{"id":"8127455e-c3c4-4e76-afd1-15a721d5204c","arxiv_id":"2412.08499","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A modified Mueller polarimeter with a large LED panel performs both 3D-PLI and ComSLI on the same brain section, yielding a combined fiber direction map.","lead":"Researchers built a microscope that combines two light-based brain fiber mapping techniques in one device. It can image the same tissue section with both methods, producing a more complete picture of nerve fiber directions.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Comparability claim rests on unquantified visual agreement after manual alignment; no angular error statistics are reported, so the central claim is not yet established.","rationale":"The paper is a credible hardware proof-of-concept: it gives a detailed construction of the Scattering Polarimeter, honest reporting of artifacts and noise, and a substantive supplementary error analysis via Mueller calculus. The claim that would make the paper novel is that the device produces results comparable to dedicated 3D-PLI and ComSLI systems. That claim currently rests on side-by-side images and manually applied rotations/offsets, with no quantitative angular error metrics. The reader's weakest assumption correctly identifies this gap, and my independent reading confirms it: the 3D-PLI comparison section explicitly describes rotating/transposing reference data and adding a global offset 'if required,' yet reports no alignment parameters or agreement statistics; the ComSLI section similarly relies on visual inspection of direction and vector maps. The multimodal map is explicitly presented as a proof of concept and depends on sample-specific thresholds, so it is less central to the headline claim. A quantitative re-analysis of the deposited data would settle whether the comparability claim holds. Since the reader's conditional verdict already calls for such quantitative comparisons, my stress-test does not change the verdict.","tokens_in":27565,"tokens_out":2862,"duration_ms":31594,"concrete_test":"Re-analyze the deposited data (Jülich DATA, DOI 10.26165/JUELICH-DATA/CUXQYF): for each of the six measured regions, after applying the same orientation correction, compute the per-pixel circular angular difference between the Scattering Polarimeter and LMP3D 3D-PLI direction maps for pixels with retardation above a pre-registered threshold (e.g., |sin(delta)| > 0.07), and report the median and RMS difference plus the fraction of pixels within ±5 and ±10 degrees. If the RMS exceeds about 10 degrees or the ±5-degree fraction is below about 90%, the word 'comparable' in the abstract is not quantitatively supported. Repeat the same calculation for ComSLI dominant directions in white matter against the reference ComSLI setup.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract claims that the Scattering Polarimeter 'generates results comparable to state-of-the-art 3D-PLI and ComSLI setups.' In the 3D-PLI Results, the reference direction maps are rotated by flipping and/or transposing the data array, and 'if required, a global direction offset was added'; transmittance maps are contrast-matched by choosing visualization ranges. No pixel-wise angular difference statistics, alignment parameter values, or agreement thresholds are given. The ComSLI comparison is likewise qualitative: direction maps and vector maps are visually compared, with the vector kernels described only as 'approximately match.' Because the alignment procedure includes free parameters (orientation flips, global offset, contrast range), systematic deviations could be concealed or compensated by these manual choices. The load-bearing assumption is therefore that visual agreement after such manual alignment demonstrates equivalence, and this assumption is unverified. The multimodal fiber direction map also relies on sample-specific classification thresholds (e.g., peak prominence >=0.3 for human vs >=1.0 for vervet; average scattering >1200 vs >430 a.u.) without a robustness analysis, but the comparability claim is the more central issue.