{"id":"22090589-2d24-4d24-bbb7-e7decfac608b","arxiv_id":"2505.20093","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Multiplex MRE with four drivers and 30-60 Hz vibration reaches near-full shear-wave coverage of the abdomen, including the pancreas, while single-driver single-frequency fails in deep organs.","lead":"This study tested combinations of multiple vibration drivers and frequencies for MR elastography in 18 healthy volunteers and 14 pancreatic cancer patients, finding that four drivers and 30-60 Hz cover nearly the whole abdomen, including the pancreas. The result gives a concrete setup recommendation for whole-abdomen stiffness imaging, with potential use in pancreatic cancer diagnosis and monitoring.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim rests on wave-coverage above a 4 µm noise threshold, but no validation links that threshold to accurate SWS reconstruction; this gap is load-bearing for 'enables tomographic mapping.'","rationale":"The reader's weakest-assumption analysis and my stress-test converge on the same load-bearing gap: coverage above a noise-derived displacement threshold is used as a proxy for tomographic stiffness accuracy. The paper is internally consistent and the experimental design is a strength: 16 configurations per volunteer, a null vibration experiment, publicly available k-MDEV inversion, and motion correction are all concrete and reproducible. The direction of the effect is also credible—more drivers and lower frequencies objectively increase wave amplitude and homogeneity, with p < 0.001. However, the central claim is worded as 'enables tomographic mapping,' which implies quantitative stiffness maps, not merely wave presence. The 4 µm threshold is derived from the same dataset used to define success, making the absolute success rates circular unless the threshold is independently validated as an accuracy threshold. A phantom calibration with known shear modulus would settle this directly: if the amplitude threshold needed for unbiased SWS exceeds 4 µm, the reported coverage percentages overstate the method's tomographic reliability. I would keep the CONDITIONAL verdict rather than downgrade further, because the missing piece is a validation step that can be added without changing the core engineering recommendation. The paper should not be accepted as a definitive technical standard until that validation is provided.","tokens_in":10034,"tokens_out":5687,"duration_ms":64529,"concrete_test":"Scan a homogeneous elasticity phantom (known SWS, e.g., 2 m/s, with a pancreas-depth insert) using the identical 4-driver, 30-60 Hz sequence and k-MDEV pipeline. Compute SWS bias as a function of local displacement amplitude and find the threshold A* at which bias is below 10%. If A* > 4 µm, recompute the Figure 4C/6 coverage percentages with A*; if whole-abdomen or pancreas coverage drops below the reported values, the central claim is not supported. If A* <= 4 µm, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The strongest claim is that four drivers plus 30-60 Hz excitation 'enables tomographic mapping of tissue stiffness across the entire abdomen.' The evidence for this is coverage: the percentage of voxels whose displacement amplitude exceeds 4 µm (Figure 4C, Figure 6). The 4 µm threshold is derived from the PDF intercept between vibration-on and vibration-off amplitudes in the same healthy-volunteer cohort (Methods, Quality Metrics (2); Results, Figure 3A). This threshold is a reasonable noise floor for wave presence, but it is not established as an accuracy threshold for stiffness reconstruction. Shear-wave amplitude is necessary for k-MDEV inversion, but not sufficient: reflections, mode conversion, partial-volume effects, and insufficient illumination geometry can still bias SWS in voxels with amplitudes above 4 µm. The only additional quality metrics are amplitude CV and Laplacian sharpness; sharpness measures anatomical contrast, not quantitative SWS accuracy. No phantom, no reference-standard comparison, and no cross-configuration consistency check is reported. Therefore coverage percentages support 'waves were present in these voxels,' not 'stiffness was correctly reconstructed in these voxels.' This is the load-bearing gap: if voxels between 4 µm and the true accuracy threshold are included, the 99.9% whole-abdomen and >60% pancreas success rates may substantially overstate tomographic reliability.