{"id":"d0335c30-0020-4dc8-803e-d58e8935aebd","arxiv_id":"2510.02847","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A ratio of dual- to single-frequency spin-lock relaxation rates, RATIO_dosl, maps T1D and MPF from three images.","lead":"A three-image spin-lock MRI method maps the dipolar relaxation time T1D, a myelin-sensitive tissue property, in white matter. The approach is fast and also yields the macromolecular pool fraction, potentially making T1D mapping practical for clinical use.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fixed-T2b conversion is load-bearing: Fig. 4(b) shows RATIO_dosl is highly sensitive to T2b, so reported T1D values may be biased unless T2b is known or its effect is bounded.","rationale":"The central claim requires that the measured ratio uniquely and accurately encodes T1D. The paper supplies strong support for the other pieces: code is available; Fig. 3 shows the analytical approximation matches exact Bloch-McConnell-Provotorov solutions; Fig. 6(a) quantifies B1/B0 behavior; phantom data confirm ihMT contrast. The weakest point is the fixed-parameter inversion, specifically T2b, because the ratio is demonstrably sensitive to T2b and the reported in vivo numbers were not validated against an independent T1D method in the same subjects. The reader's verdict 'conditional' is exactly right: the method is promising and reproducible, but the quantitative T1D values are not yet robust to a plausible biological variation in T2b. A targeted simulation can settle whether the concern actually changes the numbers. If it does, the abstract's 'quantitative' claim should be softened; if it does not, the conditional can be upgraded.","tokens_in":18942,"tokens_out":4789,"duration_ms":61425,"concrete_test":"Run the provided Bloch-McConnell-Provotorov simulations with the exact in vivo protocol parameters (Sec. 3.4.2: Δωd1/2π = 5 kHz, ω1d1/2π = 500 Hz, N = 5, TSL = 80 ms, switch times 0.5/40 ms) and generate RATIO_dosl for T2b ∈ {8, 9, 9.7, 11, 12} μs and T1D ∈ {3, 4, 5, 6, 7, 8} ms. Invert each value using the fixed-T2b = 9.7 μs dictionary. If the resulting T1D bias exceeds ~0.5 ms or the inter-bundle SD in Table 2, the constant-T2b assumption is load-bearing and the quantitative claim needs qualification.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central quantitative claim is that RATIO_dosl from three spin-lock images yields T1D values of ~3.70–4.80 ms in white matter (Abstract; Sec. 3.2; Table 2). This conversion uses a dictionary built with fixed MT parameters. In Sec. 3.4.3, T2b is fixed at 9.7 μs for all ten volunteers, and in Sec. 2 the dipolar-field parameter is D ≈ 1/(T2b√15), so the MT-pool lineshape term R_rf^b entering Eq. (8) is a direct function of T2b. Fig. 4(b) shows RATIO_dosl has pronounced sensitivity to T2b, while it is flat in MPF, R1b, and R. The robustness study in Sec. 3.3.2.2 perturbs R1b, R, and MPF but explicitly excludes T2b; the Discussion acknowledges this as an open issue. Since T2b is not a universal constant across healthy and pathological tissue, a modest T2b deviation can shift the RATIO-to-T1D mapping curve. Without a quantitative bound on that shift, the reported T1D values are conditional on an assumed lineshape, and the method's status as a quantitative T1D biomarker is not fully established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a spin-lock MRI method for quantitative mapping of the dipolar relaxation time T1D, using a ratio RATIO_dosl built from differences between dual-frequency and single-frequency spin-lock relaxation rates. The authors derive an approximate analytical expression (Eq. 10), validate it against numerical Bloch–McConnell–Provotorov simulations (Fig. 3), test the method in agar and PL161 phantoms, and demonstrate in vivo joint T1D and MPF mapping in ten healthy volunteers using only three spin-lock prepared images. Reported mean white-matter T1D values are approximately 3.70–4.80 ms. The paper includes B1 correction, two T1D estimation routes (analytical and dictionary matching), and test–retest acquisitions. The central claim is that this three-image protocol provides clinically feasible simultaneous T1D and MPF quantification.","tokens_in":19309,"tokens_out":8502,"duration_ms":72211,"significance":"If the claim holds, the method is a meaningful step toward fast quantitative ihMT imaging: it replaces multi-image ihMT and MRF T1D acquisitions with a short spin-lock protocol, and it simultaneously gives MPF. The paper has tangible strengths: the analytical approximation is compared with the full numerical solution, the code is