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces the Scattering Polarimeter, a microscope that integrates 3D-PLI and angular ComSLI into a single device using a Mueller polarimeter architecture with LCVRs and a large-area LED panel. The authors perform 3D-PLI and ComSLI measurements on two brain sections (human and vervet) with different optical properties, compare the resulting transmittance, retardation, fiber direction, and scattering maps against state-of-the-art reference setups, and construct a multimodal fiber direction map that combines 3D-PLI directions with ComSLI crossing information. The paper also provides a Mueller-matrix-based error analysis and discusses technical limitations and future hardware improvements.","tokens_in":27925,"tokens_out":2463,"duration_ms":28229,"significance":"If the comparability claim holds, this is the first single-device demonstration of correlative polarized-light and scattered-light fiber mapping with pixelwise alignment, which would facilitate faster multimodal measurements, cross-validation, and combined analysis of fiber orientation and crossings. The authors are transparent about artifacts, noise, longer exposure times, and the prototype nature of the device. The supplementary material includes fitted sinusoidal curves with residuals, R² values for representative pixels, and a detailed Mueller calculus treatment of systematic errors from LCVR retardance offsets. The data are publicly deposited, which strengthens reproducibility. The main uncertainty is the strength of the evidence for the central 'comparable results' claim, which currently rests on qualitative visual comparison.","major_comments":[{"comment":"The central claim that the Scattering Polarimeter 'generates results comparable to state-of-the-art 3D-PLI and ComSLI setups' is not supported by quantitative agreement metrics. In the 3D-PLI comparison, the reference direction maps are aligned by flipping and/or transposing the data array and adding a global direction offset (Section 'Three-Dimensional Polarized Light Imaging', with the quoted description of the rotation procedure), and transmittance maps are contrast-matched by choosing visualization ranges. The ComSLI comparison is similarly qualitative, relying on visual side-by-side inspection and vector maps with kernels chosen to 'approximately match.' No angular difference histograms, mean/median absolute angular errors, or pixel-wise error maps are reported. Because the alignment procedure includes free parameters (flips, transposition, global offset, contrast range), systematic deviations between the Scattering Polarimeter and the references could be concealed or compensated by these manual choices. The authors should provide quantitative angular error statistics over defined tissue regions (e.g., white matter, gray matter, crossing regions) for both brain samples, and report the alignment parameters actually used. Without such metrics, the comparability claim is overstated for a proof-of-concept study.","section":"Results, 3D-PLI and ComSLI"},{"comment":"The multimodal fiber direction map depends on sample-specific classification thresholds that are reported explicitly but not subjected to any robustness or sensitivity analysis. In the subsection 'Pixel classification (crossing fibers)', the thresholds for inclined parallel fibers (peak prominence ≥0.3 for human, ≥1.0 for vervet) and steep parallel fibers (average scattering >1200 a.u. for human, >430 a.u. for vervet) differ by factors of 3 or more between the two samples. The classification also uses a retardation threshold of |sin(δ)| < 0.07 for gray matter and a background threshold of I≤20 a.u. on the average scattering map. Since these thresholds directly determine which pixels are assigned to 3D-PLI versus ComSLI directions in the multimodal map, and since they are hand-set per sample, the generality of the multimodal construction is unclear. A sensitivity analysis (varying each threshold over a plausible range and reporting the resulting changes in the multimodal map) or a principled threshold-selection procedure would strengthen the claim that the multimodal map is a reliable combination rather than a demonstration tuned to these two samples.","section":"Results, Multimodal fiber direction map"}],"minor_comments":[{"comment":"There is a typo in the Introduction: 'does not rely on the briefringence' should read 'birefringence.'","section":"Introduction"},{"comment":"The vector maps are displayed with different kernel sizes for the Scattering Polarimeter (20×20 pixels) and the reference setup (10×10 pixels). The authors state that this 'was chosen to approximately match' the maps, but the size difference makes visual comparison of vector density and noise levels difficult. A figure showing both at the same kernel size, or a quantitative description of how the kernels were matched, would improve clarity.","section":"Results, ComSLI"},{"comment":"The Fourier fit quality is documented with R² values for only a few representative pixels (Supplementary Fig. 2). Since the comparability claim relies on the reliability of the 3D-PLI signal, it would be helpful to report the distribution of R² values over the imaged regions, or at least over a larger set of pixels in both white and gray matter, to demonstrate that the high fit quality is not limited to selected locations.","section":"Supplementary 