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a technical evaluation of multiplex MR elastography—combining up to four independent pneumatic drivers and vibration frequencies in the 30–60 Hz range—for whole-abdomen stiffness mapping with emphasis on the pancreas. Eighteen healthy volunteers each underwent 16 MRE acquisitions covering all combinations of one to four drivers and frequency sets (60, 50–60, 40–60, 30–60 Hz) plus a no-vibration null acquisition; 14 patients with PDAC were retrospectively analyzed with the four-driver/four-frequency setup. Coverage was defined as the percentage of voxels with harmonic displacement amplitudes above a 4 µm threshold derived from the probability density intercept of vibration-on versus vibration-off amplitudes. The authors report that multi-driver, multi-frequency MRE increases abdominal wave coverage from a median of 24.2% (single-driver, 60 Hz) to 99.9% (four drivers, 30–60 Hz), achieves >60% pancreas coverage in all healthy volunteers (vs. 2/18 for single-driver, single-frequency), and yields 99.1% pancreas coverage in PDAC patients. They conclude that multiplex MRE enables tomographic stiffness mapping of the abdomen, including the pancreas, in under 3 minutes of free-breathing acquisition.","tokens_in":10405,"tokens_out":6591,"duration_ms":62528,"significance":"The study provides a valuable systematic comparison of driver/frequency configurations and a practical recommendation for a feasible clinical setup. Its strengths include a rigorous prospective protocol (17 examinations per healthy volunteer, with a null control), clear statistical testing of driver and frequency effects, an openly available inversion pipeline, and a separate patient cohort. The coverage data convincingly show that four drivers and 30–60 Hz excitation propagate waves more uniformly throughout the abdomen than single-driver/single-frequency excitation. However, the central claim that the multiplex configuration 'enables tomographic mapping' is not directly established: the evidence is based on wave-presence coverage above a noise floor, not on validation of reconstructed shear-wave speed accuracy. If that gap is closed with a phantom or reference-standard comparison, the work would be a solid technical recommendation; in its current form, the significance is real but the conclusion is overreaching.","major_comments":[{"comment":"The 4 µm threshold is defined operationally as the PDF intercept of vibration-on versus vibration-off amplitudes in the same healthy-volunteer cohort, and all coverage and success claims are based on it. This threshold is a sensible noise floor for wave presence, but the manuscript provides no evidence that displacement amplitude above this threshold is sufficient for accurate shear-wave speed reconstruction by k-MDEV. Reflections, mode conversion, partial-volume effects, and unfavorable illumination geometry can bias SWS even in voxels with amplitudes above 4 µm; image sharpness (variance of Laplacian) measures anatomical contrast, not quantitative accuracy. No phantom with known stiffness, no reference-standard comparison, and no cross-configuration consistency analysis is presented. Consequently, the reported coverage percentages support 'waves were present in these voxels,' not 'stiffness was correctly reconstructed in these voxels,' and the conclusion that multiplex MRE 'enables tomographic mapping' is not established. A phantom validation or a comparison of SWS against an independent reference, or at least an analysis demonstrating that SWS is insensitive to driver configuration in well-covered voxels, is needed to support the central claim.","section":"Methods, Quality Metrics (2); Results, Figures 3A, 4C, 6"},{"comment":"The reported SWS values depend substantially on the number of drivers even for a fixed frequency set; for example, in the kidney at 60 Hz, SWS increases from 2.30±0.27 m/s (one driver) to 2.77±0.30 m/s (four drivers), and at 30–60 Hz from 1.65±0.22 to 2.06±0.29 m/s; similar but smaller shifts are seen in liver and pancreas. If the method were already tomographically accurate wherever the wave amplitude exceeds the threshold, the reconstructed SWS for a given frequency set should not change systematically with driver count. This driver-count dependence suggests that coverage alone does not guarantee inversion accuracy and that the four-driver configuration may still yield biased SWS estimates in some regions. The authors should quantify and discuss these differences, and validate the recommended configuration against a reference standard before claiming that the multiplex setup enables tomographic mapping.","section":"Results, Supplemental Table S1"}],"minor_comments":[{"comment":"The statement 'Superficial organs were adequately assessed in all configurations' is contradicted by Figure 6, where single-driver 60 Hz MRE achieved a >60% coverage success rate in only 4/18 (22%) of volunteers for the liver; please rephrase or provide a different definition of 'adequately