{"id":"4519bd2f-e091-40de-85e9-83c173c18d7c","arxiv_id":"2608.08860","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A preview-based, offset-free model-predictive controller tracks a simulated pulsating tissue surface, cutting contact placement error versus feedback baselines while keeping lateral shear within budget via a slack-relaxed octagonal constraint.","lead":"This simulation study tests a controller that predicts the rhythm of brain tissue motion and steers a surgical insertion tool relative to that moving surface, not toward a fixed point. It reports much smaller placement errors and lateral shear in simulated neural-thread insertion, with real tissue as the open question.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The preview advantage is partly by construction: the benchmark plant and the predictor share the same harmonic exosystem, and Table VIII shows the benefit collapses under broadband/ectopic motion.","rationale":"The reader's weakest assumption identifies the same point. I read the paper as a scoped simulation study; the harmonic exosystem limitation is stated honestly in Section VI-F and Appendix D-B. Nevertheless, this is the single most load-bearing condition because the preview mechanism is exercised on the same model it is built to predict, and the only provided test of the violated assumption (Table VIII) shows the headline margin collapses. A concrete out-of-distribution benchmark would settle whether the advantage extends to realistic tissue motion. Since the paper already flags the limitation and the reader's CONDITIONAL verdict asks for scope alignment and artifacts, I do not move the verdict. The rigid-tip issue is a further scope restriction but is acknowledged and less decisive. I also note the missing Appendix?? placeholders block independent reproduction; that is a process condition, not the main intellectual weakness.","tokens_in":23859,"tokens_out":7638,"duration_ms":85438,"concrete_test":"Rerun the Section VII benchmark with the plant surface generated outside the exosystem family while keeping the controller's harmonic assumption: e.g., replace d(t) with recorded or synthesized human cortical pulsation that has the same RMS and line-spectrum content plus a 40 um RMS broadband/colored-noise component and an ectopic transient (the Table VIII case), and report the same metrics for preview MPC and delayed-feedback impedance across the N=30 protocol. If contact RMS and shear remain below spec and significantly better than the no-preview baseline, the concern is resolved; if they degrade to the ~37 um/7.5 mm/s range shown in Table VIII, the preview dominance is an in-distribution artifact and the paper should be conditioned on out-of-distribution validation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central simulation claim rests on Section III-B's exosystem being the exact generator of the Section VII-A plant pulsation. The predictor's internal model and the plant's disturbance model coincide (same S, same output map), so Table IV's 12.0/1.9 um results describe a matched-model, best-case scenario for preview. The paper's own robustness table (Table VIII) is the stress test: adding a 40 um RMS broadband component plus an ectopic beat raises contact RMS to 37.4 um (from 1.9) and measured contact shear to 7.49 mm/s (from 0.22), i.e. it removes most of the advertised margin and exceeds the 20 um placement spec and 0.80 mm/s shear budget. Because no baseline (delayed-feedback impedance) is reported for that broadband case, the paper does not establish that preview still dominates when its core assumption fails. The flexible-thread limitation is real but secondary; the main claim is already scoped to rigid tips, whereas the exosystem match is presented as a formal carrier of the preview benefit (Section III-B) and is the load-bearing transfer from simulation to pulsating-tissue motivation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a preview-based relative-motion controller for robotic neural-thread insertion into a cortical surface that pulsates with cardiac and respiratory motion. The controller combines a Kalman-filter estimate of a harmonic exosystem model of surface motion with latency forward-propagation, an offset-free acceleration-bias observer, and a receding-horizon QP that enforces actuator and lateral-shear constraints; a soft-octagon shared-slack variant restores feasibility under sensing noise. The paper reports MuJoCo simulations: 1-DOF free-space/contact RMS relative placement of 12.0/1.9 um versus 18.3/176.8 