4"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the scope of physics.med-ph and presents a genuinely useful proof-of-concept device. The main issue is that the abstract and results make a strong claim of 'comparable results' that is not yet quantitatively established. I believe this is fixable within the manuscript's scope by adding angular error statistics and reporting the alignment parameters; it does not require new experiments, since the raw data are already available. The multimodal classification thresholds also need a robustness discussion. I have no concerns about the citation pattern or novelty disclosure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a real hardware advance — a modified Mueller polarimeter that runs 3D-PLI and ComSLI on the same section without moving the sample, plus a first attempt at a multimodal direction map. The device is the contribution, and the engineering is described in enough detail to reproduce. They also ship data with a DOI, which is more than most papers in this space do.\n\nWhat I liked: the clever use of a large LED panel for both vertical and oblique illumination, and the PSG bypass for ComSLI is a neat trick. They tested on two samples with very different optical behavior (old transparent vervet vs. fresh scattering human) and they are upfront about artifacts, noise, long exposure times, and the LCVR ellipticity problem. The Fourier fit residuals in Supplementary 4 give an honest look at systematic errors. The multimodal map is more than a badge: they explain the rules for choosing between 3D-PLI and ComSLI directions pixel-by-pixel, and Figure 5 shows a real SNR improvement over single-mode ComSLI.\n\nThe soft spot is real and central. The claim that the Scattering Polarimeter generates results comparable to state-of-the-art setups is supported only by visual inspection after manual processing: reference maps are flipped or transposed, a global direction offset is added 'if required', and transmittance ranges are contrast-matched. No angular-difference histograms, no mean or median error, no agreement threshold. With that many free alignment parameters, systematic orientation biases could be concealed. I am not saying the results are wrong — the maps do look similar, and the R-squared values suggest clean signals — but the comparability claim is not yet demonstrated quantitatively. This is fixable: report the alignment parameters, compute per-pixel angular differences in white matter and crossing regions, and give error metrics against the reference setups.\n\nThe multimodal classification also uses sample-specific thresholds (peak prominence >=0.3 for human vs >=1.0 for vervet; scattering >1200 vs >430 a.u.) with no robustness analysis. That is a minor point relative to the missing quantitative comparison, but worth mentioning.\n\nWho this is for: anyone building correlative polarimetric/scattering instruments, and neuroscientists who want to know whether combined 3D-PLI/ComSLI is practical. This paper deserves a serious referee. I would send it out, but with a request for quantitative validation before acceptance. My own verdict is promising proof-of-concept, not yet a fully validated instrument.","headline":"A genuine first integration of 3D-PLI and ComSLI in one microscope, but the 'comparable results' claim currently rests on visual agreement after hand-tuned alignment rather than measured errors.","tokens_in":28334,"tokens_out":3087,"would_cite":true,"duration_ms":32892,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper introduces a single microscope that combines polarized-light and scattered-light fiber mapping and shows it matches separate state-of-the-art setups on real brain sections.","keywords":["scattering polarimetry","3D-PLI","ComSLI","nerve fiber mapping","multimodal imaging","Mueller polarimeter","brain sections","fiber crossings"],"falsifier":"On the same section and pixel grid, compute the fiber direction angle with the Scattering Polarimeter and with a reference 3D-PLI setup, align the data using the paper's own procedure, and measure the median absolute angular difference in the corpus callosum; if the median exceeds a few degrees or varies systematically with brain region, the comparability claim fails. A complementary check would count false-positive crossing detections in regions known to contain only parallel fibers, where the multimodal map should not introduce extra directions.","tokens_in":27358,"feed_emoji":"🧠","tokens_out":9411,"duration_ms":86986,"temperature":0.7,"pith_summary":"This paper introduces the Scattering Polarimeter, a microscope that performs Three-Dimensional Polarized Light Imaging (3D-PLI) and Computational Scattered Light Imaging (ComSLI) on the same brain section in immediate succession, without moving the sample. 