assessed.'","section":"Abstract, Results"},{"comment":"The sentence 'mean amplitudes and IQR decreased with -0.38 µm/Hz ... and -0.06 µm/Hz' should read 'decreased at rates of 0.38 µm/Hz and 0.06 µm/Hz' to avoid the awkward double negative.","section":"Results, Figure 3B and Table 2"},{"comment":"The sentence 'Patients with PDAC were investigated with an MRE setup similar to the multiplex study... Details of MRI scanners and sequences are provided in [13]' refers to a prior publication; because the PDAC results are used to validate the recommendation, the imaging parameters and processing pipeline should be described with sufficient detail in this manuscript or in a way that is transparent to the reader.","section":"Methods, MRE Setup and Acquisition"},{"comment":"The reference values for liver, kidney, spleen, and pancreas SWS are presented without confidence intervals or information on the number of subjects used for each value; please specify the configuration and sample size for these reference values.","section":"Discussion"}],"recommendation":"major_revision","confidential_remarks":"The paper relies heavily on the authors' own tomoelastography and multifrequency MRE methods, and the PDAC cohort is drawn from a prior publication; this is transparently disclosed. For a journal in physics/medicine, the manuscript's fit is good as a technical report. The main editorial concern is whether the term 'tomographic mapping' should be softened to 'wave coverage' unless additional validation is provided."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a useful, well-executed technical study that gives the community a clear recipe—four drivers, 30–60 Hz, under three minutes—for whole-abdomen MRE coverage. The novelty is real: nobody has done the factorial comparison of driver count and frequency set in the same subjects. The coverage data are internally consistent, the statistics are appropriate, and the PDAC cohort, though small, supports feasibility.\n\nWhat it does well: 18 volunteers scanned with 16 configurations plus a null run is a serious acquisition effort. The PDF-based 4 µm threshold for wave presence is derived cleanly. The organ-specific success rates at 25/60/95% coverage give a useful robustness picture. The discussion is honest about not going beyond four drivers and about the retrospective patient selection.\n\nWhere it is soft: the conclusion says multiplex MRE 'enables tomographic mapping of tissue stiffness,' but the evidence is wave coverage above 4 µm, not stiffness accuracy. Amplitude above a noise floor is necessary but not sufficient for reliable k-MDEV inversion—reflections, mode conversion, and illumination geometry can still bias SWS in voxels that pass the amplitude test. The threshold is also derived from the same healthy-volunteer data that defines success, though as an operational noise-floor definition that is defensible. What is missing is a phantom or reference-standard comparison, or at least a consistency check showing that SWS values in overlapping regions agree across configurations. Given the known frequency dispersion in Table S1, a cross-configuration comparison would have been feasible and would directly test whether coverage equals invertible stiffness. Without that, the 99.9% whole-abdomen success rate is really a wave-presence rate.\n\nThat said, the direction of the effect is not in doubt: more drivers and lower frequencies clearly improve wave penetration and homogeneity, and the practical recommendation is reasonable. The overreach is in the wording of the conclusion, not in the underlying data.\n\nWho it is for: anyone setting up abdominal MRE or planning multi-organ stiffness studies. A serious referee should engage—the methods are reproducible, the volunteer dataset is valuable, and the recommendation will likely become a technical reference. My recommendation: send it to review, and push for either phantom validation or a softened conclusion.","headline":"A practical, systematic MRE setup comparison with a clear recommendation, but the 'tomographic mapping' claim leans on a wave-presence proxy rather than validated stiffness accuracy.","tokens_in":10926,"tokens_out":1988,"would_cite":true,"duration_ms":21063,"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":"Four drivers and 30–60 Hz waves map whole-abdomen stiffness","keywords":["MR elastography","multifrequency","multi-driver","shear wave speed","pancreas","pancreatic ductal adenocarcinoma","abdominal stiffness mapping","tomoelastography"],"falsifier":"Acquire multiplex MRE with four drivers at 30–60 Hz of a tissue-mimicking phantom with known shear modulus, then compare shear-wave speed estimates in voxels whose displacement amplitudes sit just above