{"id":"d2ffc9fa-5a1d-4827-b983-37e533fb78ac","arxiv_id":"1908.01473","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A paraxial x-ray beam's intensity evolution through a thin sample is described by adding a diffusion term to the transport-of-intensity equation, with a Kramers-Moyal generalization for anisotropic scattering.","lead":"This paper derives a Fokker-Planck equation that simultaneously models phase contrast and small-angle scattering in paraxial x-ray imaging. It gives a unified framework that extends the transport-of-intensity equation to include diffusive dark-field signals.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (9) places D inside the Laplacian, but the paper's own §2.2/§3 derivations put D outside; for varying D the two disagree and Eq. (9) predicts spurious dark-field contrast.","rationale":"The reader accepted with high confidence, identifying the narrow-fan second-order truncation as the weak assumption. I agree that is a scope limitation, but the stress-test pass found a more specific internal issue: even granting the second-order truncation, the operator ordering in Eq. (9) is inconsistent with the derivations in the same paper. The first-principles route (Eqs. 35, 38) and the Kramers–Moyal route (Eqs. 44, 50) both place the diffusion coefficient outside the second derivative; Eq. (9) places it inside. These differ whenever D varies transversally, producing terms the microscopic model does not contain. Because Eq. (9) is the equation named in the abstract and quoted as the central claim, this is load-bearing for the paper's headline statement. The paper's later use of Eq. (13) and the D-commuting assumption shows a fix is readily available, so the concern warrants a conditional acceptance with an amendment or qualification rather than rejection. No numerical or experimental validation is present, but the derivations otherwise appear internally consistent and the issue is local to the operator form.","tokens_in":23867,"tokens_out":15271,"duration_ms":173432,"concrete_test":"Perform a 1D SAXS-only numerical experiment: take I_s(x)=1, F=0.2 constant, D(x)=1+0.3 cos(2πx/p) with p much larger than the pixel size, Δ small but fixed, and a normalized symmetric narrow Gaussian kernel K(x,x') with width sqrt(D(x')Δ). Compare the propagated intensity from (a) the exact convolution Eq. (42); (b) the finite-difference Fokker–Planck form Eq. (13), F D Δ ∂²I; and (c) the Eq. (9)/Eq. (12) form, F Δ ∂²(D I). If (c) deviates from (a) by roughly Δ F D''(x) while (b) matches (a) to the Taylor-truncation error, the D-inside form is not the correct first-principles limit and Eq. (9) needs qualification.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The most load-bearing concern is an operator-ordering inconsistency internal to the central claim. In the first-principles derivation, the SAXS contribution is B≈F[1+L² d²/dx²]I_s (Eqs. (35) and (38)); the general Kramers–Moyal form, Eq. (50), has second-order terms F∑D^{(2)}_{m,2-m}∂²I with the diffusion tensor outside the derivatives. Reducing Eq. (50) to an isotropic fan gives F D ∇²⊥I, which is exactly the term used in Eq. (13), not the F∇²⊥[D I] appearing in Eq. (9). For spatially varying D the two differ by 2F∇D·∇I + F I ∇²D. These extra terms are not higher-order corrections in Δ; they survive at fixed small Δ and are absent from the microscopic kernel model. Concretely, if I_s(x)=1 and F is constant but D(x) varies, Eq. (9) predicts I(z=Δ)−I_s ≈ Δ F D''(x), whereas the normalized, first-moment-zero kernel of Eq. (42) leaves a uniform intensity unchanged and Eq. (13) gives zero. The paper only later restricts the simpler Eq. (13) to slowly varying D that commutes with the Laplacian, but Eq. (9) is stated unconditionally and is the equation highlighted in the abstract and in the reader's strongest claim. The claim should be amended to the D-outside form or explicitly qualified by ∇D≈0; otherwise forward/inverse uses with position-dependent SAXS fan width inherit a systematic bias.