Pith. sign in

REVIEW 3 major objections 7 minor 20 references

Applying the Method of Critical Fluctuations on Human Electrocardiograms

T0 review · 3 major / 7 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A physics-based test of ECG criticality flags 100% of infarction records and 88% of healthy controls.

desk verdict The 35-record extension is real, but the diagnostic conclusion outruns the method's validation. read the letter →

arxiv 1908.06408 v1 pith:FUUEKF3P submitted 2019-08-18 physics.med-ph physics.bio-ph

Yiannis Contoyiannis , Fotis Diakonos , Myron Kampitakis This is my paper · ORCID
classification physics.med-phphysics.bio-ph
keywords electrocardiogramMethodofCriticalFluctuationscriticalitypower-lawlaminarlengthsmyocardialinfarctionhealthycontrolautocorrelationfunction
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper applies the Method of Critical Fluctuations (MCF) to human electrocardiograms to test whether cardiac health shows up as physical criticality in the ECG signal. MCF extracts the distribution of waiting times (laminar lengths) between voltage levels in the high-frequency 'grass' fluctuations, and healthy critical dynamics are identified by a power-law distribution with exponent between 1 and 2. Applied to 35 ECG records from a public diagnostic database, the method finds critical dynamics in 22 of 25 healthy-control ECGs (88%) and in none of the 10 myocardial-infarction records, which all sit far from criticality. The authors conclude that roughly one in ten ECGs labeled healthy may actually be pathological, and they propose criticality as a physical criterion that can reveal healthy and diseased states beyond ordinary ECG reading.

What carries the argument

The load-bearing object is the Method of Critical Fluctuations (MCF): a procedure that detects critical dynamics in an experimentally recorded time series by locating a fixed point $V_L$ where high-frequency 'grass' fluctuations begin, scanning an upper level $V_U$, and computing the distribution $P(L)$ of laminar lengths $L$ spent between $V_L$ and $V_U$. Criticality is declared when the fitted exponent $p_2$ of the power-law part of $P(L)$ satisfies $1<p_2<2$ and the exponential exponent $p_3$ is close to zero; departure from criticality shows as $p_2$ falling and $p_3$ growing. A wider zone $\Delta V_U$ over which these conditions hold is read as a more stable critical state. The autocorrelation function of the most critical record is used as a second, related measure.

What would settle it

Run MCF blinded on a larger ECG cohort with an independent gold-standard diagnosis; the claim fails if a confirmed myocardial-infarction record passes the criticality condition ($1<p_2<2$, $p_3\approx 0$) or if a cohort of clinically verified healthy subjects systematically fails it. A direct check is to submit the three flagged control records (p165, p245, p242) to independent cardiological assessment to see whether they are in fact diseased.

Watch

Extended reading notes

Core claim

The central claim is that a healthy heart's electrical fluctuations are in a critical state, and this criticality is measurable in an ECG. For a segment of high-frequency fluctuations, the authors define laminar lengths as waiting times inside the zone between the fixed point $V_L$ and a varied level $V_U$; when the laminar-length distribution follows a power law with $p_2$ in (1,2) and $p_3\approx 0$ over a range of $V_U$, the signal is critical. By this criterion, all 10 myocardial-infarction records are non-critical while 22 of 25 healthy-control records are critical, giving 100% and 88% agreement. The paper therefore claims that MCF can serve as a diagnostic indicator and that the three non-critical control records may be misdiagnosed. For the record with the strongest criticality (p121), the autocorrelation function shows characteristic symmetries that the authors propose as a measure of optimal heart functionality.

Load-bearing premise

The load-bearing premise is that a healthy heart is in a physical critical state, so that detecting critical fluctuations in an ECG is treated as proof of health; the paper inherits this equation from frog-heart data and never independently verifies it in humans.