um for delayed-feedback impedance and 286.1/275.5 um for lab-frame PD; 3-DOF lateral shear reduction from 1.34 to 0.50 mm/s; a 5-15 um noise sweep where the soft-octagon controller meets the 0.80 mm/s budget at 10 um with 0/10 violations versus 10/10 for a cost-only variant; and a two-vertex Lyapunov certificate over a -40%/+50% reflected-mass box. The paper is explicitly simulation-only with a rigid tip and lists flexible-thread mechanics, biological damage thresholds, and hardware-realistic sensing and timing as future work.","tokens_in":24043,"tokens_out":8694,"duration_ms":98367,"significance":"If the simulation results are taken at face value, the paper makes a useful task-level contribution: it integrates preview, offset-free contact regulation, and a coupled shear constraint in a single insertion-tool controller and evaluates it with unusually disciplined reporting, including Monte Carlo ablations, reproducible scripts, a numerical LMI audit, and open disclosure of the force/placement and feasibility trade-offs. The central caveat is that the simulated plant's pulsation is generated by the same harmonic exosystem the predictor assumes, so the headline preview advantage is partly a matched-model, best-case result; the paper's own robustness table shows that a broadband/ectopic component removes most of the margin, and no baseline is reported for that case. The flexible-thread limitation is real but secondary because the main claims are explicitly scoped to a rigid tip.","major_comments":[{"comment":"The simulated cortical surface is generated by the same harmonic exosystem that the predictor assumes, so the Table IV improvements (12.0 vs. 18.3 um free-space; 1.9 vs. 176.8 um contact) are matched-model, best-case results for preview. The only test that breaks this assumption (Table VIII, 40 um broadband plus ectopic beat) yields 37.4 um contact RMS and 7.49 mm/s shear, both far outside the stated 20 um / 0.80 mm/s specifications, but the table does not include the delayed-feedback impedance or lab-frame PD baselines. Without those rows, the reader cannot tell whether preview still dominates, ties, or loses when its core assumption fails; please add the same baselines to the broadband/ectopic and combined rows and discuss the ranking.","section":"§VII-A, §III-B, Tables IV and VIII"},{"comment":"The text characterizes the broadband disturbance result as 'graceful degradation,' but the measured numbers (contact RMS 1.9 to 37.4 um, shear 0.22 to 7.49 mm/s, peak force 3.43 to 7.93 mN) are a 20- to 34-fold degradation that violates both design specifications. 'Graceful' is therefore not supported unless it is defined relative to a baseline; I recommend either quantifying the degradation against the feedback and lab-frame baselines or removing the term.","section":"§VI-F, Table VIII"}],"minor_comments":[{"comment":"The two-vertex Lyapunov certificate is proved only for the constraint-inactive feedback component, as Remark 1 clearly states, but the abstract's phrasing 'the controller's actual finite-horizon error-feedback gain' could be misread as certifying the constrained QP that produces the benchmark numbers; please add an explicit qualifier in the abstract.","section":"§VI-G, abstract"},{"comment":"Several unresolved cross-reference placeholders remain, e.g., 'Appendix??' in Section VII-A and in Appendices B-A, B-B, and C-A; these should be resolved before publication.","section":"§VII-A, Appendices"},{"comment":"The column 'Fripple' is not defined; please add a definition in the table caption or in the text near Section VII-D.","section":"Table IV"},{"comment":"There are minor typographical issues such as 'non-harmonicsurface' with a missing space; a proofreading pass is needed.","section":"§VI-F"}],"recommendation":"major_revision","confidential_remarks":"This is a well-scoped simulation paper with unusually honest limitations. My main concern is the matched-model construction of the benchmark and the missing baseline under broadband/ectopic motion, which I believe can be fixed within the manuscript's scope by adding the requested comparisons and rephrasing the 'graceful degradation' claim. I have no concerns about novelty disclosure or citation behavior."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing you should know: this is a carefully scoped simulation study that does what it claims, and it is more honest about its weaknesses than most. The actual new content is narrow but real—task-level synthesis of preview, offset-free contact regulation, and a feasibility-restored octagonal shear row for relative-depth insertion into a 3D pulsating cortical surface. No prior work I can see benchmarks that exact combination. The soft-slack octagon is a sensible fix to QP infeasibility, and the 2x2 ablation showing preview helps free space while offset-free kills the contact offset is clean.