{"id":"b27dd3b3-4a24-4f5a-a61a-7be39591f3f9","arxiv_id":"2505.06427","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A single parameter in a cardiac model switches between chaotic and organized spiral-wave states, and the resulting intermittency increases mean termination time by up to three orders of magnitude.","lead":"Simulations of a heart-tissue model show that adjusting a single recovery-time parameter creates an intermittent mixture of chaotic and orderly electrical activity, and this mixture greatly lengthens the time before the activity dies out. The result offers a concrete mechanism for why some atrial fibrillation episodes alternate with normal rhythm and may self-terminate.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The intermittency and the three-orders-of-magnitude increase in mean termination time are only demonstrated on a single N=180 domain; no domain-size check separates parameter-induced metastability from a finite-size artifact.","rationale":"The reader's conditional verdict hinges on the possibility that the observed intermittency is a finite-size artifact of the N=180 domain. This is the most load-bearing concern because it attacks the mechanism itself, not merely the error bars: if global non-SDC intervals only exist because the simulation box is small, then the paper's claim that varying tau_w^- can greatly increase Tterm 'similar to changing domain area' is not established outside that specific box, and the clinical extrapolation in the Concluding Remarks loses its main support. The manuscript's own introduction highlights that Tterm grows exponentially with domain area, so a single box size cannot separate a parameter-induced metastable regime from a small-box collapse of SDC to a few spirals. The statistical issue (n=7-8 at tau=215-300 ms and no error bars in Fig. 5e) is real and should also be addressed, but it affects confidence intervals rather than the existence of the regime; an N-sweep with more runs would provide both. Since the reader's CONDITIONAL verdict already requires a domain-size check, this stress-test does not change the verdict.","tokens_in":8735,"tokens_out":8846,"duration_ms":96728,"concrete_test":"Repeat the Section II.B protocol at N=240 and N=288 for tau_w^- = 150, 185, 200, and 215 ms, with at least 20 independent runs per condition, keeping dx, dt, and all other parameters fixed. Compute the intermittency parameter f (Eq. 8) and Tterm (Fig. 5e). If f drops toward zero and/or the Tterm ratio between 150 and 200 ms shrinks by more than an order of magnitude as N grows, the reported intermittency is a finite-size artifact. If both persist at N=288, the finite-size objection is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that intermittent non-SDC intervals are an intrinsic dynamical regime that sharply increases the mean termination time. The only evidence comes from an N=180 box with no-flux boundaries (Section II.B), and the paper never varies N. This is load-bearing because the non-SDC state consists of a small, fixed number of spiral waves in the entire box; in a larger domain, SDC would contain many more tips, and the probability of a global fluctuation down to a few stable tips is expected to be exponentially small in area, consistent with the domain-size dependence cited in the Introduction (Refs. 28-30 and [34]). Thus the intermittent intervals, and the resulting Tterm growth, could be a finite-size metastability specific to this box rather than a generic property of SDC. The coexistence example in Fig. 7 is local and does not test the global non-SDC intervals used to define f and to compute Tterm. A parameter sweep at fixed N=180 cannot distinguish 'the system develops an intermittent regime' from 'the box is small enough that the chaotic state intermittently collapses to a few spirals'.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the Fenton-Karma (FK) cardiac model in a 2D domain and reports that varying a single time constant, τ−w, takes the system from continuous spiral defect chaos (SDC) through a regime of intermittent SDC, in which intervals of chaos are punctuated by quasi-stable intervals with a small, constant number of spiral waves, to a final state with a single stable spiral wave. The authors define an intermittency parameter f, measure the probability distribution of spiral tip numbers, and compute the mean termination time Tterm. They find that as τ−w approaches the stable-spiral regime, Tterm increases sharply, by almost three orders of magnitude between τ−w = 150 ms and 200 ms, and that the number of intermittent non-SDC intervals per simulation correlates positively with termination time. They also report local coexistence of quasi-stable spiral waves with SDC and suggest clinical relevance for paroxysmal