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes an x-ray Fokker–Planck equation for paraxial near-field imaging of thin samples, which augments the transport-of-intensity equation (TIE) with a diffusive term that models unresolved small-angle x-ray scattering (SAXS). Two derivations are presented: a phenomenological merging of the TIE and a diffusion equation in Sec. 2.1, and a first-principles derivation in Sec. 2.2 based on Fresnel propagation, ensemble averaging over a Gaussian stationary fast phase, and a second-order narrow-fan expansion. Section 3 generalises the result to a Kramers–Moyal equation with a hierarchy of SAXS-fan moments, and Section 4 outlines forward- and inverse-problem applications, including a two-distance phase-retrieval formula. The central claim is that Eq. (9) (and its one-dimensional form, Eq. (10)) simultaneously captures propagation-based phase contrast and dark-field blurring, with the Kramers–Moyal form, Eq. (50), handling anisotropic scattering fans.","tokens_in":24217,"tokens_out":4508,"duration_ms":48106,"significance":"If the proposed equations are valid, they provide a compact and practical formalism for modelling combined coherent and diffuse transport in x-ray imaging, with potential applications to phase retrieval, dark-field tomography, and imaging with visible light, electrons, and neutrons. The paper's strengths include a transparent microscopic derivation that gives independent physical meaning to the scattering fraction F, the phase depth σ_φf, the blur width L, and the diffusion coefficient D (Eqs. (28), (39), and (40)), and a Kramers–Moyal hierarchy that connects the Fokker–Planck equation to a rigorous moment expansion of the SAXS fan. The discussion of detector-resolution dependence of what is classified as SAXS is also valuable. However, the headline equation Eq. (9) has an operator-ordering inconsistency with the microscopic derivations that needs to be resolved before the paper can be accepted as is.","major_comments":[{"comment":"The diffusion term F(x,y)∇²⊥[D(x,y;Δ)I] places the diffusion coefficient inside the Laplacian, but the first-principles derivation in §2.2 (Eq. (38)) and the Kramers–Moyal expansion in §3 (Eq. (50)) both place D outside the derivatives, giving FD∇²⊥I (in the isotropic case). For spatially varying D the two forms differ by 2F∇D·∇I + FI∇²D. These are not higher-order corrections in Δ; they survive at fixed small Δ. Concretely, for I_s(x)=1 with constant F and varying D(x), Eq. (9) predicts I(z=Δ)−I_s ≈ ΔFD''(x), whereas the kernel model Eq. (42) leaves a uniform intensity unchanged and Eq. (13) gives zero. Eq. (9) is stated unconditionally and is highlighted in the abstract, so this is a load-bearing inconsistency that should be corrected by replacing ∇²⊥(DI) with D∇²⊥I (or with ∇⊥·(D∇⊥I)) whenever D varies, or by explicitly restricting Eq. (9) to ∇D≈0.","section":"§2.1.3, Eq. (9)"},{"comment":"The phenomenological diffusive current J(2)_⊥ = −∇⊥(DI) in Eq. (7) is nonstandard: combined with continuity it produces ∇²(DI), which is not the usual Fick-law form ∇·(D∇I). The uncritical use of this flux in the merging argument is the origin of the ordering discrepancy with the microscopic derivation. The authors should either justify why the current should contain I∇D or align the phenomenological derivation with the microscopic result by using the standard diffusive current −D∇I, which would yield F∇⊥·(D∇⊥I).","section":"§2.1.2, Eq. (7)"},{"comment":"The claim that the domain of applicability of the Fokker–Planck and Kramers–Moyal equations is 'at least as broad as that of the transport-of-intensity equation' is overstated. The diffusive term relies on the second-order narrow-fan truncation (Eqs. (34)–(35) and Eq. (44)), which requires each local SAXS fan to be narrow, smooth, and forward-peaked and the propagation distance Δ to be small enough for the Taylor expansion to be accurate. The TIE itself does not require these additional conditions. The abstract and conclusion should be qualified, for example by saying 'as broad as the TIE in the near-field limit of narrow SAXS fans.'","section":"Abstract"}],"minor_comments":[{"comment":"There is a typographical repetition: 'the the Fokker–Planck equation' should read 'the Fokker–Planck equation'.","section":"Introduction, p. 2"},{"comment":"The arrow notation 'weak SAXS−−−−−−→' is unusual and could be replaced with a clearer statement such as 'in the weak-SAXS limit σ²_φf ≪ 1, F(x) ≈ σ²_φf(x)'.","section":"Eq. (28)"},{"comment":"The notation kx=φ′s(x)−kx/Δ in the definition of d(x;x0) is confusing because kx appears on both sides with different meanings; using a separate symbol for the Fourier variable would improve readability.","section":"Eq. (33)"},{"comment":"The argument that Q=1 follows solely from energy conservation is terse; a sentence clarifying that the kernel integrates to unity (cf. Eq. (45)) would make the step easier to follow.","section":"§2.2, Eq. (34)"}],"recommendation":"major_revision","confidential_remarks":"The