is performed at a single loading condition.","major_comments":[{"comment":"The binary removal test for Bruch's membrane is performed only at the baseline loading condition (IOP=20 mmHg, ICP=10 mmHg), where it changes average LC strains by about 1.2%. Section 4 then uses the reduced model over IOP 10-30 mmHg and ICP 5-15 mmHg without re-testing the effect of omitting Bruch's membrane at those pressures. Because the mechanical role of a thin shell may vary with the pressure differential across the eye, the claim that accurate ONH representation omits Bruch's membrane is not supported outside the baseline state. The authors should run the binary removal test across the IOP/ICP grid and report the range of strain change, or explicitly restrict the conclusion to the tested condition.","section":"Section 'Effect of Materials on Lamina Cribrosa Mechanics' (Figure 6) and Section 4"},{"comment":"The validation compares only qualitative trends (strain increases with IOP and ICP, highest strains in inferior and temporal subregions) against literature studies with different loading protocols and species; there is no quantitative comparison and no comparison of the reduced model against the full model on the same pressure grid. The phrase 'accurate ONH mechanical representation' is therefore stronger than the evidence: a full-model/reduced-model comparison would directly quantify the error introduced by omitting Bruch's membrane and would support the central claim.","section":"Section 4 (Effect of Pressure Variations on Lamina Cribrosa Mechanics)"},{"comment":"The necessity ranking is based on two subjects' eyes, and results from only one subject are shown. The statement that 'all significant observations were consistent for both subjects' is asserted but not documented. Please report the second subject's sensitivity and binary-removal results, for example in supplementary material, or prominently state this limitation in the abstract and conclusions; as written, the generalizability of the tissue-necessity ranking is not verifiable.","section":"Methods, Data Acquisition and first paragraph of Results"},{"comment":"The posterior LC boundary and the geometry of the pia mater, dura mater, Bruch's membrane, border tissue, and annular ring are estimated from literature rather than measured for the subjects, and no sensitivity analysis is performed on these geometric parameters. Since the main recommendation (omitting Bruch's membrane) depends on the assumed thickness and position of a non-visible tissue, the robustness of the ranking to the assumed geometry should be tested (e.g., thickness variations) or explicitly acknowledged as a limitation in the abstract and conclusions.","section":"Methods Section 2 and Table 1"}],"minor_comments":[{"comment":"Equation 1 uses an empirical formula for axial length derived from spherical equivalent; please state whether the resulting axial length was compared with any available biometric data for the two subjects, or note that it is an estimate.","section":"Equation 1"},{"comment":"The table in Figure 3 reports strain values as percentages, but the units are not explicitly stated in the table caption; please add a clear statement that all strain values are in percent.","section":"Table in Figure 3"},{"comment":"The references list Feola et al. 2016a and 2016b with nearly identical titles and journal information; please verify whether these are the same paper cited twice, and if so, consolidate the citations.","section":"References (Feola et al., 2016a and 2016b)"},{"comment":"Figure 6 plots a single average of the relative change in compressive and tensile strains, but some removed tissues affect these strain components differently; please report the tensile and compressive changes separately or explicitly state that the panel shows their average.","section":"Figure 6 description"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for a biomechanics journal and the modeling pipeline is plausible, with a clear sensitivity framework and no parameter fitting to the target outcomes. The main risk is overclaiming from n=2 and a qualitative validation; the most actionable request is to re-test the Bruch's membrane removal across the IOP/ICP grid and to document the second subject's results. If those additions are made, the recommendation could move to minor revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead the ONH modeling paper. The useful thing here is a systematic sensitivity framework: for two in vivo subject-specific geometries, they vary each tissue's modulus and also do binary removal tests for the low-impact tissues, then report effects on LC, retina, and optic nerve strains. The findings are not brand new—they confirm scleral stiffness dominance and Bruch's membrane's minor role noted by Sigal and others—but the structure is cleaner and more explicit than most prior work, and the method is transparent enough to reproduce.\n\nThe core weakness is the gap between what the data support and what the abstract claims. The binary removal test shows Bruch's membrane changes LC strain by about 1.2% at one loading condition (IOP 20, ICP 10). That is the entire basis for omitting it. The later IOP/ICP sweep is run with the reduced model (no Bruch's membrane) and validated only qualitatively against literature trends. The stress-test note is right: the effect of Bruch's membrane could be pressure-dependent, and this design never checks it. A simple follow-up re-running the removal test across the same 3x3 pressure grid would settle it. Until that is done, \"accurate representation\" overstates the evidence; \"likely not required\" is fine, which is actually what the conclusions mostly say.