{"id":"5d6cd9dd-2c00-445f-a932-8dc48c1b6da1","arxiv_id":"2608.03208","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A modular FPGA-based DAQ system with 16-bit, 100 MS/s digitization and on-board DSP acquires microdosimetric spectra in clinical proton and carbon beams, achieving an ENC of about 175 electrons.","lead":"The Spectacular DAQ system is a modular, FPGA-based data acquisition platform for microdosimetry, tested with a diamond detector in proton and carbon-ion therapy beams at MedAustron. It aims to make microdosimetric measurements faster and more practical for clinical ion-beam therapy.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Pileup rejection at clinical rates is unquantified; without rate/pileup data the reported spectra may be distorted, so the central claim is not yet supported.","rationale":"The paper describes a promising modular DAQ platform and provides proof-of-principle spectra, but the most load-bearing part of the argument is the assertion that the acquisition works at clinical rates without pileup distortion. The reader's weakest-assumption analysis identifies exactly this: the PUR scheme's effectiveness is asserted, not demonstrated. My own reading finds no internal inconsistency—the hardware description is coherent, the ENC measurement is a genuine data point, and the authors honestly flag the proton cutoff as a noise limitation. However, the carbon full-dynamic-range claim and the quantitative comparison to GATE cannot be assessed without knowing whether the accepted-event spectrum is rate-independent. The proposed two-rate test directly probes whether residual pileup changes the spectral shape. If the test shows rate independence, the central claim survives; if not, the spectra do not support the stated clinical-rate capability. Thus the conditional verdict should stand, pending this pileup characterization.","tokens_in":7339,"tokens_out":4539,"duration_ms":57740,"concrete_test":"Re-acquire the carbon depth-dose spectra at, e.g., the 2.9 cm WET position at two beam intensities separated by a factor of ~5 (or with added attenuation), using identical PUR settings and live-time normalization. If the normalized yd(y) distributions agree within statistical uncertainties, pileup distortion is not significant at the tested rate; if they differ, the reported spectra are rate-dependent and the 'full clinical dose-rate' claim needs qualification. Additionally, report the measured fast-channel trigger rate, accepted-event rate, and inhibited-event fraction for each run; a pileup fraction > a few percent or a rate-dependent spectral shift would settle the issue.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—acquisition of undistorted microdosimetric spectra under full clinical dose-rate conditions (Section 4)—rests on the fast-slow PUR scheme (Section 3) preventing spectral distortion. No event rate, pileup fraction, dead time, or comparison against a pileup-free reference is reported. The scheme is only specified as a 50 ns fast shaper with inhibition up to fourth order; events separated by less than the fast-channel resolving time, or fifth-and-higher-order coincidences, are not covered, and a finite inhibition dead time can itself distort the accepted-event distribution. Since pileup predominantly adds artificial high-y events, the claimed full dynamic range for carbon and the dose-weighted spectra yd(y) could be biased. The authors cite micro-spill time structure (refs. [22,23]) as enabling improved event separation, but no measured spill-synchronized rates or spill-phase utilization are given. This is an addressable omission, not an internal contradiction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript describes Spectacular, a modular DAQ platform for solid-state microdosimetry. The system is based on a Xilinx Zynq UltraScale+ SoC, a 16-bit 100 MS/s ADC, interchangeable front-end daughterboards, and FPGA/CPU/host processing. The authors report a discrete CSA with ~175 e ENC at 1 µs shaping time, verified linearity up to 250 fC, and a fast-slow pileup rejection scheme using a 50 ns fast shaper, a 1 µs slow shaper, and inhibition up to fourth order. Beam tests at MedAustron with a 10 µm diamond microdosimeter produced microdosimetric spectra for 62.4 MeV protons and 120 MeV/u carbon ions at several depths, calibrated with the proton edge and compared qualitatively with GATE simulations. The central claims are that the platform can acquire spectra under full clinical dose-rate