operator-ordering inconsistency in Eq. (9) is the main technical obstacle; it is local and fixable by aligning the phenomenological flux with the microscopic and Kramers–Moyal results. The authors should also ensure that the companion paper (ref. 76), which uses Eq. (10), is checked against the corrected form. The paper otherwise appears suitable for publication in the journal after the requested revisions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something real: it assembles known ingredients from Nesterets, Yashiro, Goodman, and the authors' own prior work into an explicit Fokker–Planck equation for near-field x-ray imaging, and it gives two derivations plus a Kramers–Moyal extension. The writing is clear, the physical picture is sensible, and the microphysical identifications — F = 1 - exp(-σ_φf²) and D = L²/Δ — turn the phenomenological coefficients into something concrete. The Kramers–Moyal hierarchy of SAXS-fan moments gives a useful formal umbrella. I don't hold the lack of numerical or experimental demonstration against it; this is a theory paper and the scope limits are stated honestly.\n\nThe soft spot is the one the stress-test note flags, and it is real. Equation (9), the equation highlighted in the abstract, puts D inside the Laplacian: F∇²[DI]. The first-principles derivation in §2.2 and the Kramers–Moyal equation in Eq. (50) both give the second-order diffusion term with D outside the derivatives — in the isotropic case, F D ∇²I. The paper acknowledges in §2.2 that the microscopic derivation only reduces to the D-inside form when D commutes with the Laplacian, but Eq. (9) is stated unconditionally and is the form quoted as the central claim. For position-dependent D, the two forms differ by 2F∇D·∇I + FI∇²D. That is not a negligible higher-order correction; it is a different model. If the intended model is the D-outside form, Eq. (9) should be amended or qualified. If the D-inside Fick form is intended, the derivation does not justify it. This is fixable and the correct D-outside form is present in the paper, but it needs to be sorted out before the equation gets used in forward or inverse problems.\n\nCitations: nothing offensive. The paper leans on the authors' prior work, but the cited results are standard in the field and the new synthesis is clearly distinguished. This is a paper for the x-ray phase-contrast and dark-field imaging community, and for anyone doing tensor tomography or speckle tracking. It deserves a serious referee and probably a major revision, not a desk rejection. I would bring it to a reading group because the discussion alone is worth the time.","headline":"A genuinely useful unification of phase contrast and SAXS diffusion, but the headline equation has an operator-ordering inconsistency that needs fixing before forward models should use it.","tokens_in":24780,"tokens_out":2855,"would_cite":true,"duration_ms":31919,"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":"A single Fokker–Planck equation merges x-ray phase contrast with dark-field blur.","keywords":["x-ray imaging","Fokker-Planck equation","Kramers-Moyal equation","transport-of-intensity equation","small-angle x-ray scattering","phase contrast imaging","dark-field imaging","paraxial optics"],"falsifier":"Take a sample whose scattering fan is known to be broad or structured and image it at three or more propagation distances; if the projected thickness recovered from each pair of distances via Eq. (61) changes with the chosen pair, or if the inferred $D_{\\rm eff}/\\Delta$ is not constant, the second-order diffusion truncation behind the Fokker–Planck equation has broken down.","tokens_in":23642,"feed_emoji":"🔬","tokens_out":6639,"duration_ms":61036,"temperature":0.7,"pith_summary":"Near-field x-ray images of thin samples are usually modelled either as coherent propagation-based phase contrast or as diffusive dark-field scattering, but real samples show both at once. The paper establishes a single equation, the x-ray Fokker–Planck equation, that describes the downstream intensity when coherent refraction and unresolved-sample small-angle scattering act together. It derives the equation twice: by fusing the transport-of-intensity equation with a diffusion equation through an energy-conserving current, and from an ensemble of unresolved random phases at the sample exit surface. The same machinery yields a Kramers–Moyal generalisation for anisotropic or structured scattering fans. If the claims hold, forward modelling becomes a