\n\nOther soft spots are real but more standard: n=2 with results from one subject shown; non-visible tissues (pia, dura, Bruch's, border tissue, annular ring) are assigned literature dimensions and properties rather than measured; no UQ; no data or code released. The anisotropy tests are mentioned but not shown, which is mildly annoying given the paper claims they did not change the rankings. None of these sink the paper, but they cap its strength. The material properties are not fit to the outcomes being predicted, and the pressure-response trends do match several independent experimental and computational studies, so the circularity burden is low.\n\nI'd send this to a serious referee. The modeling community gets a clear, reusable protocol for deciding which tissues are needed, and the pressure sweep is a useful benchmark. The fix is straightforward: test the necessity decision across the pressure range, report the second subject in supplementary form, and soften the abstract. I'd bring it to reading group and would consider citing it once the pressure-dependence question is answered.","headline":"A careful sensitivity study whose main practical claim—drop Bruch's membrane—is plausible but only tested at one pressure condition, so the abstract overstates the evidence.","tokens_in":20623,"tokens_out":2453,"would_cite":false,"duration_ms":27085,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Finite element modeling of two healthy eyes shows scleral stiffness dominates optic nerve head mechanics and Bruch's membrane can be omitted.","keywords":["optic nerve head biomechanics","finite element modeling","subject-specific eye model","lamina cribrosa strain","intraocular pressure","intracranial pressure","sclera stiffness","Bruch's membrane"],"falsifier":"Measure lamina cribrosa strain quantitatively in human eyes in vivo or ex vivo (for example, with OCT-based displacement tracking or marker-based loading) and compare the magnitude and the inferior/temporal pattern with the model's predictions; if the measured strains differ substantially from the predicted values, the tissue-necessity ranking would need revision. Alternatively, repeat the sensitivity and removal tests using subject-specific measured scleral stiffness and measured Bruch's membrane thickness; if sclera stiffness no longer dominates or removing Bruch's membrane changes lamina cribrosa strain well beyond the reported 1.2 percent, the central claim would fail.","tokens_in":19627,"feed_emoji":"👁️","tokens_out":9357,"duration_ms":84545,"temperature":0.7,"pith_summary":"This paper asks which tissue regions must be present in a subject-specific computer model of the optic nerve head to faithfully predict how the lamina cribrosa, retina, and optic nerve deform under pressure. Using two healthy eyes imaged in vivo and finite element analysis with literature-based properties for tissues that are not visible on imaging, it finds that the stiffness of the sclera has the largest influence on the strain response of all three regions, while Bruch's membrane has a negligible effect. The authors conclude that accurate optic nerve head mechanics can be obtained by including only the tissue regions identified as necessary, and they show that the resulting reduced model produces the behavior expected from earlier studies: lamina cribrosa strain increases with both intraocular and intracranial pressure and is highest in the inferior and temporal sectors. The practical consequence would be simpler and more consistent finite element models for studying pressure-related optic nerve head loading in glaucoma.","feed_headline":"Sclera stiffness rules optic nerve head strain","feed_subtitle":"A two-eye finite element study finds Bruch's membrane barely matters, so models can drop it without losing accuracy.","key_machinery":"The central object is a three-dimensional, subject-specific finite element model of the optic nerve head reconstructed from six radial swept-source OCT slices, embedded in a spherical eye globe sized from each subject's spherical equivalent. Tissues are modeled as homogeneous, isotropic, nearly incompressible materials, with the sclera, lamina cribrosa, pia mater, dura mater, and annular ring treated as Neo-Hookean and the remaining tissues as linear elastic. Two probes carry the argument: a local sensitivity analysis that raises and lowers each tissue's Young's modulus by 25 percent, and binary tests that remove or replace the lowest-impact tissues one at a time. Outcomes are judged by average and peak principal tensile and compressive strains in the four subregions of the lamina cribrosa, in the retina, and in the optic nerve, with peak strains defined as 95th and 5th percentiles to avoid mesh artifacts.","core_discovery":"The central claim is that a finite element model of the optic nerve head built from in vivo imaging of two healthy subjects, with each major ocular tissue assigned a distinct material property, can determine which tissue regions are mechanically necessary. In these models, varying the sclera's Young's modulus by 25 percent produced the largest change in the principal strains of the lamina cribrosa, retina, and optic nerve; removing Bruch's membrane changed lamina cribrosa strains by only about 1.2 percent on average. The paper therefore asserts that accurate ONH mechanical representation can be obtained by including only the tissues identified as necessary, with Bruch's membrane excluded. It supports this by showing that the reduced model reproduces literature-consistent trends: strain in the lamina cribrosa rises when intraocular pressure or intracranial pressure increases, and the highest strains appear in the inferior and temporal subregions.","pith_inferences":["Because the paper attributes Bruch's membrane's negligible effect to its thinness and distance from the lamina cribrosa, a parametric sweep of its thickness and stiffness would map the range where that conclusion holds and where it would break.","The two-subject demonstration suggests population stability is untested; repeating the removal and sensitivity tests across eyes with varied axial length, age, and