\n\nCredit where due: the paper reports its own failure modes. It discloses the hard force-constraint infeasibility, the -50% mass instability, the force/placement trade-off, and the sparse-sensing frequency-drift failure where fixed-frequency preview is worse than feedback. That is the right way to present engineering simulation evidence.\n\nNow the soft spots, in proportion. The biggest one is exactly what the stress-test note flags: the preview advantage is partly by construction. The plant pulsation is generated from the same harmonic exosystem the predictor assumes, so the 12 um vs 18 um free-space gain is a matched-model best case. Table VIII shows the consequence—add a 40 um broadband component and an ectopic beat, and contact error goes from 1.9 to 37 um and shear from 0.22 to 7.49 mm/s. That removes most of the advertised margin, and no baseline is reported for that case, so we don't know if preview still beats feedback when its core assumption fails. That is a genuine gap, not a manufactured one. The rigid-tip limitation is secondary because the paper scopes itself to that.\n\nOther issues are minor and mostly self-flagged: the promised reproduction scripts and appendices are missing (the text has literal 'Appendix??' placeholders), the headline 1-DOF table is a single seed even though the ablation uses 30 seeds, and the abstract's certificate statement is stronger than Remark 1's scope, which explicitly covers only the constraint-inactive feedback law, not the running QP. None of these is load-bearing enough to kill the paper; all are fixable in revision.\n\nBottom line: the central simulation claim—preview plus offset-free MPC reduces relative placement error and lateral shear in this pulsating-tissue model—holds up within its stated scope. It is not a clinical demonstration and the transfer to real tissue is unproven, but the paper says that itself. Control engineers working on physiological motion compensation and neural-interface insertion will get value; the clinical community should read it as motivation, not validation. This deserves a serious referee. I would send it to review with the condition that the artifacts ship, the abstract is aligned with the body's scoping, and either a broadband-motion baseline is added or the claim is explicitly narrowed to harmonic motion.","headline":"Carefully scoped simulation study whose preview benefit is partly by construction—worth refereeing, with conditions, not desk rejection.","tokens_in":24654,"tokens_out":2361,"would_cite":true,"duration_ms":24676,"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":"The paper claims that preview-based relative-motion control with offset-free disturbance rejection can keep an insertion-tool tip within about two microns of a pulsating tissue surface in simulation, beating delayed-feedback and lab-frame…","keywords":["relative-motion control","physiological motion compensation","model predictive control","offset-free disturbance rejection","neural-thread insertion","brain-surface pulsation","lateral shear constraint","physics simulation"],"falsifier":"Run the controller against a recorded cortical-motion trace with broadband noise and occasional ectopic beats in a physical flexible-thread setup; if contact-phase RMS placement error grows far beyond a few microns or the 0.80 mm/s shear budget is violated at 10 µm RMS per-axis sensing noise, the central claim fails. A simpler computational check is to replant the simulation with the paper's own non-harmonic 40 µm RMS plus ectopic disturbance and test whether the soft-octagon controller still completes every seed without a shear violation; the paper's own results indicate it would not.","tokens_in":23540,"feed_emoji":"🧠","tokens_out":8301,"duration_ms":80843,"temperature":0.7,"pith_summary":"The paper argues that neural-thread insertion into pulsating cortex should be treated as a relative-motion problem: regulate the tool tip against the predicted moving tissue surface, not a fixed lab-frame target. It develops a preview-based model predictive controller that estimates cardiac and respiratory surface motion from delayed measurements, predicts it over a short horizon, and cancels persistent contact force bias with an