Editorial extensions

If this is right

  • If criticality is genuinely a marker of healthy cardiac tissue, MCF offers a complementary screening signal that does not rely on matching ECG morphology to a template.
  • The reported 12% non-critical healthy controls become a concrete claim: about one in ten patients with a normal-looking ECG may be misdiagnosed and could warrant further clinical investigation.
  • The gradation of $p_3$ and the width of the critical zone $\Delta V_U$ provide a quantitative scale of cardiac health, from strongly critical (record p121) to pathology-like, rather than a binary label.
  • For myocardial infarction, the perfect 10/10 separation suggests the method could be tested as an aid in cases where ordinary ECG reading is ambiguous.
  • Because the paper explicitly stops short of medical physiology, a validated link between criticality and cardiac pathophysiology would need to be established before clinical use.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Going beyond the paper: the three healthy-labeled but non-critical records (p165, p245, p242) are the natural test cases—if independent cardiological assessment finds them diseased, MCF is a genuinely prospective diagnostic; if it finds them healthy, the 'criticality equals health' premise is incomplete.
  • Going beyond the paper: the same MCF procedure could be applied to other labeled ECG classes such as arrhythmia, ischemia, or cardiomyopathy to see whether non-criticality is specific to infarction or common to all pathology, which would sharpen or weaken the diagnostic claim.
  • Going beyond the paper: the autocorrelation symmetries reported for record p121 could be computed automatically for the other records; if the three non-critical controls lack those symmetries, the autocorrelation profile might replace manual $V_U$ scanning as a faster screening statistic.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 7 minor

Summary. The manuscript applies the Method of Critical Fluctuations (MCF) to human ECG time-series from the PTB Diagnostic ECG Database. For each ECG, the authors manually select stationary segments, identify a fixed point V_L and scan an upper voltage V_U, then fit the laminar-length distribution with a power-law-with-exponential form to estimate exponents p2 and p3. They report that 22 of 25 healthy-control records (88%) satisfy their criticality condition (1 < p2 < 2, p3 approximately 0), while all 10 myocardial-infarction records are non-critical (100%). They interpret the three discordant healthy-control records as possible misdiagnoses and also present the autocorrelation function of the record p.121 as a candidate signature of optimal cardiac function.

Significance. If the equivalence between physical criticality and cardiac health were established on human data, the approach could offer a non-invasive, physiology-based screening marker, and the use of a public database with clearly reported agreement rates is a strength. The descriptive rates (88% and 100%) are a useful pilot observation. However, the central diagnostic conclusion depends on a criticality-health link imported from earlier frog-heart work, and the manual segmentation and threshold choices are not yet specified well enough for independent reproduction. As it stands, the paper is a promising but preliminary demonstration rather than a validated diagnostic claim.