vs. persistent atrial fibrillation.","tokens_in":8933,"tokens_out":5659,"duration_ms":58720,"significance":"If the central claim survives scrutiny, this is a valuable contribution to the dynamics of excitable media and cardiac arrhythmia modeling: it identifies a single-parameter route from short-lived SDC to much longer-lived intermittent activity, and it provides quantitative observables (the fraction f of time in the non-SDC state, tip-number distributions, and lifetime distributions) that could be compared with clinical data. The paper also reproduces and extends earlier master-equation results by relating termination-time distributions to the probability of small tip numbers. The authors are explicit about the computational limitations, and the qualitative phenomenology is well illustrated by the time series and snapshots. However, the quantitative claim of a three-orders-of-magnitude increase in Tterm currently rests on a single system size and on a small number of simulations for the most interesting parameter values, which limits the strength of the conclusions.","major_comments":[{"comment":"The central quantitative claim is not yet supported against a finite-size artifact. All simulations use a single N=180 domain with no-flux boundaries. The non-SDC intervals consist of a small, fixed number of spiral waves in the entire box. Since previous work, cited in the Introduction (Refs. 28-30), shows that Tterm grows exponentially with domain size, the observed collapse from many-tip SDC to a few quasi-stable tips could be a metastable, finite-size effect that would disappear or become negligible in larger domains. The authors should vary N (e.g., 120, 180, 240, 360) and show how f, the tip-number distribution, and Tterm scale with system size, or provide a theoretical argument that the fluctuation to small tip numbers is not controlled by area. Without such a test, the extrapolation to clinical atrial fibrillation in the Concluding Remarks is not justified.","section":"II.B, Fig. 5e"},{"comment":"The quantitative claim of a sharp, three-orders-of-magnitude increase in Tterm between τ−w = 150 ms and 200 ms is not accompanied by error bars for Tterm, and several of the data points are based on very few simulations: only 7 simulations for τ−w = 215 ms and 250 ms, and 8 for τ−w = 300 ms. Because the intermittent-regime termination-time distributions are heavy-tailed (Fig. 5b-d), the sample mean is highly sensitive to rare long-lived realizations and the uncertainty is large. The authors should provide bootstrap confidence intervals or standard errors for every point in Fig. 5e, report the number of simulations for each point, and discuss whether any simulations were terminated by a maximum-time cutoff (censoring), which would bias the mean. The qualitative trend may be robust, but the specific magnitude and the transition point need statistical support.","section":"Fig. 5e, Section III.C"},{"comment":"The definition of a non-SDC interval is arbitrary: any interval longer than 1.5 s (≥19 rotations) in which the number of spiral tips remains stable. The intermittency parameter f and, consequently, the location of regime boundaries in Fig. 3 depend on this threshold and on the precise meaning of \"stable\" (is zero change in tip number required, or is a small tolerance allowed?). The authors should test the sensitivity of f to the threshold (e.g., 1.0, 2.0, 3.0 s) and state the tip-counting tolerance. Without this, the quantitative transition curve f(τ−w) is not well defined.","section":"Section III.A, Eq. (8)"},{"comment":"The quantity TSDC, the average time spent in the SDC state before termination, is not precisely defined. It is unclear whether TSDC is the sum of all SDC intervals in a simulation, or the average length of individual SDC intervals, and how the coexistence case of Fig. 7, in which quasi-stable spirals appear locally while the rest of the domain is in SDC, is treated. The exponential fit to TSDC (blue line in Fig. 5e) is also not described: which points are included, what is the fitting form, and what is the goodness of fit? These details are needed to support the claim that the growth in Tterm is not simply due to longer non-SDC intervals.","section":"Section III.C, Fig. 5e"}],"minor_comments":[{"comment":"There are several typographical issues: \"Fig. 1(a))\" has an extra parenthesis; \"liftetimes\" should be \"lifetimes\"; the superscripted \"a\" in \"For this, we compute... a As shown in Fig. 6a&b\" appears to be a LaTeX artifact; and the notation τ−w would read more cleanly as τ_w^- .","section":"Throughout"},{"comment":"The caption uses τSDC while the text uses TSDC for the same quantity; please unify the notation.","section":"Fig. 5e caption"},{"comment":"The authors use the term \"quasi-stable\" for spiral waves in non-SDC intervals but do not define it precisely; a short definition (e.g., a tip that survives for at least some multiple of the dominant period and does not change its distance to the boundary) would improve clarity.","section":"Section III.A, Fig. 2"},{"comment":"The statement that the intermittent behavior is not restricted to the particular parameter, parameter set, or model is not supported by any displayed data, figure, or reference to a supplementary file. Either add a supporting figure or weaken the claim to a qualitative observation.","section":"Concluding Remarks"},{"comment":"The range for τ−w is given as \"50-300 ms\" with a hyphen; using an en dash and explicitly stating the units for all parameters (e.g., usi_c is dimensionless) would be clearer.","section":"Table I"}],"recommendation":"major_revision","confidential_remarks":"The paper's core phenomenological observation is interesting and likely correct, but the main quantitative claim is under-supported by the absence of a finite-size analysis and by the lack of error bars on the termination-time data. The authors should be encouraged to add a domain-size study and more careful statistics; these are feasible within the scope of the manuscript. I also note the heavy use of self-citations (Refs. 12, 13, 27, 34) to the authors' own group; this is not inappropriate here, but the generality claims in the Concluding Remarks would be more convincing with independent validation or a supplementary figure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line up front: this is a genuine new observation—intermittent non-SDC intervals in the FK model, with a sharp increase in mean termination time as τ−w approaches the single-spiral regime. The paper deserves a serious referee, but the main quantitative claim is under-supported by one missing control: they never vary the domain size.\n\nWhat's actually new: previous work on SDC termination used a master-equation picture where Tterm is controlled by p(NPS=1). Here they show a parameter regime where the system alternates between SDC and periods with a small fixed number of stable tips, and they quantify this with an intermittency parameter f. The time series, tip distributions, and lifetime distributions in Fig. 6c are convincing that something qualitatively different happens around τ−w=200 ms. The correlations between number of non-SDC intervals and Tterm (Fig. 6a,b) are a nice direct check. The bootstrap error bars on f are a plus.\n\nWhere it's soft: The three-order-of-magnitude jump in Tterm in Fig. 5e has no error bars on the mean, and the largest values (τ−w=215, 250, 300 ms) rest on 7-8 simulations. Those are expensive runs, but a few more and a bootstrap CI would help. Bigger issue: everything is on one N=180 box. The non-SDC state is a few stable tips sitting in the whole box. In a larger domain, SDC would have many more tips, and the chance of fluctuating down to a few tips should be exponentially smaller; that's exactly the domain-size dependence cited in their own intro. So it's plausible the intermittent regime is metastability of a small box, not a generic property of the model. They should test at least one larger N at τ−w=200 and show the regime persists. The coexistence example in Fig. 7 is local and doesn't address the global intermittency used to define f.\n\nThe clinical extrapolation in the conclusion is appropriately hedged and doesn't bother me. The self-citations are to the group's own master-equation work and are legitimate foundations.\n\nIf I were the referee: ask for error bars on Fig. 5e, more runs at the critical parameters, and a domain-size check. The core observation will likely survive, but the claim as stated—that tuning recovery dynamics can convert short-lived chaos into long-lived intermittency—is only proven for one box size. That's a normal, fixable gap.\n\nRecommendation: send it out. A good referee will catch the domain-size issue, and the paper has enough solid content to be worth their time.","headline":"Genuine new observation of intermittent non-SDC states in the FK model with a sharp Tterm increase, but one missing domain-size control keeps the central claim from being fully settled.","tokens_in":9474,"tokens_out":1989,"would_cite":true,"duration_ms":19328,"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":"Lengthening one recovery-time parameter converts cardiac spiral chaos into long-lived intermittent activity, raising mean termination time by nearly three orders of magnitude.","keywords":["cardiac arrhythmias","spiral defect chaos","intermittency","termination time","Fenton-Karma model","excitable media","phase singularity","atrial fibrillation"],"falsifier":"Repeat the same protocol at $\\tau_w^-=200$ ms in larger