lamina cribrosa thickness would show whether the tissue-necessity ranking is general.","If scleral stiffness truly dominates the lamina cribrosa response, then non-invasive estimates of peripapillary scleral stiffness could become a clinically useful input for patient-specific glaucoma risk models, a step the paper does not take.","The pattern that thin stiff membranes matter less than their geometry would suggest that other models of the eye can safely use literature geometry for such tissues while concentrating measurement effort on the sclera."],"forward_implications":["Optic nerve head models can omit Bruch's membrane without materially changing predicted lamina cribrosa, retina, or optic nerve strains, simplifying segmentation while preserving accuracy.","Scleral stiffness is the input that matters most: errors in the scleral modulus will propagate into the largest errors in predicted lamina cribrosa strain, so subject-specific characterization of the sclera should be prioritized.","For the pia mater, border tissue, and annular ring, geometric presence matters more than precise stiffness values, because their removal changes strains substantially even though their modulus variations do not.","The reduced model reproduces the reported increase of lamina cribrosa strain with elevated intraocular and intracranial pressure and the inferior/temporal localization of peak strain, supporting its use in pressure-response studies.","A consistent tissue-inclusion standard of this kind could reduce the disagreements among eye models that currently differ in which tissue regions they represent."],"supporting_citations":[{"why":"Supplies the baseline Neo-Hookean ground-matrix stiffness values for the sclera and lamina cribrosa, the tissues whose property changes most affect the results.","marker":"(Wang et al., 2016)"},{"why":"Supplies pia mater and dura mater geometry, retina and optic nerve elastic moduli, and the 5th/95th percentile convention used for reporting peak strains.","marker":"(Feola et al., 2016a)"},{"why":"Supplies the dimensions and material parameters for Bruch's membrane, border tissue, annular ring, and choroid, including the shell geometry used for Bruch's membrane.","marker":"(Wang et al., 2018)"},{"why":"Supplies the elastic modulus assigned to the Bruch's membrane and border tissue.","marker":"(Jin et al., 2018)"},{"why":"Supplies the Neo-Hookean properties of pia mater and dura mater and supports the isotropy and compressibility assumptions for the models.","marker":"(Chuangsuwanich et al., 2020)"},{"why":"Provides the prior finite element comparison showing pia mater inclusion changes lamina cribrosa strains and frames the sclera as a protective shield, against which the sensitivity findings are checked.","marker":"(Sigal et al., 2004)"},{"why":"Experimental ex vivo comparison showing increased lamina cribrosa tensile and compressive strains with elevated cerebrospinal fluid pressure, used to validate the pressure-response trend.","marker":"(Feola et al., 2017)"},{"why":"Supplies the regional validation observation of highest lamina cribrosa strains in the inferior and temporal quadrants under elevated pressure.","marker":"(D. E. Midgett et al., 2017)"},{"why":"In vivo OCT study of healthy eyes showing higher lamina cribrosa strains in the inferotemporal region after intraocular pressure elevation, used as a validation comparison.","marker":"(Beotra et al., 2018)"},{"why":"Documents that the posterior lamina cribrosa boundary is frequently not visible in OCT, which is why the model relies on literature-based dimensions for that boundary.","marker":"(Girard et al., 2015)"}],"fun_headline_variants":["Sclera stiffness drives optic nerve head strain","Bruch's membrane? Drop it from eye models","Optic nerve head strain: sclera matters, Bruch's doesn't","Minimal eye model: sclera only, no Bruch's membrane","Sclera stiffness rules, Bruch's membrane droppable"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The ranking of which tissues are necessary assumes that the literature-based material properties and estimated dimensions of the non-visible tissues (posterior lamina cribrosa boundary, pia mater, dura mater, Bruch's membrane, border tissue, annular ring) are representative for the two subjects' eyes; if those inputs are not representative, the sensitivity and removal rankings, and the recommendation to drop Bruch's membrane, could change.","fun_headline_variants_meta":{"raw":{"variants":["Sclera stiffness drives optic nerve head strain","Bruch's membrane? Drop it from eye models","Optic nerve head strain: sclera matters, Bruch's doesn't","Minimal eye model: sclera only, no Bruch's membrane","Sclera stiffness rules, Bruch's membrane droppable"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000618,"raw_usage":{"total_tokens":2841,"prompt_tokens":890,"completion_tokens":1951,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":506,"completion_tokens_details":{"reasoning_tokens":1862}},"tokens_in":506,"tokens_out":1951,"duration_ms":15053,"temperature":1.0,"reasoning_tokens":1862,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:54:33.098316+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure lamina cribrosa strain quantitatively in human eyes in vivo or ex vivo (for example, with OCT-based displacement tracking or marker-based loading) and compare the magnitude and the inferior/temporal pattern with the model's predictions; if the measured strains differ substantially from the predicted values, the tissue-necessity ranking would need revision. Alternatively, repeat the sensitivity and removal tests using subject-specific measured scleral stiffness and measured Bruch's membrane thickness; if sclera stiffness no longer dominates or removing Bruch's membrane changes lamina cribrosa strain well beyond the reported 1.2 percent, the central claim would fail.","supporting_citations":[],"review_version":1}