major comments (3)
  1. [Introduction and Conclusions] The conclusion that 'in approximately 1 out of 10 ECGs which presented the typical characteristics of a healthy ECG, the diagnosis may not be accurate' rests on the premise, stated in the Introduction, that 'When heart tissue is functioning properly (healthy state), then it is in a physical critical state.' This equivalence is imported from the earlier frog-heart study [10] and is never independently validated on human data in this manuscript. The paper itself concedes in the Conclusions that 'this work interprets the heart operation in terms of Physical Critical Phenomena and is not involved with Medical Physiology.' Consequently, the three discordant healthy-control records (p165, p245, p242) are as plausibly MCF false negatives as they are misdiagnoses; the central diagnostic claim needs either an external validation of the criticality-health link or a clearly weakened formulation.
  2. [MCF application steps (bulleted list) and Fig.4] The classification protocol has substantial manual freedom. The fixed point V_L is 'localized' visually, the upper boundary V_U is moved to a new position without a pre-specified rule, stationary segments are chosen by eye using 'the criterion of stationarity,' and the criticality condition 'p3 approximately 0' is never given a numerical threshold: in Fig.4a, p3=0.006 is treated as critical, while p3=0.29 is treated as non-critical, but no cutoff is defined. These degrees of freedom can inflate the reported 88% and 100% agreement rates and make independent reproduction difficult. Please specify an algorithmic protocol, including a p3 cutoff, a stopping rule for the V_U scan, and a stability criterion for segment selection, and apply it blinded to the diagnostic labels.
  3. [Results, 'We have analyzed 25 ECGs...' and 'We have analyzed 10 ECGs...'] The headline rates are reported without statistical quantification. With 10 infarction cases, the 100% agreement rate has an exact 95% confidence interval of approximately [69%, 100%], so the claim of 'absolute agreement' is compatible with a wide range of true sensitivity. Please report confidence intervals or exact binomial tests for both rates, and ideally evaluate the procedure on a held-out or independent set rather than on the same 35 records used to illustrate the method.
minor comments (7)
  1. [Equation (1) and following sentence] The text refers to 'Eq. (11)' where it should refer to Eq. (1).
  2. [Reference [16]] Reference [16], cited for the first MCF application to human ECG, is titled 'The Earth as a living planet: Human-type diseases in the earthquake preparation process'; please clarify whether this is the correct source or provide the correct prior human-ECG reference.
  3. [Figures 2 and 3] The time-axis labels are inconsistent between Fig.2 ('t(1/1K s)') and Fig.3 ('t(10^{-3} s)'), which will confuse readers about the sampling scale.
  4. [Abstract] The phrase 'In contrary' should be changed to 'In contrast'.
  5. [Reference [5]] The title of reference [5] contains 'mFractal'; this appears to be a typo for 'Fractal'.
  6. [Fig.7 and autocorrelation discussion] The autocorrelation analysis is presented without a quantitative criterion linking the observed 'symmetries' to criticality; please mark it as illustrative or add a quantitative measure.
  7. [Criticality condition, Eq. (3)] The condition 'p3 approximately 0' should be stated as a numerical criterion (for example, p3 below a specified tolerance) to avoid ambiguity in reproducing the method.

Circularity Check

2 steps flagged · score 6.0 of 10

The 88%/100% agreement rates are empirical, but the headline inference that 12% of healthy-labeled ECGs are misdiagnoses is just the MCF criticality criterion restated as a medical conclusion.

  1. self definitional [Abstract and Conclusions]
    "Using the concept of criticality as basic criterion for the characterization of the recorded ECG as that of a healthy person ... Basic conclusion of our work is that in approximately 1 out of 10 ECGs which presented the typical characteristics of a healthy ECG, the diagnosis may not be accurate."

    The paper uses the MCF criticality conditions (1<p2<2, p3 approximately 0) as the criterion for calling an ECG healthy. The three PTB-labeled healthy controls that fail these conditions are therefore classified as non-healthy by construction, and the conclusion that their PTB diagnosis 'may not be accurate' is simply the same criterion re-expressed as a medical inference. No independent medical evidence is supplied that a non-critical MCF result corresponds to infarction or any pathology; the disagreement with external labels is interpreted as a misdiagnosis only because the criticality criterion is assumed to define health.

  2. self citation load bearing [Introduction, paragraph beginning 'In 2003 we applied the MCF']
    "In 2003 we applied the MCF in ECGs produced from frog heart [10]. The results of the MCF analysis are summed up in the phrase: When heart tissue is functioning properly (healthy state), then it is in a physical critical state."