conditions, that the full dynamic range for carbon ions is covered, and that pulse pileup is mitigated by the implemented PUR scheme.","tokens_in":7604,"tokens_out":5284,"duration_ms":63393,"significance":"If the rate capability and spectral fidelity claims are substantiated, this would be a useful contribution to clinical microdosimetry instrumentation. The paper has clear strengths: the ENC was measured with the sensor attached in the irradiation room, the spectra are calibrated against the external proton-edge standard, the comparison to independent GATE simulations provides a reference for the measured shapes, and the hardware architecture is documented in enough detail to be reproduced. The authors also explicitly disclose the proton low-energy cutoff, which is a welcome limitation statement. However, the central claim of acquiring undistorted spectra under full clinical dose-rate conditions is not yet supported because no quantitative rate or pileup information is provided. This is an addressable gap, not an internal inconsistency, and should be fixed before publication.","major_comments":[{"comment":"The central claim that spectra were acquired under full clinical dose-rate conditions rests on the fast-slow PUR scheme, but no event rate, pileup fraction, or pileup-free reference is reported. The scheme is specified only as a 50 ns fast shaper and inhibition 'up to the fourth order'; the fast-channel resolving time, inhibition window, dead time, and inhibited fraction are missing. Events closer than the fast-channel resolving time, and fifth-and-higher-order coincidences, are unaddressed, and the inhibition dead time itself can bias the accepted-event distribution. Since residual pileup adds artificial high-y events, the reported yd(y) spectra could be distorted. Please report rates per depth, quantify PUR efficiency/dead time, and validate one spectrum against a reduced-rate or pileup-free acquisition. The cited micro-spill structure [22,23] does not replace measured spill-synchroniz","section":"§3 Processing; §4 Characterization"},{"comment":"Figure 4 shows no error bars or uncertainties, and the agreement with GATE is only qualitative. Because the conclusions depend on spectral shapes and dynamic-range coverage, please add statistical and calibration uncertainties and a quantitative comparison (e.g., y_F, y_D, or bin-wise residuals) for representative depths. Also specify the GATE physics list, detector geometry, and energy-deposition model used; citing [13] is insufficient.","section":"§4, Figure 4"},{"comment":"The proton spectra are subject to a lower cutoff of 3.5 keV/µm, so shallow-depth spectra are not fully resolved and the dose-weighted yd(y) is truncated. This materially qualifies the claimed clinical-rate acquisition for protons. Please quantify the fraction of events and dose below the cutoff from the GATE simulation, show its impact on y_F and y_D, and state this limitation explicitly in the abstract and conclusions.","section":"§4, proton spectra"}],"minor_comments":[{"comment":"Typographical errors: 'By enabling optimization fast testing' in the abstract; 'Spectacularisbasedaroud' and 'daugtherboards' in §3. Please proofread throughout.","section":"Abstract; §3"},{"comment":"Terminology is inconsistent: 'pile-up' and 'pileup' are both used, and the simulation code is sometimes 'Gate' and sometimes 'GATE'. Please standardize.","section":"§2–§4"},{"comment":"References [1]–[3] and [22]–[24] lack complete bibliographic information (journal, volume, pages, or DOI). Please complete the entries.","section":"References"},{"comment":"The phrase 'up to the fourth order' is ambiguous: does it mean fourth-order coincidences or inhibition of up to four subsequent pulses? Please define the order and the corresponding time window.","section":"§3 Processing"},{"comment":"No nominal clinical dose rates or particle fluences are stated for the beam tests. Please report the dose rate and, if available, the count rate observed by the detector at each depth.","section":"§4 Characterization"},{"comment":"Caption contains a missing space ('10µmdiamond') and the legend labels may be hard to distinguish for color-blind readers. Consider markers or distinct line styles.","section":"Figure 1"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about unquantified pileup lands. The instrumentation contribution is real and the spectral results are plausible, but the clinical-rate claim cannot be evaluated without rate and pileup data. I view this as a major revision rather than a rejection, since the missing measurements are within the scope of the presented system and facility access."