domains, for example $N=360$ and $N=720$, keeping all other parameters fixed. If the intermittency parameter $f$ drops toward zero or the sharp increase in $T_{\\rm term}$ moves to different parameter values, the reported effect is a finite-size artifact; if the non-SDC intervals and the increase persist, the claim stands.","tokens_in":8511,"feed_emoji":"❤️","tokens_out":7563,"duration_ms":69758,"temperature":0.7,"pith_summary":"This paper argues that spiral defect chaos, the disordered self-renewing wave state thought to underlie atrial fibrillation, is not inevitably a short-lived state in cardiac tissue models. Increasing the recovery time constant $\\tau_w^-$ in the Fenton-Karma model carries the system from continuous chaos through an intermittent regime in which chaotic intervals alternate with intervals holding a small, stable number of spiral waves. The authors quantify this intermittency and show that the mean time until all electrical activity stops rises sharply, by almost three orders of magnitude between $\\tau_w^-=150$ ms and $200$ ms. The result matters because it identifies a route, other than tissue size, by which arrhythmic activity can become long-lived.","feed_headline":"A single parameter stretches arrhythmia survival 1,000-fold","feed_subtitle":"In a cardiac model, longer recovery time turns brief spiral chaos into long-lived intermittent waves.","key_machinery":"The paper's central quantitative tool is the intermittency parameter $f$, defined by averaging, over successive non-SDC/SDC interval pairs, the fraction $f_i=t_{\\text{non-SDC},i}/(t_{\\text{non-SDC},i}+t_{\\text{SDC},i})$ of each pair spent in the non-SDC state; $f=0$ is continuous SDC and $f=1$ is a single stable spiral. A non-SDC interval is operationally defined as an interval longer than 1.5 s (at least 19 rotations) in which the number of spiral wave tips stays constant. The companion quantity is the probability distribution $p(N_{\\rm PS})$ of the number of phase singularities: in continuous SDC the mean termination time is known to scale as $1/p(N_{\\rm PS}=1)$, and the intermittent regime introduces a bump at small $N_{\\rm PS}$ that eventually dominates the distribution. These measures are what link the visually distinct dynamical regimes to the reported thousand-fold increase in survival time.","core_discovery":"Varying the single recovery-time parameter $\\tau_w^-$ moves the Fenton-Karma model from continuous spiral defect chaos (SDC) to an intermittent regime and finally, for $\\tau_w^-\\ge 330$ ms, to a single stable spiral wave. In the intermittent regime, SDC intervals are separated by non-SDC intervals during which the number of phase singularities remains small and constant for many rotations; the intermittency parameter $f$, the average fraction of time spent in the non-SDC state, grows from 0 toward 1 as $\\tau_w^-$ increases. The mean termination time $T_{\\rm term}$ grows linearly with $\\tau_w^-$ in the continuous SDC regime and then sharply, by nearly three orders of magnitude, between $\\tau_w^-=150$ ms and $200$ ms. Simulations with more non-SDC intervals terminate later, and individual spiral wave lifetimes grow to more than $10^3$ s. The authors also report quasi-stable spiral waves coexisting with SDC in different parts of the same domain, increasingly for $\\tau_w^-\\ge 200$ ms, and state that similar intermittency occurs for other parameter choices and in a generic two-variable model.","pith_inferences":["Editorial inference: the sharp rise in $T_{\\rm term}$ resembles a critical slowing down at the SDC-to-spiral transition; one testable consequence is that non-SDC interval durations near the transition should obey a power-law distribution, a quantity the paper does not report.","Editorial inference: the coexistence of quasi-stable spirals with SDC in separate subregions suggests that ordered regions act as local stabilizers; perturbing recovery parameters only in a subdomain could create or destroy such coexistence in a controlled way.","Editorial inference: because survival times are known to depend exponentially on domain size, the reported thousand-fold increase might be amplified or suppressed by system size; comparing $f$ across domain sizes would separate the intrinsic intermittent regime from a finite-size effect.","Editorial inference: clinically, if non-SDC intervals correspond to organized activity, the model predicts a counterintuitive hazard: drugs that flatten the action-potential-duration restitution curve (effectively raising $\\tau_w^-$) could lengthen fibrillatory episodes even as they make the waves more orderly."],"forward_implications":["In the continuous SDC regime the mean termination time increases linearly with $\\tau_w^-$, but entering the intermittent regime changes the growth