publicly available, B1 correction is incorporated, and phantom and in vivo demonstrations are included. However, the quantitative T1D values rest on the assumption that T2b and R are fixed across tissue and subjects; this is explicitly acknowledged in the paper but not bounded. The three-image ratio also requires a derivation that is not fully clear as printed. These issues affect the quantitative biomarker claim, so the contribution is promising but not yet fully established.","major_comments":[{"comment":"The conversion from RATIO_dosl to T1D fixes T2b=9.7 μs, obtained from a single volunteer's Z-spectrum fit, and uses this value for all subjects. Fig. 4(b) shows pronounced sensitivity of RATIO_dosl to T2b, and the robustness study in Sec. 3.3.2.2 deliberately varies R1b, R, and MPF but excludes T2b. The Discussion acknowledges this. This is a load-bearing limitation: if T2b varies modestly across subjects or tissue, the reported 3.70–4.80 ms range could be biased without any stated uncertainty. Please supply a quantitative bias analysis over a plausible T2b range, or measure/correct T2b on a per-subject basis, before the absolute T1D values are presented as tissue biomarkers.","section":"Sec. 3.4.3 / Fig. 4(b)"},{"comment":"R_dosl is defined as a difference between dual-frequency and single-frequency spin-lock relaxation rates. However, the printed Eq. (10) and Eq. (11) use the same single-frequency image, R1ρ^single(1), in both the numerator and the denominator: RATIO_dosl = (R1ρ^dual(1)-R1ρ^single(1))/(R1ρ^dual(2)-R1ρ^single(1)). Since the single-frequency term in Eq. (9) depends on Δω_s and ω1_s, and condition 2 uses different offsets/powers, it is not obvious that this three-image form equals the intended ratio R_dosl,1/R_dosl,2. The step from Eq. (9) to Eq. (10) should be shown in full or the cancellation must be justified explicitly. If the cancellation does not hold, the method requires a fourth image and the central 'three images' claim needs revision.","section":"Eq. (10)-(11), Sec. 3.1"},{"comment":"The dictionary parameters (MPF=13.6%, R=20 s−1, T2b=9.7 μs) are determined from one volunteer and then applied to all ten volunteers. No in vivo comparison with an independent T1D quantification method (e.g., the multi-ihMTR approach of Varma et al. or the MRF approach of West et al.) is made on the same subjects; the Discussion states that precise validation of true T1D remains challenging. Since the quantitative claim is a specific T1D range, a direct cross-method comparison or a histology-based calibration would materially support the result. At minimum, the sensitivity of the final T1D estimates to the dictionary parameters should be quantified beyond the existing R1b/R/MPF robustness study.","section":"Sec. 3.4.3 / Table 2 / Discussion"}],"minor_comments":[{"comment":"Typo: 'approcah' should be 'approach'.","section":"Abstract"},{"comment":"The sentence 'an independe nce R1 maps acquisition' contains a spacing typo and the R1 mapping method is not described; please specify the sequence used.","section":"Sec. 3.4.2"},{"comment":"The notation M_a^1, M_a^2, M_s^1, and M_Tog is used without a clear definition of which images correspond to which symbols. Please spell this out explicitly, especially the role of the magnetization-reset image.","section":"Eq. (11)"},{"comment":"In the provided manuscript text, Table 2 appears as a caption without the actual numeric entries. The table must be populated so values can be inspected.","section":"Table 2"}],"recommendation":"major_revision","confidential_remarks":"The T2b sensitivity and the derivation of the three-image ratio are the two points that determine whether this paper is publishable as a quantitative method. The authors are transparent about the first, but transparency is not a substitute for a quantitative bound or per-subject correction. The second could be a typographical artifact, but as printed the algebra is incomplete and should be verified. If both points are resolved satisfactorily, the paper would be a useful contribution to the ihMT/spin-lock literature."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this paper does something real. The new idea is RATIO_dosl, a T1D-specific ratio built from three off-resonance spin-lock images, with a rotary-echo sequence that alternates positive and negative offsets to produce dual- vs single-frequency spin-lock. If the model holds, that's a substantial acquisition-speed improvement over Varma's multi-ihMTR and West's MRF, and they also get MPF from the same three images at no extra cost.