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read the Spectacular DAQ paper. My take: it's a solid hardware contribution that is not yet a solid physics result. The new thing is the platform itself—Zynq SoC, 16-bit 100 MSa/s ADC, interchangeable daughterboards, on-board FPGA pulse processing, programmable supplies, test-pulse and active reset. That's genuinely useful. The microdosimetry community has been using external digitizers and off-board processing; integrating everything on one board is a real step. The ENC measurement is credible, and the authors are honest that it falls short of their own <100 e target.\n\nThe prominent claim—spectra acquired under full clinical dose-rate conditions—does not hold up yet. The fast-slow PUR scheme is described as a 50 ns fast shaper with inhibition up to fourth order, but no event rate, pileup fraction, dead time, or comparison to a low-rate reference is reported. Without that, the reader cannot know whether the spectra are distorted. The stress-test note is right: this is the load-bearing claim and the evidence is missing. The proton spectra also required a post-hoc 3.5 keV/µm cutoff, so the 'full dynamic range' claim only applies to carbon. Figure 4 has no error bars, and the GATE comparison is qualitative.\n\nThe paper is nevertheless worth refereeing. It is clearly written, the architecture is reproducible from the description, and the authors flag their own limitations. The missing pieces are addressable: report measured rates, PUR efficiency, and a comparison run at reduced dose rate. I would not desk-reject this; I would send it to a referee who knows microdosimetry and ask for those additions.\n\nFor a reading group, it's a reasonable maybe—the pileup question is relevant to anyone doing spectroscopy in pulsed beams. I wouldn't cite it in my own work yet, but if the authors provide the rate/pileup characterization in a revision, it becomes citable.\n\nSend it to review, with a clear request for the missing quantification.","headline":"A genuinely useful modular DAQ platform, but the clinical-dose-rate pileup claim is unsupported without rate and PUR characterization.","tokens_in":8053,"tokens_out":2795,"would_cite":false,"duration_ms":28526,"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":"A modular data-acquisition system with roughly 175-electron noise records proton and carbon-ion microdosimetric spectra at full clinical dose rates, covering the full carbon lineal-energy range.","keywords":["microdosimetry","linear energy transfer","particle therapy","data acquisition system","pulse pileup rejection","diamond microdosimeter","charge-sensitive amplifier","digital pulse shaping"],"falsifier":"Acquire the same lineal-energy spectrum twice at the same beam position: once at full clinical dose rate and once with the beam attenuated until pileup is negligible. If the two $y\\,d(y)$ distributions differ — a rate-dependent shift in the carbon or proton edge, or a suppression of high- or low-LET events — the pileup rejection is not complete. An equivalent test is to histogram the inter-event times from the $50\\,\\mathrm{ns}$ fast channel and compare against the expectation from the measured micro-spill structure; a mismatch would show that fourth-order inhibition is insufficient.","tokens_in":7283,"feed_emoji":"⚛️","tokens_out":19173,"duration_ms":171835,"temperature":0.7,"pith_summary":"This paper argues that a single modular data-acquisition platform can meet both conflicting demands of clinical microdosimetry: resolving the small energy deposits (of order $10^3$ electrons) from $\\mu$m-scale detectors while rejecting pileup at therapeutic ion-beam rates near $10^{10}\\,\\mathrm{s^{-1}}$. The authors built \"Spectacular\", centered on a Zynq UltraScale+ system-on-chip, a 16-bit $100\\,\\mathrm{MSa\\,s^{-1}}$ ADC, and interchangeable front-end daughterboards, and tested it at a clinical ion-therapy center with a custom discrete charge-sensitive amplifier (equivalent noise charge of $\\sim 175\\,e$ at the optimal $1\\,\\mu\\mathrm{s}$ shaping time) and a $10\\,\\mu\\mathrm{m}$ diamond microdosimeter. With fast-slow pileup rejection (a $50\\,\\mathrm{ns}$ fast shaper inhibiting acquisition up to fourth order), the system recorded lineal-energy spectra for $62.4\\,\\mathrm{MeV}$ protons and $120\\,\\mathrm{MeV/u}$ carbon ions