offset-free disturbance observer. In simulation the controller reaches 12.0 µm free-space and 1.9 µm contact RMS relative-placement error, versus 18.3/176.8 µm for delayed-feedback impedance and 286.1/275.5 µm for lab-frame PD; in three degrees of freedom it reduces lateral shear from 1.34 to 0.50 mm/s while keeping the QP feasible under sensing noise. The price is higher peak contact force (3.43 versus 2.00 mN), and the paper is explicit that the results are simulation-only with a rigid contact-point tip.","feed_headline":"Tissue-preview control places neural threads within 2 µm","feed_subtitle":"Instead of aiming at a fixed point, the controller rides the moving surface, cutting contact error and lateral shear in simulation.","key_machinery":"The central object is the physiological-motion exosystem, a bank of marginally stable oscillators (a constant mode plus sinusoids at the cardiac and respiratory frequencies and harmonics) whose output is the tissue-surface displacement. Once its state is estimated from delayed surface measurements and forward-propagated by the sensing latency, the entire near-future surface trajectory is available for preview at the cost of one linear propagation. Around that preview sit three mechanisms: an integrating acceleration-bias state that cancels persistent contact reaction and model error, which makes contact tracking offset-free; a receding-horizon QP that renders an impedance port about the moving equilibrium while enforcing actuator and lateral-relative-velocity limits; and a shared nonnegative slack on the octagonal lateral-velocity constraint that keeps the QP feasible under degraded sensing. A constant state matrix with a parameter-affine input matrix lets the stability certificate be checked at two vertices of the reflected-mass box, giving the reported two-vertex Lyapunov certificate.","core_discovery":"The paper's claim is an engineering control claim: for an insertion tool modeled as a rigid contact point, regulating the tip relative to a previewed, latency-delayed model of the pulsating cortical surface yields micron-level placement in a physics simulation. Concretely, the 1-DOF controller reaches 12.0 µm free-space and 1.9 µm contact RMS relative-placement error, against 18.3/176.8 µm for delayed-feedback impedance and 286.1/275.5 µm for lab-frame PD; in 3 DOF, the coupled soft-octagon controller reduces lateral shear from 1.34 to 0.50 mm/s and, at 10 µm RMS per-axis sensing noise, completes all 10 seeds with no measured 0.80 mm/s budget violation while a cost-only controller violates in 10/10 seeds. The paper does not claim a clinical efficacy result; it explicitly leaves flexible-thread mechanics, validated force limits, and hardware sensing as required future work.","pith_inferences":["Beyond the paper, the same preview-plus-offset-free architecture is a candidate for other periodic physiological targets, such as retinal, cardiac, or respiratory motion, whenever the motion can be modelled as a few sinusoids and the sensing latency is known.","A natural next test is to replace the exosystem-generated plant with a recorded, non-harmonic cortical motion trace; the paper's own 40 µm RMS plus ectopic-beat case suggests this is where the controller's margin would erode most.","The paper's measured gap between the quasi-static force cap and realized peak force indicates that hardware force sensing and a dynamic contact model are prerequisites for any force-safety guarantee, not optional refinements."],"forward_implications":["If the central claim is correct, contact-phase depth error during thread placement can be held near two microns without a force sensor, because the offset-free observer cancels the persistent contact reaction.","The coupled lateral-velocity constraint would keep shear near a 0.80 mm/s budget at per-axis sensing noise up to roughly 12.5 µm RMS, whereas cost-only control violates the budget in every seed at 10 µm.","The two-vertex certificate implies the constraint-inactive feedback stays stable over a reflected-mass range of -40% to +50%, but a -50% mass error pushes the running QP infeasible, so the robustness margin is narrower than a naive symmetric box.","Deployment would need to resolve the force trade-off: the benchmarked controller produces 3.43 mN peak contact force versus 2.00 mN for the under-penetrating feedback baseline, because it drives the tip to full commanded depth."],"supporting_citations":[{"why":"Supplies the predictive physiological-motion-compensation baseline and the 15 ms sensing-latency