to a sharp, nearly exponential rise.","Termination-time probability distributions switch from exponential in continuous SDC to long-tailed in the intermittent regime, reflecting runs with one or several long non-SDC intervals.","Recovery time acts as a control axis alongside domain size: shorter recovery gives short-lived chaos, longer recovery near the single-spiral boundary gives activity that can outlive the simulation budget.","The same intermittency appearing under other parameter choices and in a generic two-variable model indicates that the mechanism is not locked to one parameter set of the Fenton-Karma model.","If the mapping to atrial fibrillation holds, the results imply that interventions altering recovery dynamics could change whether fibrillation self-terminates or persists."],"supporting_citations":[{"why":"Supplies the Fenton-Karma ion-current model and parameter set used for all simulations.","marker":"[31]"},{"why":"Defines the tip instability and far-field breakup mechanisms that distinguish SDC from the ordered regime.","marker":"[23]"},{"why":"Establishes the baseline result that continuous-SDC termination time scales with $1/p(N_{\\rm PS}=1)$, the relation the intermittent regime modifies.","marker":"[34]"},{"why":"Provides clinical and statistical evidence that fibrillation termination times are exponentially distributed, the baseline the long-tailed intermittent distribution departs from.","marker":"[27]"},{"why":"Supplies the phase-singularity tracking algorithm used to count spiral tips and measure their lifetimes.","marker":"[32]"},{"why":"Supports the claim that intermittency is not special to this parameter set by showing it in a generic two-variable model.","marker":"[35]"},{"why":"Supplies the action-potential-duration restitution mechanism invoked to explain the near-tip spiral instability.","marker":"[33]"}],"fun_headline_variants":["Intermittent chaos lengthens spiral wave life by 1000x","One parameter stretches spiral chaos lifetime 1000-fold","Intermittent spirals delay fibrillation termination by 1000x","Tuning recovery time delays cardiac spiral death 1000-fold"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The long-lived orderly intervals are treated as a genuine dynamical state, but every simulation uses the same $N=180$ tissue size, and survival times in these models are known to grow with domain size, so the intervals and the thousand-fold increase could be artifacts of the fixed box rather than a property of the model.","fun_headline_variants_meta":{"raw":{"variants":["Intermittent chaos lengthens spiral wave life by 1000x","One parameter stretches spiral chaos lifetime 1000-fold","Intermittent spirals delay fibrillation termination by 1000x","Tuning recovery time delays cardiac spiral death 1000-fold"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000329,"raw_usage":{"total_tokens":1858,"prompt_tokens":991,"completion_tokens":867,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":607,"completion_tokens_details":{"reasoning_tokens":793}},"tokens_in":607,"tokens_out":867,"duration_ms":7159,"temperature":1.0,"reasoning_tokens":793,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:42:39.860710+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the same protocol at $\\tau_w^-=200$ ms in larger domains, for example $N=360$ and $N=720$, keeping all other parameters fixed. If the intermittency parameter $f$ drops toward zero or the sharp increase in $T_{\\rm term}$ moves to different parameter values, the reported effect is a finite-size artifact; if the non-SDC intervals and the increase persist, the claim stands.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the tip instability and far-field breakup mechanisms that distinguish SDC from the ordered regime."},{"cited_title":"Vidmar and W.-J","cited_arxiv_id":null,"evidence_quote":"Establishes the baseline result that continuous-SDC termination time scales with $1/p(N_{\\rm PS}=1)$, the relation the intermittent regime modifies."},{"cited_title":"Dharmaprani, M","cited_arxiv_id":null,"evidence_quote":"Provides clinical and statistical evidence that fibrillation termination times are exponentially distributed, the baseline the long-tailed intermittent distribution departs from."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the phase-singularity tracking algorithm used to count spiral tips and measure their lifetimes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supports the claim that intermittency is not special to this parameter set by showing it in a generic two-variable model."},{"cited_title":"Nolasco and R","cited_arxiv_id":null,"evidence_quote":"Supplies the action-potential-duration restitution mechanism invoked to explain the near-tip spiral instability."}],"review_version":1}