along the depth-dose curve at full clinical dose rates; for carbon the whole dynamic range was covered, while for protons a lower cutoff of $3.5\\,\\mathrm{keV}\\,\\mu\\mathrm{m^{-1}}$ was required because entrance-channel low-LET events were not yet resolvable. If the claim holds, microdosimetry moves from research prototypes toward routine clinical use as an experimental source of radiation-quality data for treatment planning and quality assurance in particle therapy; the authors state that a dedicated CSA ASIC and revised boards are expected to uncover the currently inaccessible proton range.","feed_headline":"175 e of noise maps ion-beam microdosimetry in the clinic","feed_subtitle":"Modular Zynq platform with fast-slow pileup rejection captures full carbon-ion spectra; proton low-LET events need a quieter amplifier.","key_machinery":"The carrying mechanism is one reconfigurable pipeline: a daughterboard-mounted charge-sensitive amplifier feeds a 16-bit, $100\\,\\mathrm{MSa\\,s^{-1}}$ ADC, and the FPGA shapes the digitized trace with a finite-impulse-response kernel implementing a fourth-order quasi-Gaussian filter. Two time constants do the work: a $1\\,\\mu\\mathrm{s}$ slow shaper extracts amplitudes with maximum signal-to-noise ratio, while a $50\\,\\mathrm{ns}$ fast shaper flags closely spaced events and vetoes acquisition up to fourth order (the fast-slow pileup-rejection method). The FPGA resets the CSA on threshold crossings, synchronizes to the accelerator extraction trigger, and delta-compresses the $200\\,\\mathrm{MB\\,s^{","core_discovery":"On its own terms, the paper claims that the Spectacular platform — a Zynq UltraScale+ SoC, 16-bit ADC at $100\\,\\mathrm{MSa\\,s^{-1}}$, and a standardized daughterboard interface — turns microdosimetry readout into a reconfigurable chain: swap the front end, keep the digitization, shaping, and pileup rejection. A discrete CSA reached ENC $\\sim 175\\,e$ at $1\\,\\mu\\mathrm{s}$ shaping. With a fourth-order quasi-Gaussian digital shaper and a $50\\,\\mathrm{ns}$ fast channel vetoing up to fourth order, it recorded spectra for $62.4\\,\\mathrm{MeV}$ protons and $120\\,\\mathrm{MeV/u}$ carbon ions at full clinical dose rate, calibrated by the proton edge. Carbon covered the full dynamic range; protons requi","pith_inferences":["The paper reports no per-spill event rates or pileup fractions, so the \"full clinical dose rate\" claim stands or falls with the fast-slow rejection assumption; a rate sweep or a comparison against a pileup-free reference spectrum would settle it directly. (This is my inference, not the paper's claim.)","The digitize-early, shape-digitally pipeline is parameterized only by filter time constants, so the same architecture transfers to other high-rate single-event spectroscopy problems, such as LET spectrometry in mixed radiation fields or neutron/gamma discrimination.","Because synchrotron beams arrive in a micro-spill structure correlated with the machine orbital frequency, gating on the spill phase could reject a class of pileup events that a free-running $50\\,\\mathrm{ns}$ veto misses; the paper's extraction-trigger synchronization option is the natural first step toward testing that.","The multiplexed single-ADC prototype leaves the four-channel claim untested under simultaneous load; a four-channel acquisition would reveal whether the shared digitizer introduces dead time or cross-talk that single-channel results cannot show."],"forward_implications":["If the platform performs as claimed, the same carrier board evaluates different charge-sensitive front ends, ASICs, and detectors without redesigning the digitization and processing chain, accelerating sensor development for microdosimetry.","Carbon-ion beams can be characterized across the entire lineal-energy dynamic range at clinical dose rates, supplying experimental radiation-quality data for treatment-planning and quality-assurance workflows.","Once the noise target below $100\\,e$ is reached with the dedicated CSA ASIC and revised boards, the proton entrance-channel range below $3.5\\,\\mathrm{keV}\\,\\mu\\mathrm{m^{-1}}$ becomes accessible, closing the remaining gap.","Synchronizing acquisition with the extraction trigger and exploiting the micro-spill time structure of synchrotron beams offers an additional lever against pileup on top of the fast-slow veto.","Multi-channel simultaneous readout, currently limited to one digitized channel in this prototype, is the natural next step enabled by the