    The central premise that criticality equals cardiac health is imported from the authors' own prior frog-heart paper (ref. [10]) and is not independently validated on human ECGs in this manuscript. This self-citation is load-bearing because it is the only support for interpreting the three discordant healthy-control records as possible misdiagnoses rather than as MCF false negatives. The paper's closing disclaimer concedes the missing link: 'this work interprets the heart operation in terms of Physical Critical Phenomena and is not involved with Medical Physiology. However, a link between the two descriptions is necessary and is the subject of further work.' Thus the headline diagnostic claim rests on an unverified premise supplied by a self-citation chain.

full rationale

The reported agreement rates are not themselves circular: the PTB diagnostic labels are external, and the MCF classifications are compared with those labels to obtain 88% and 100%. That part of the paper is a descriptive, externally anchored result. The circularity lies in the interpretive leap from 'these three healthy-labeled ECGs do not satisfy MCF criticality' to 'those diagnoses may be inaccurate.' That leap is equivalent to the paper's own criterion, since the criterion is explicitly used as the characterization of a healthy ECG. The premise that criticality indicates healthy cardiac tissue is also carried entirely by self-citation to the authors' prior frog-heart work, and the paper itself acknowledges that the link to medical physiology remains to be established. Manual degrees of freedom in selecting stationary segments and scanning V_U are better framed as reproducibility or correctness risks rather than circularity, but they reinforce the concern that the classification procedure is not fully independent of researcher choices. Overall, the central diagnostic conclusion reduces to the assumed criticality=health equivalence, while the raw agreement statistics retain independent content; this is a partial, construction-level circularity rather than a fully vacuous derivation.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on two kinds of imported assumptions: the MCF theoretical machinery (intermittent map, power-law waiting-time distributions) and the physiological interpretation that criticality equals cardiac health. The exponents p2 and p3 are fitted per segment and supply the classification, while V_L/V_U are manual analysis choices. No genuinely new entities, particles, or forces are introduced. The main burden is that the health-criticality link is carried over from frog-heart work and is not validated on human outcomes in this study.

free parameters (3)
  • p2 power-law exponent per ECG segment = Examples: 1.28 for p121, 0.31 for p058, 0.10 for p165
    Fit parameter in P(L) approximately L^(-p2) e^(-p3 L); the criticality classification requires p2 > 1 (with the text also citing 1 < p2 < 2). Values come from fitting each laminar-length histogram.
  • p3 exponential cutoff exponent per ECG segment = Examples: 0.006 for p121, 0.29 for p058, 0.32 for p165
    Second fit parameter in Eq. (3); criticality requires p3 approximately 0, but no explicit numerical threshold is given in the paper.
  • V_L fixed point and V_U upper boundary positions = V_L = -0.05 for p121; V_U varied; not reported for other patients
    Manual choices defining the laminar region. The existence of a zone Delta V_U where 1 < p2 < 2 and p3 approximately 0 is the criticality criterion, so these thresholds directly determine the classification results.
assumptions (4)
  • domain assumption A physical system in a critical state produces laminar-length distributions following P(l) ~ l^(-p_l) with 1 < p_l < 2 and p3 approximately 0; this is the MCF criticality criterion.
    Inherited from the authors' prior works [7,8,9,10]; stated in the methods section and used as the classification rule without independent validation in this paper.
  • domain assumption Criticality in the relaxation-phase grass fluctuations indicates healthy cardiac tissue, while deviation indicates pathology.
    Stated in the Introduction: 'When heart tissue is functioning properly (healthy state), then it is in a physical critical state.' This premise is used to interpret the three healthy-control disagreements as possible misdiagnoses.
  • domain assumption The analyzed stationary ECG segments are representative of the patient's cardiac state, and non-stationary segments can be excluded without bias.
    The paper selects segments by the criterion of stationarity but gives no algorithm or threshold; the manual choice could change the reported agreement rates.
  • domain assumption The intermittent map approximation with a stochastic shift parameter describes critical fluctuations in ECG time series.
    The derivation of Eqs. (2) and (3) relies on the critical map and the distribution of laminar lengths from intermittency theory (refs [13,14,15]); this framework is adopted rather than derived in the present paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Applying the Method of Critical Fluctuations on Human Electrocardiograms." pith.science (2026). https://pith.science/paper/FUUEKF3P