3D-PLI gives stable in-plane fiber directions and retardance from birefringence, while ComSLI uses oblique scattered light to resolve multiple crossing fiber directions per pixel. On a 60 µm vervet monkey section and a 50 µm human section, the authors report that the new device produces transmittance, retardation, and fiber direction maps comparable to those from a state-of-the-art 3D-PLI microscope and a dedicated ComSLI setup. They then fuse both modalities into a multimodal fiber direction map that shows up to three directions per pixel, combining 3D-PLI's stability with ComSLI's crossing detection. The value of the claim is that correlative, pixel-aligned multimodal mapping of dense nerve fiber architecture would no longer require two separate instruments or image registration.","feed_headline":"One microscope maps brain fibers two complementary ways","feed_subtitle":"A new polarimeter unites polarized-light and scattered-light fiber mapping, with pixel-aligned directions and crossings.","key_machinery":"The load-bearing mechanism is the Scattering Polarimeter itself: a Mueller polarimeter whose polarization state generator and analyzer use four liquid-crystal variable retarders and two fixed linear polarizers, with a large-area LED panel that supplies both vertical illumination (for 3D-PLI and Mueller polarimetry) and oblique illumination segments (for ComSLI). Because both measurements share one camera and one sample position, the resulting parameter maps are inherently pixel-aligned. The multimodal fiber direction map then carries the argument's final step: a hand-tuned classification of each pixel by expected fiber architecture decides whether 3D-PLI or ComSLI directions are more trustworthy, and up to three directions are displayed per pixel.","core_discovery":"The central claim is that a modified Mueller polarimeter, built from two linear polarizers, four liquid-crystal variable retarders, and a large patternable LED panel, can act as both a 3D-PLI system and a ComSLI system, generating results equivalent to dedicated single-mode setups. Vertical illumination through the polarization state generator and analyzer yields the 3D-PLI sinusoidal signal, from which transmittance, retardation, and fiber direction are extracted; oblique illumination segments bypass the polarization generator so that the same camera records scattering profiles from which up to three fiber directions per pixel are derived. The authors show this on human and vervet monkey brain sections and go on to construct a multimodal fiber direction map: a pixel-classification routine identifies white matter, gray matter, and crossing regions, then selects for each pixel the most reliable direction from 3D-PLI or ComSLI, exhibiting an improved signal-to-noise ratio in gray matter and parallel-fiber regions while preserving crossing information. The authors present this as a proof of concept that scattering polarimetry can serve as a correlative, multimodal nerve fiber imaging method.","pith_inferences":["Beyond the paper: a quantitative validation study reporting pixel-wise angular error against the reference setups would let other labs reproduce the comparability claim; the current report compares visually after manual alignment.","Beyond the paper: if the multimodal direction map proves reliable, it could serve as a micrometer-resolution ground truth for validating diffusion MRI tractography in post-mortem brains, especially for crossing-fiber regions.","Beyond the paper: the reported transmittance asymmetry in the older, more transparent vervet sample could be corrected by calibrating the LCVR retardance-voltage curves near their quarter-wave operating points, a direct, testable hardware fix implied by the paper's Mueller calculus.","Beyond the paper: comparing 3D-PLI and ComSLI directions per pixel could be developed into a label-free assay of myelin integrity, since loss of myelin changes birefringence without changing scattering."],"forward_implications":["Combined 3D-PLI and ComSLI measurements on one section can be done with a single device, eliminating sample transfer and image registration steps.","The multimodal fiber direction map can display up to three fiber directions per pixel, retaining 3D-PLI's stability for parallel fibers and gray matter while adding ComSLI's crossing detection.","Regions where 3D-PLI and ComSLI directions disagree may flag changes in tissue optical properties, for instance myelin degeneration that alters birefringence but leaves scattering largely unchanged.","Because the hardware is a Mueller polarimeter, the same device can in principle also deliver full Mueller matrix parameters such as depolarization and diattenuation, adding tissue-classification contrasts without