four-channel carrier design and FPGA-side processing."],"supporting_citations":[{"why":"Supplies the direct-synthesis method for the fourth-order quasi-Gaussian filter whose FIR kernel performs the 1 µs amplitude shaping.","marker":"[25]"},{"why":"Provides the particle-edge calibration procedure (proton edge) used to convert measured pulse heights into lineal energy.","marker":"[12]"},{"why":"Provides the Monte Carlo depth-dose curves and reference spectra used for comparison with the measured spectra.","marker":"[13]"},{"why":"Describes the fabrication of the CVD diamond microdosimeter array (ten-micrometre Schottky diodes) used in the beam tests.","marker":"[16]"},{"why":"Documents the micro-spill time structure of medical accelerator beams and its impact on pulse pileup, motivating the pileup-rejection design and trigger synchronization.","marker":"[23]"},{"why":"Exemplifies prior solid-state microdosimetry readouts limited to front-end and digitization, the gap the modular platform fills.","marker":"[14]"}],"fun_headline_variants":["Modular DAQ: swap front-ends, keep pileup rejection","Zynq SoC readout: 175-e noise at full clinical ion rates","Proton and carbon spectra, one platform: Spectacular DAQ","Reconfigurable microdosimetry: test any sensor on the same DAQ","Pileup vetoed, spectra captured: modular DAQ for ion beams"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The fast-slow pileup rejection (a $50\\,\\mathrm{ns}$ fast shaper vetoing up to fourth order) is assumed to keep spectra free of pileup distortion at full clinical dose rates, but the paper reports no measured event rate, pileup fraction, or comparison against a pileup-free reference.","fun_headline_variants_meta":{"raw":{"variants":["Modular DAQ: swap front-ends, keep pileup rejection","Zynq SoC readout: 175-e noise at full clinical ion rates","Proton and carbon spectra, one platform: Spectacular DAQ","Reconfigurable microdosimetry: test any sensor on the same DAQ","Pileup vetoed, spectra captured: modular DAQ for ion beams"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000426,"raw_usage":{"total_tokens":2085,"prompt_tokens":877,"completion_tokens":1208,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":621,"completion_tokens_details":{"reasoning_tokens":1108}},"tokens_in":621,"tokens_out":1208,"duration_ms":14763,"temperature":1.0,"reasoning_tokens":1108,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T23:26:15.250876+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Acquire the same lineal-energy spectrum twice at the same beam position: once at full clinical dose rate and once with the beam attenuated until pileup is negligible. If the two $y\\,d(y)$ distributions differ — a rate-dependent shift in the carbon or proton edge, or a suppression of high- or low-LET events — the pileup rejection is not complete. An equivalent test is to histogram the inter-event times from the $50\\,\\mathrm{ns}$ fast channel and compare against the expectation from the measured micro-spill structure; a mismatch would show that fourth-order inhibition is insufficient.","supporting_citations":[{"cited_title":"Ohkawa, M","cited_arxiv_id":null,"evidence_quote":"Supplies the direct-synthesis method for the fourth-order quasi-Gaussian filter whose FIR kernel performs the 1 µs amplitude shaping."},{"cited_title":"Conte, D","cited_arxiv_id":null,"evidence_quote":"Provides the particle-edge calibration procedure (proton edge) used to convert measured pulse heights into lineal energy."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Monte Carlo depth-dose curves and reference spectra used for comparison with the measured spectra."},{"cited_title":"Verona, G","cited_arxiv_id":null,"evidence_quote":"Describes the fabrication of the CVD diamond microdosimeter array (ten-micrometre Schottky diodes) used in the beam tests."},{"cited_title":"Measurements of the micro-spill structure of medical cyclotron and synchrotron beams and its impact on pulse pileup","cited_arxiv_id":"2605.07508","evidence_quote":"Documents the micro-spill time structure of medical accelerator beams and its impact on pulse pileup, motivating the pileup-rejection design and trigger synchronization."},{"cited_title":"Rosenfeld,Novel detectors for silicon based microdosimetry, their concepts and applications,","cited_arxiv_id":null,"evidence_quote":"Exemplifies prior solid-state microdosimetry readouts limited to front-end and digitization, the gap the modular platform fills."}],"review_version":1}