the 4 µm threshold against the phantom's known value; if those voxels are systematically biased, the coverage percentages overstate tomographic accuracy.","tokens_in":9849,"feed_emoji":"","tokens_out":4676,"duration_ms":52730,"temperature":0.7,"pith_summary":"Standard single-driver MR elastography often fails to send measurable shear waves into deep abdominal organs, and the pancreas is the hardest case. This paper argues that multiplex MRE — four independent drivers vibrating the abdomen at 30–60 Hz simultaneously — is the technical configuration that makes whole-abdomen tomographic stiffness mapping practical. In 18 healthy volunteers, coverage with wave amplitudes above the 4 µm noise threshold rose from a median of 24.2% (single driver, 60 Hz) to 99.9% (four drivers, 30–60 Hz), and the pancreas was sufficiently covered in every subject only in the multiplex configuration. The same setup covered 99.1% of the pancreas and 96.3% of the abdomen in 14 patients with pancreatic ductal adenocarcinoma, in a free-breathing scan under three minutes. The paper treats wave amplitude above 4 µm as the quality gate for reliable stiffness reconstruction and, on that basis, recommends this setup for tomographic abdominal MRE.","feed_headline":"Four drivers and 30–60 Hz waves map whole-abdomen stiffness","feed_subtitle":"Single-driver MRE fails in the pancreas; four drivers and four frequencies succeed in every healthy volunteer and PDAC patient.","key_machinery":"The load-bearing mechanism is multiplex shear-wave excitation: four independent compressed-air drivers placed around the chest, driven at four frequencies (30, 40, 50, 60 Hz) whose wave fields are combined in one acquisition. Lower frequencies produce larger displacement amplitudes (mode 15.5 µm at 30 Hz versus 4.6 µm at 60 Hz), and more drivers make the field more homogeneous, pushing the fraction of abdominal voxels above the 4 µm noise threshold from roughly a quarter to nearly all. Displacement amplitudes above that threshold define wave coverage, and the k-MDEV inversion method converts the multifrequency shear-wave data into shear-wave speed maps.","core_discovery":"The paper's central claim is that multiplex MRE with at least four drivers and multiple vibration frequencies between 30 and 60 Hz enables tomographic mapping of tissue stiffness across the entire abdomen, including the pancreas. The authors show that both ingredients matter: the number of drivers raises and homogenizes shear-wave displacement amplitudes, while including lower frequencies adds larger-amplitude waves, and only the four-driver, four-frequency combination reached 99.9% median abdominal coverage in healthy volunteers and 99.1% pancreatic coverage in PDAC patients. They state this as a technical recommendation for any center wanting tomographic abdominal MRE, noting that the setup costs little, uses parallel pressure tubes from one air outlet, and fits into routine protocols at under three minutes of scan time.","pith_inferences":["The paper validates coverage, not stiffness accuracy; the next testable step is to compare multiplex-MRE shear-wave speed in the pancreas and in PDAC against histology or endoscopic ultrasound elastography in the same lesions.","Because the 4 µm threshold was calibrated in healthy, normal-BMI volunteers, applying the same threshold to patients with ascites, high BMI, or post-surgical anatomy is an assumption the paper flags for future testing; a per-subject noise calibration would remove that risk.","If amplitude coverage is confirmed to imply accuracy, multiplex MRE could be combined with other quantitative MRI readouts to screen for systemic mechanical changes, such as early liver involvement before metastasis in pancreatic cancer."],"forward_implications":["A single free-breathing MRE acquisition under three minutes can produce tomographic stiffness maps of liver, spleen, kidney, and pancreas at once.","Pancreatic tissue, the organ most often failed by single-driver MRE, is covered above the quality threshold in all 18 healthy volunteers and all 14 PDAC patients with the multiplex setup.","Including frequencies down to 30 Hz is essential, since lower frequencies carry the largest wave amplitudes and improve the sharpness of the stiffness maps.","The reported healthy reference shear-wave speeds (pancreas 1.31 ± 0.12 m/s, liver 1.36 ± 0.10 m/s, kidney 2.06 ± 0.29 m/s, spleen 2.08 ± 0.32 m/s) can serve as a baseline for detecting stiffness changes in abdominal disease."],"supporting_citations":[{"why":"Provides the prior multi-driver liver MRE result that this study extends and must outperform on whole-abdomen coverage.","marker":"[12]"},{"why":"Supplies the previously published multifrequency, multi-driver