single finite-difference update, and inverse problems can recover phase, thickness, and dark-field diffusion from one or two near-field images.","feed_headline":"One equation merges phase contrast and x-ray dark-field blur","feed_subtitle":"A diffusion-style update models coherent refraction and small-angle scattering together in near-field x-ray images.","key_machinery":"The engine of the paper is a continuity equation for transverse energy flow in which the current has two parts: a coherent Poynting-vector current $I\\nabla_\\perp\\phi/k$ and a diffusive Fick-type current $-\\nabla_\\perp[DI]$ entering with a weight $F$. Combining these gives the finite-difference Fokker–Planck update of Eq. (11). The diffusion coefficient is $\\Delta$-dependent, $D=L^2/\\Delta$, so that the small-angle-scattering blur expands as the propagation distance itself rather than as its square root. For the extended form, the Kramers–Moyal equation replaces the scalar $D$ with a hierarchy of diffusion tensors defined as moments of the position-dependent scattering kernel, allowing elliptical and structured scattering fans to be tracked. The hierarchy also supplies a principled truncation point, since Pawula's theorem forbids stopping at any order above second short of keeping everything.","core_discovery":"On the paper's own terms, the central claim is that the intensity of a paraxial quasi-monochromatic beam evolves as $$\n\\frac{\\partial I}{\\partial z}=-\\frac{1}{k}\\nabla_\\perp\\cdot[I\\nabla_\\perp\\phi]+F\\nabla_\\$perp^{2}$[DI].\n$$ The first term is the transport-of-intensity contribution from coherent refraction; the second adds a diffusive blur whose strength is the local fraction of the beam converted to small-angle scattering times a diffusion coefficient. The paper derives this equation twice: first by adding the coherent and diffusive currents in an energy-conserving way, then from first principles starting from a decomposition of the exit-surface phase into a slowly varying part and an unresolved random part. The first-principles route makes the physics explicit: the scattering fraction is $F=1-\\exp(-\\sigma_{\\varphi_f}^2)$, the diffusion coefficient is $D=L^2/\\Delta$, and the blur width $L$ grows linearly with propagation distance. When the scattering fan is anisotropic or highly structured, the same construction produces the Kramers–Moyal equation, whose successive terms are moments of the local scattering kernel.","pith_inferences":["If the two-distance inversion is stable, it suggests a practical protocol: collect two defocus distances and solve for thickness and dark-field simultaneously, rather than acquiring a full scattered-intensity map.","The diffusion-tensor moments in the Kramers–Moyal equation could be inverted directly from ordinary near-field images, connecting phase retrieval with tensor tomography without dedicated grating or raster-scan setups.","Because the Fokker–Planck update is linear in propagation distance for fixed $D$, time-lapse imaging of a slowly evolving sample could be modelled by the same equation with time as the evolution parameter, an extension the paper does not develop.","The same second-order expansion of the scattering kernel suggests that resolution-dependent classification of speckle as coherent versus diffuse could be used to tune detectors or binning to suppress dark-field blur in phase-contrast imaging."],"forward_implications":["A near-field image of a thin sample with unresolved microstructure can be computed by one finite-difference Fokker–Planck update from the exit-surface intensity, phase, and diffusion coefficient, without resolving the sample's internal speckle.","Two near-field images at different propagation distances can be combined, through Eqs. (60)–(61), to recover the projected thickness of a single-material sample independently of its dark-field scattering; the dark-field coefficient then follows algebraically.","The same equations apply to paraxial imaging with visible light, electrons, and neutrons, because they depend only on the paraxial scalar wave equation and energy conservation.","The geometric-flow speckle-tracking equation gains a diffusive term, so a single augmented equation can in principle reconstruct both phase and effective diffusion coefficient from speckle images.","Anisotropic or structured scattering fans can be represented by the Kramers–Moyal tensors, which can be measured by raster-scanning a focused beam over the sample."],"supporting_citations":[{"why":"Supplies the