new hardware."],"supporting_citations":[{"why":"supplies the 3D-PLI sinusoidal signal model and the relation between retardation and fiber inclination that the new device's 3D-PLI analysis implements","marker":"12"},{"why":"describes the 3D-PLI measurement procedure and blockface-based registration that the reference 3D-PLI set-up and the reported comparisons build on","marker":"13"},{"why":"establishes scatterometry-based scattered light imaging that underlies ComSLI's full scattering patterns","marker":"14"},{"why":"introduces angular ComSLI and the peak-pair method for extracting in-plane fiber directions, the algorithm the new device uses for crossing detection","marker":"15"},{"why":"provides the software toolbox used to evaluate the ComSLI measurements and compute the fiber direction and vector maps","marker":"19"},{"why":"gives the Lu-Chipman decomposition used in the Mueller polarimetry analysis and in the error-propagation discussion","marker":"16"},{"why":"supplies the eigenvalue calibration method used to calibrate the measured Mueller matrices of the Scattering Polarimeter","marker":"31"}],"fun_headline_variants":["One polarimeter does two fiber-imaging jobs","Single microscope maps brain fibers with both light modes","Dual-mode polarimeter for combined brain fiber imaging","Merging polarized and scattered light for fiber maps","Scattering polarimetry unites two fiber-imaging techniques"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim that the new device matches state-of-the-art setups rests on visual agreement after manually rotating, transposing, offsetting, and contrast-matching the reference images, with no quantitative angular-difference statistics or pixel-wise error metrics reported to rule out systematic deviations.","fun_headline_variants_meta":{"raw":{"variants":["One polarimeter does two fiber-imaging jobs","Single microscope maps brain fibers with both light modes","Dual-mode polarimeter for combined brain fiber imaging","Merging polarized and scattered light for fiber maps","Scattering polarimetry unites two fiber-imaging techniques"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000599,"raw_usage":{"total_tokens":2832,"prompt_tokens":1010,"completion_tokens":1822,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":626,"completion_tokens_details":{"reasoning_tokens":1747}},"tokens_in":626,"tokens_out":1822,"duration_ms":17575,"temperature":1.0,"reasoning_tokens":1747,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T17:44:04.425317+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"On the same section and pixel grid, compute the fiber direction angle with the Scattering Polarimeter and with a reference 3D-PLI setup, align the data using the paper's own procedure, and measure the median absolute angular difference in the corpus callosum; if the median exceeds a few degrees or varies systematically with brain region, the comparability claim fails. A complementary check would count false-positive crossing detections in regions known to contain only parallel fibers, where the multimodal map should not introduce extra directions.","supporting_citations":[{"cited_title":"NeuroImage54, 1091–1101, DOI: https://doi.org/10.1016/j.neuroimage.2010.08.075 (2011)","cited_arxiv_id":null,"evidence_quote":"supplies the 3D-PLI sinusoidal signal model and the relation between retardation and fiber inclination that the new device's 3D-PLI analysis implements"},{"cited_title":"Adaptive Semi-Supervised Intent Inferral to Control a Powered Hand Orthosis for Stroke","cited_arxiv_id":"2011.00034","evidence_quote":"describes the 3D-PLI measurement procedure and blockface-based registration that the reference 3D-PLI set-up and the reported comparisons build on"},{"cited_title":"Neuroanat.15, DOI: https://doi.org/10.3389/fnana.2021.767223 (2021)","cited_arxiv_id":null,"evidence_quote":"establishes scatterometry-based scattered light imaging that underlies ComSLI's full scattering patterns"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"introduces angular ComSLI and the peak-pair method for extracting in-plane fiber directions, the algorithm the new device uses for crossing detection"},{"cited_title":"& Menzel, M","cited_arxiv_id":null,"evidence_quote":"provides the software toolbox used to evaluate the ComSLI measurements and compute the fiber direction and vector maps"},{"cited_title":"& Chipman, R","cited_arxiv_id":null,"evidence_quote":"gives the Lu-Chipman decomposition used in the Mueller polarimetry analysis and in the error-propagation discussion"},{"cited_title":"& Drevillon, B","cited_arxiv_id":null,"evidence_quote":"supplies the eigenvalue calibration method used to calibrate the measured Mueller matrices of the Scattering Polarimeter"}],"review_version":1}