MRE cohort from which the 14 PDAC patients were retrospectively drawn.","marker":"[13]"},{"why":"Establishes the externally placed pressurized-air driver design used for multiplex excitation.","marker":"[14]"},{"why":"Documents pancreas MRE repeatability in pancreatic cancer patients and healthy controls, motivating the clinical feasibility question.","marker":"[15]"},{"why":"Supplies the motion-correction and breathing-artifact reduction approach applied before MRE post-processing.","marker":"[16]"},{"why":"Provides the k-MDEV multifrequency inversion method used to reconstruct shear-wave speed maps.","marker":"[19]"},{"why":"Supplies the open-access processing pipeline used for multifrequency MRE inversion and its validation.","marker":"[20]"},{"why":"Underpins the noise-threshold approach used to separate true shear-wave displacement from vibration-off background amplitudes.","marker":"[22]"}],"fun_headline_variants":["Four-driver MRE maps abdominal stiffness including pancreas","Multiplex MRE: four drivers enable whole-abdomen stiffness maps","Two-frequency MRE with four drivers covers entire abdomen","Single-driver fails on pancreas; four drivers succeed","Near-complete stiffness maps with multiplex MRE in abdomen"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion rests on treating shear-wave displacement amplitude above a fixed 4 µm threshold, calibrated from vibration-off noise in healthy volunteers, as proof that stiffness maps in those voxels are tomographically reliable, and on carrying that threshold into patients unchanged.","fun_headline_variants_meta":{"raw":{"variants":["Four-driver MRE maps abdominal stiffness including pancreas","Multiplex MRE: four drivers enable whole-abdomen stiffness maps","Two-frequency MRE with four drivers covers entire abdomen","Single-driver fails on pancreas; four drivers succeed","Near-complete stiffness maps with multiplex MRE in abdomen"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000694,"raw_usage":{"total_tokens":3223,"prompt_tokens":1116,"completion_tokens":2107,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":732,"completion_tokens_details":{"reasoning_tokens":2025}},"tokens_in":732,"tokens_out":2107,"duration_ms":17101,"temperature":1.0,"reasoning_tokens":2025,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T13:58:21.061102+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Acquire multiplex MRE with four drivers at 30–60 Hz of a tissue-mimicking phantom with known shear modulus, then compare shear-wave speed estimates in voxels whose displacement amplitudes sit just above the 4 µm threshold against the phantom's known value; if those voxels are systematically biased, the coverage percentages overstate tomographic accuracy.","supporting_citations":[{"cited_title":"Abdominal MR elastography with multiple driver arrays: performance and repeatability","cited_arxiv_id":null,"evidence_quote":"Provides the prior multi-driver liver MRE result that this study extends and must outperform on whole-abdomen coverage."},{"cited_title":"Tomoelastography for Measurement of Tumor Volume Related to Tissue Stiffness in Pancreatic Ductal Adenocarcinomas","cited_arxiv_id":null,"evidence_quote":"Supplies the previously published multifrequency, multi-driver MRE cohort from which the 14 PDAC patients were retrospectively drawn."},{"cited_title":"Tomoelastography of the prostate using multifrequency MR elastography and externally placed pressurized‐air drivers","cited_arxiv_id":null,"evidence_quote":"Establishes the externally placed pressurized-air driver design used for multiplex excitation."},{"cited_title":"Elastography of the Pancreas: Bowel Preparation and Repeatability Assessment in Pancreatic Cancer Patients and Healthy Controls","cited_arxiv_id":null,"evidence_quote":"Documents pancreas MRE repeatability in pancreatic cancer patients and healthy controls, motivating the clinical feasibility question."},{"cited_title":"Reduction of breathing artifacts in multifrequency magnetic resonance elastography of the abdomen","cited_arxiv_id":null,"evidence_quote":"Supplies the motion-correction and breathing-artifact reduction approach applied before MRE post-processing."},{"cited_title":"Tomoelastography by multifrequency wave number recovery from time-harmonic propagating shear waves","cited_arxiv_id":null,"evidence_quote":"Provides the k-MDEV multifrequency inversion method used to reconstruct shear-wave speed maps."},{"cited_title":"Rapid MR elastography of the liver for subsecond stiffness sampling","cited_arxiv_id":null,"evidence_quote":"Underpins the noise-threshold approach used to separate true shear-wave displacement from vibration-off background amplitudes."}],"review_version":1}