figure.","marker":"[1]"},{"why":"Supplies the internal-model principle that makes periodic surface displacement previewable from an exosystem.","marker":"[3]"},{"why":"Provides the adaptive sinusoidal-rejection foundation for tracking unknown or drifting cardiac and respiratory frequencies.","marker":"[4]"},{"why":"Underpins the offset-free MPC disturbance model used to cancel the persistent contact reaction.","marker":"[5]"},{"why":"Supplies the extended-state-observer idea behind the integrating acceleration-bias state.","marker":"[6]"},{"why":"The QP solver whose warm-start performance the reported solve times depend on.","marker":"[10]"},{"why":"Establishes the constant-state, affine-input two-vertex Lyapunov certificate that the paper specializes to the moving preview equilibrium.","marker":"[11]"},{"why":"Provides the only direct cortical-motion measurements, used to ground and to flag the conservatism of the benchmark amplitudes.","marker":"[27]"}],"fun_headline_variants":["Preview control rides pulsating tissue to 2 µm accuracy in simulation","Predicting tissue motion shrinks neural-thread placement error to 2 µm","Pulsating-tissue preview control achieves micron-level thread placement","Neural-thread tool previews tissue motion, hits 2 µm in sim","Preview-based control keeps neural-thread tip on target in pulsatile tissue"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the cortical surface moves as a steady offset plus a few smooth up-and-down waves at heartbeat and breathing frequencies, and the simulated test tissue is made from exactly that same recipe.","fun_headline_variants_meta":{"raw":{"variants":["Preview control rides pulsating tissue to 2 µm accuracy in simulation","Predicting tissue motion shrinks neural-thread placement error to 2 µm","Pulsating-tissue preview control achieves micron-level thread placement","Neural-thread tool previews tissue motion, hits 2 µm in sim","Preview-based control keeps neural-thread tip on target in pulsatile tissue"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000685,"raw_usage":{"total_tokens":3211,"prompt_tokens":1151,"completion_tokens":2060,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":767,"completion_tokens_details":{"reasoning_tokens":1964}},"tokens_in":767,"tokens_out":2060,"duration_ms":17482,"temperature":1.0,"reasoning_tokens":1964,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:22:57.737618+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the controller against a recorded cortical-motion trace with broadband noise and occasional ectopic beats in a physical flexible-thread setup; if contact-phase RMS placement error grows far beyond a few microns or the 0.80 mm/s shear budget is violated at 10 µm RMS per-axis sensing noise, the central claim fails. A simpler computational check is to replant the simulation with the paper's own non-harmonic 40 µm RMS plus ectopic disturbance and test whether the soft-octagon controller still completes every seed without a shear violation; the paper's own results indicate it would not.","supporting_citations":[{"cited_title":"Active filtering of physiological motion in robotized surgery using predictive control,","cited_arxiv_id":null,"evidence_quote":"Supplies the predictive physiological-motion-compensation baseline and the 15 ms sensing-latency figure."},{"cited_title":"Adaptive algorithms for the rejection of sinusoidal disturbances with unknown frequency,","cited_arxiv_id":null,"evidence_quote":"Provides the adaptive sinusoidal-rejection foundation for tracking unknown or drifting cardiac and respiratory frequencies."},{"cited_title":"Disturbance models for offset-free model-predictive control,","cited_arxiv_id":null,"evidence_quote":"Underpins the offset-free MPC disturbance model used to cancel the persistent contact reaction."},{"cited_title":"From PID to active disturbance rejection control,","cited_arxiv_id":null,"evidence_quote":"Supplies the extended-state-observer idea behind the integrating acceleration-bias state."},{"cited_title":"OSQP: An operator splitting solver for quadratic programs,","cited_arxiv_id":null,"evidence_quote":"The QP solver whose warm-start performance the reported solve times depend on."},{"cited_title":"ARViS: A bleed-free multi-site automated injection robot for accurate, fast, and dense delivery of virus to mouse and marmoset cerebral cortex,","cited_arxiv_id":null,"evidence_quote":"Provides the only direct cortical-motion measurements, used to ground and to flag the conservatism of the benchmark amplitudes."}],"review_version":1}