@misc{pith2026190806408,
  author       = {Pith},
  title        = {Pith review of: Applying the Method of Critical Fluctuations on Human Electrocardiograms},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FUUEKF3P}},
  note         = {Machine review of arXiv:1908.06408}
}
read the original abstract

In this work we apply the Method of Critical Fluctuations (MCF)on human Electrocardiogram (ECG) time-series. The method is able to reveal critical characteristics, in terms of physical behavior, in experimentally recorded signals. Using the concept of criticality as basic criterion for the characterization of the recorded ECG as that of a healthy person, we find a 100% verification of the characterization Myocardial infarction. In contrary in the cases of the characterization Healthy control we find a 88% agreement. We also consider the autocorrelation function for the ECG time-series which obeys optimally the criteria of criticality and we observe the appearance of specific characteristic symmetries in the corresponding profile.

Figures

Figures reproduced from arXiv: 1908.06408 by the authors.

Figure 1
Figure 1. Fig.1. A detail from frog’s ECG. The fluctuations i [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Fig.2. In Human ECG [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 6
Figure 6. Fig.6. (a) Detail from ECG p.165 characterized as c [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 20 canonical work pages

  1. [10]

    Criticality in the relaxation phase of the spontan eous contracting atria isolated from the heart of the frog (Rana ridibunda )

    Y. Contoyiannis, F. Diakonos, C. Papaefthimiou and G. Theophilidi s. “Criticality in the relaxation phase of the spontan eous contracting atria isolated from the heart of the frog (Rana ridibunda )” Phys. Rev. Lett. 93, 098101 (2004)

    work page 2004

  2. [1]

    Himanshu Gothwal, Silky Kedawat, Rajesh Kumar “Card iac arrhythmias detection in an ECG beat signal using fast fourier transform and artificial neural network“, J.Biomedical Science and Engineeri ng, 2011, 4, 289-296, doi:10.4236/jbise.2011.44039

    work page arXiv 2011

  3. [2]

    B.V.P Prasad & Velusamy Parthasarathy (2018)“Detec tion and classification of cardiovascular abnormalities using FFT based mul ti-objective genetic algorithm“,Biotechnology & Biotechnological Equipme nt,32:1,183-193,DOI: 10.1080/13102818.2017.1389303

    work page arXiv 2018

  4. [3]

    Naoko Aoyagi , Ken Kiyono , Zbigniew R. Struzik , and Yoshiharu Yamamoto “Changes in the Hurst Exponent of Heart Rate Variab ility during Physical Activity“, AIP Conference Proceedings 780, 599 (200 5); https://doi.org/10.1063/1.2036824

    work page doi:10.1063/1.2036824

  5. [4]

    DOI: 10.1016/j.bspc.2017.09.027

    Alexander Z Tzabazis, Andreas Eisenried, David C Yeomans “Wavelet analysis of heart rate variability: Impact of wavelet select ion “, Biomedical Signal Processing and Control 40:220-225, February 2018 . DOI: 10.1016/j.bspc.2017.09.027

    work page doi:10.1016/j.bspc.2017.09.027 2018

  6. [5]

    Jonathan Senand Darryl McGill “Fractal analysis of heart rate variability as a predictor of mFractal analysis of heart rate variab ility as a predictor of mortality: A systematic review and meta-analysis“, Chaos 28, 072101 (2018); https://doi.org/10.1063/1.5038818

    work page doi:10.1063/1.5038818 2018

  7. [6]

    Bifurcation theory and cardiac arrhythmias

    Hrayr S Karagueuzian, Hayk Stepanyan and William J Mandel “Bifurcation theory and cardiac arrhythmias”, Am J Cardiovasc Dis. 2013; 3(1): 1–16

    work page 2013

  8. [7]

    Intermittent dynamics of critical fluctuations

    Y. Contoyiannis , F. Diakonos and A. Malakis . “Intermittent dynamics of critical fluctuations” Phys.Rev. Lett. 89, 35701(2002)

    work page 2002

Show all 20 references
  1. [8]