\n\nWhat's done well: the analytical approximation is checked against full Bloch-McConnell-Provotorov simulation with <1% relative error over the tested range (Fig. 3). Phantom behavior matches ihMT expectations—PL161 shows T1D contrast, agar does not. The in vivo white-matter values (3.70–4.80 ms) sit in the range of prior reports, and they ran test-retest with a 7–10 day interval. Code is on GitHub. For a methods paper, that is a solid evidence base.\n\nThe main soft spot is the one the authors themselves flag. Conversion of RATIO_dosl to T1D assumes the MT parameters, especially T2b, remain constant (Sec. 3.4.3). Fig. 4(b) shows the ratio is essentially flat against MPF, R1b, and R, but has pronounced sensitivity to T2b. The robustness study perturbs R1b, R, and MPF but deliberately excludes T2b, so the reported T1D values are conditional on a fixed lineshape. If T2b varies across subjects or pathology, the estimates shift. That makes the method's status as an absolute quantitative biomarker provisional, not invalid—and it is honestly acknowledged in the Discussion, so it is a limitation rather than an oversight. Less central but still real: Eq. 11 has a garbled line with mismatched subscripts, which will cost anyone reimplementing it time; and there is no same-subject comparison against an established T1D method, though that is a reasonable validation gap for a first description.\n\nBottom line: the central claim—three spin-lock images give simultaneous T1D and MPF—is supported by the simulations and the phantom/in vivo data shown. I would want a T2b sensitivity bound or a multi-T2b dictionary before trusting absolute values across pathologies, but this is a clear step forward, not a desk reject. I'd send it to peer review and would cite it if I work in MT imaging. Recommendation: engage with it.","headline":"T1D mapping via a three-image spin-lock ratio is a genuine speedup over existing ihMT methods, but the fixed-T2b conversion is a real, honestly-flagged caveat that needs a bound before the absolute T1D values carry weight.","tokens_in":19853,"tokens_out":2346,"would_cite":true,"duration_ms":57731,"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":"From just three spin-lock images, a single ratio yields quantitative maps of the dipolar relaxation time T1D and the macromolecular proton fraction.","keywords":["dipolar order","inhomogeneous magnetization transfer","spin-lock MRI","T1D mapping","macromolecular proton fraction","myelin microstructure","quantitative magnetization transfer","rotary echo"],"falsifier":"Acquire RATIO_dosl and an independent T1D reference in the same white matter while also estimating T2b from quantitative MT fitting; if the apparent T1D correlates with T2b rather than with the reference, the constant-T2b assumption is falsified. A simpler phantom check: fix a PL161 sample with constant T1D, change the T2b value used in the dictionary, and observe whether the reported T1D moves.","tokens_in":18812,"feed_emoji":"🧠","tokens_out":5012,"duration_ms":38511,"temperature":0.7,"pith_summary":"This paper claims that a dimensionless ratio, RATIO_dosl, computed from three off-resonance spin-lock prepared images, carries a measure of dipolar order that is largely independent of water relaxation and most magnetization-transfer parameters. If the claim holds, T1D maps—previously requiring long multi-image ihMT acquisitions or MR fingerprinting—can be obtained in about a minute per slice, and the same three images also yield the macromolecular proton fraction (MPF). The authors support the claim with numerical simulations, phantom experiments on a lipid emulsion with strong inhomogeneous magnetization transfer contrast, and test-retest in vivo scans of ten healthy volunteers, reporting white-matter T1D values around 3.70–4.80 ms. A sympathetic reader would take the central contribution to be the reduction of dipolar-order quantification to a single ratio accessible with a standard spin-lock sequence.","feed_headline":"Three spin-lock images map myelin's dipolar relaxation time","feed_subtitle":"A ratio of dual- vs single-frequency spin-lock rates yields T1D and MPF from one scan, with white-matter T1D around 4 ms.","key_machinery":"The load-bearing object is RATIO_dosl, the ratio of two R_dosl values, each being the difference between dual- and single-frequency spin-lock relaxation rates. Taking the ratio cancels the water-pool contribution and, with MPF, R1b, T2b, and R held constant, makes the measured quantity a function of T1D alone. A rotary-echo spin-lock pulse chain alternates frequency offset and phase: short switch time (0.5 ms) produces effective