transport-of-intensity equation that the Fokker–Planck equation generalises, fixing the coherent current term.","marker":"[5]"},{"why":"Provides the Fokker–Planck and Kramers–Moyal formalism used throughout, including the forward-Kolmogorov form.","marker":"[13]"},{"why":"Gives the decomposition of the propagated intensity into a damped specular term and a diffuse small-angle-scattering term that anchors the first-principles derivation.","marker":"[44]"},{"why":"Introduces the split of exit-surface phase into slowly varying and unresolved fast-varying components used in the derivation.","marker":"[47]"},{"why":"Extends the phase-decomposition approach to anisotropic unresolved microstructure, supporting the Kramers–Moyal generalisation.","marker":"[48]"},{"why":"Supplies the speckle correlation-function split and renormalisation used to derive the decoherence factor and scattering kernel.","marker":"[50]"},{"why":"Provides the operator approximation of blur as a second-order derivative, which is the step that turns the scattering convolution into the diffusion term.","marker":"[66]"},{"why":"Gives the single-material homogeneous-object phase-retrieval formula whose algebraic form the two-distance dark-field-free thickness recovery inherits.","marker":"[7]"},{"why":"Justifies truncating the Kramers–Moyal expansion at second order or not at all, defining the hierarchy of valid models.","marker":"[70]"}],"fun_headline_variants":["X-ray Fokker-Planck unifies phase and dark-field imaging","One equation for coherent and diffusive x-ray transport","Fokker-Planck bridges phase contrast and scattering blur","New equation merges refraction and dark-field in x-rays","Paraxial imaging unified: phase plus small-angle scattering"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that each local small-angle scattering fan is narrow, smooth, and almost forward-peaked, so that the blur it produces can be captured by a second-order Taylor expansion in spatial derivatives over the propagation distances used.","fun_headline_variants_meta":{"raw":{"variants":["X-ray Fokker-Planck unifies phase and dark-field imaging","One equation for coherent and diffusive x-ray transport","Fokker-Planck bridges phase contrast and scattering blur","New equation merges refraction and dark-field in x-rays","Paraxial imaging unified: phase plus small-angle scattering"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000917,"raw_usage":{"total_tokens":3972,"prompt_tokens":1017,"completion_tokens":2955,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":633,"completion_tokens_details":{"reasoning_tokens":2871}},"tokens_in":633,"tokens_out":2955,"duration_ms":21928,"temperature":1.0,"reasoning_tokens":2871,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:12:06.586030+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a sample whose scattering fan is known to be broad or structured and image it at three or more propagation distances; if the projected thickness recovered from each pair of distances via Eq. (61) changes with the chosen pair, or if the inferred $D_{\\rm eff}/\\Delta$ is not constant, the second-order diffusion truncation behind the Fokker–Planck equation has broken down.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the decomposition of the propagated intensity into a damped specular term and a diffuse small-angle-scattering term that anchors the first-principles derivation."},{"cited_title":"& Momose, A","cited_arxiv_id":null,"evidence_quote":"Introduces the split of exit-surface phase into slowly varying and unresolved fast-varying components used in the derivation."},{"cited_title":"Distribution of unresolvable anisotropic microstructures revealed in visibility-contrast images using x-ray Talbot interferometry","cited_arxiv_id":null,"evidence_quote":"Extends the phase-decomposition approach to anisotropic unresolved microstructure, supporting the Kramers–Moyal generalisation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the speckle correlation-function split and renormalisation used to derive the decoherence factor and scattering kernel."},{"cited_title":"E., Stevenson, A","cited_arxiv_id":null,"evidence_quote":"Provides the operator approximation of blur as a second-order derivative, which is the step that turns the scattering convolution into the diffusion term."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Justifies truncating the Kramers–Moyal expansion at second order or not at all, defining the hierarchy of valid models."}],"review_version":1}