    Monitoringof a preseismic phase from its electromagnetic precursors

    Y. Contoyiannis, P. Kapiris and K. Eftaxias . “Monitoringof a preseismic phase from its electromagnetic precursors”. Phys. Rev. E 71, 1(2005)

    2005

  2. [9]

    Kosmidis, Yiannis F

    Efstratios K. Kosmidis, Yiannis F. Contoyiannis , Costas Papatheodoropoulos ,Fotios K. Diakonos “ Traits of criticality in membrane potential fluctuations of pyramidal neurons in the CA1 region of rat hippocam pus “. European Journal of Neuroscience ,https://doi.org/10.1111/ej...

    2018 doi

  3. [11]

    Intermittency in Stock Market Dynamics

    A.Ozun, Y.Contoyiannis, F.Diakonos, L.Magafas and M.Hanias“Intermittency in Stock Market Dynamics”.Journal of Trading. Summe r 2014, Vol 9, No 3, pp26-33

    2014

  4. [12]

    Intermittency-induced criticality in a resis tor-inductor-diode circuit

    Stelios M. Potirakis , Yiannis Contoyiannis , Fotios K. Diakonos, and Michael P. Hanias“Intermittency-induced criticality in a resis tor-inductor-diode circuit ”. Physical Review E 95, 042206 (2017)

    2017

  5. [13]

    Unimodal ma ps and order parameter fluctuations in the critical region

    Y.F. Contoyiannis and F.K. Diakonos . “Unimodal ma ps and order parameter fluctuations in the critical region” Phys. Rev. E 76,031138 (2007)

    2007

  6. [14]

    Deterministic Chaos

    H. Schuster, “ Deterministic Chaos” Weinheim, Germany: VCH, 1998

    1998

  7. [15]

    Statistical Mechanics

    K. Huang “Statistical Mechanics” 2nd New York: Wiley 1987

    1987

  8. [16]

    The Earth as a living planet: Human-type diseases in the earthquake preparation p rocess

    Y. Contoyiannis, S.M. Potirakis, K. Eftaxias, “The Earth as a living planet: Human-type diseases in the earthquake preparation p rocess” Nat. Hazards Earth Syst. Sci 13(2013) 125-139

    2013

  9. [17]

    Bousseljot, R., Kreiseler, D., and Schnabel, A.: Nu tzung der EKG-Signal datenbank CARDIODAT der PTB uber das Internet, Biom edical Engineering/Biomedizinische Technik, 40, 317–318, 1995

    1995

  10. [18]

    and Bousseljot, R.: Automatisierte EKG-Auswertung Mit Hilfe der EKG-Signaldatenbank CARDIODAT der PTB , Biomedical Engineering/Biomedizinische Technik, 40, 319–320, 1995

    Kreiseler, D. and Bousseljot, R.: Automatisierte EKG-Auswertung Mit Hilfe der EKG-Signaldatenbank CARDIODAT der PTB , Biomedical Engineering/Biomedizinische Technik, 40, 319–320, 1995

    1995

  11. [19]

    L., Amaral, L

    Goldberger, A. L., Amaral, L. A. N., Hausdorff, J. M., Ivanov, P. C., Peng, C.-K., and Stanley, H. E.: Fractal dynamics in physiology: Alterations with disease and aging, Proc. Natl. Aca. Sci., 99, 2466–2472, 2002

    2002

  12. [20]

    Intermittent criticality revealed in ULF magnetic fields prior to the 11 March 2011 Tohoku earthquake ( /g1839/g3050=9)

    Y.Contoyiannis, S.M.Potirakis, K.Eftaxias, M.Hayaka wa, A.Schekotov “Intermittent criticality revealed in ULF magnetic fields prior to the 11 March 2011 Tohoku earthquake ( /g1839/g3050=9)”, Physica A 452(2016)19–28

    2016

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.