dual-frequency irradiation, long switch time (40 ms) produces single-frequency irradiation. T1D is recovered either by inverting the analytical expression or by matching against a simulated dictionary, with a B1 map correcting the actual RF amplitude.","core_discovery":"The paper's central claim is that the difference between dual-frequency and single-frequency spin-lock relaxation rates defines an ihMT-specific rate R_dosl, and the ratio of R_dosl at two constrained spin-lock settings is a one-to-one function of the dipolar relaxation time T1D under fixed MT parameters. This ratio, RATIO_dosl, is shown by simulation to track T1D across 1–10 ms with low relative error when offset, spin-lock amplitude, and duration are chosen in a favorable range (5 kHz, 500 Hz, 80 ms in vivo). A rotary-echo spin-lock pulse train with a short switch time acts as dual-frequency irradiation and a long switch time as single-frequency irradiation, so both rates are sampled in pr","pith_inferences":["Because RATIO_dosl is sensitive to T2b, and T2b is assumed constant, any disease process that changes the macromolecular lineshape would masquerade as a T1D change; testing this would require independent T2b measurement in the same tissue.","The method assumes a single T1D component and a dipolar-order fraction of unity; if multi-component dipolar reservoirs are present in myelin, the reported T1D is an effective average rather than a pool-specific value, and a two-component extension would clarify the interpretation.","The orientation dependence of the spin-lock ihMT signal is left open; a simple check would be to rotate a fixed anisotropic phantom or excised nerve relative to B0 and see whether estimated T1D shifts.","Since MPF and T1D show different contrast in the maps, the ratio may provide an independent axis for tissue classification; one could test empirically whether the joint (MPF, T1D) pair separates demyelinating lesions from normal white matter better than either alone."],"forward_implications":["T1D mapping can be added to a clinical protocol as three spin-lock prepared images, with about one minute acquisition per slice, instead of the eight or more ihMT-weighted images used by earlier quantification schemes.","The same acquisition yields MPF from the two dual-frequency images, so macromolecular content and dipolar-order microstructure can be compared voxel-by-voxel without extra scan time.","If translated to 3D FSE/TSE readouts, whole-brain T1D/MPF mapping should be feasible in roughly five minutes, making the measure practical for studies of myelination.","B1 inhomogeneity, which chiefly biases the line-shape amplitude, can be corrected retrospectively with a B1 map, and B0 offsets of +/-100 Hz have little effect.","Simulations indicate that an SNR of about 40 is sufficient to keep the T1D estimate within a few percent bias."],"fun_headline_variants":["Spin-lock MRI: three images map dipolar relaxation time","Ratio of spin-lock rates gives T1D from 3 images","Dipolar order mapping: spin-lock MRI does it in 3 scans","Simultaneous T1D and MPF from a single spin-lock ratio"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The conversion of RATIO_dosl to T1D assumes the MT model parameters MPF, R1b, T2b, and exchange rate R are constant; the paper's own simulations show RATIO_dosl is especially sensitive to T2b, so if T2b varies across subjects or tissue the estimated T1D is biased.","fun_headline_variants_meta":{"raw":{"variants":["Spin-lock MRI: three images map dipolar relaxation time","Ratio of spin-lock rates gives T1D from 3 images","Dipolar order mapping: spin-lock MRI does it in 3 scans","Simultaneous T1D and MPF from a single spin-lock ratio"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000273,"raw_usage":{"total_tokens":1557,"prompt_tokens":914,"completion_tokens":643,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":658,"completion_tokens_details":{"reasoning_tokens":572}},"tokens_in":658,"tokens_out":643,"duration_ms":8144,"temperature":1.0,"reasoning_tokens":572,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T12:38:14.092424+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Acquire RATIO_dosl and an independent T1D reference in the same white matter while also estimating T2b from quantitative MT fitting; if the apparent T1D correlates with T2b rather than with the reference, the constant-T2b assumption is falsified. A simpler phantom check: fix a PL161 sample with constant T1D, change the T2b value used in the dictionary, and observe whether the reported T1D moves.","supporting_citations":[],"review_version":1}