Beta · we're expanding our cohort — vets & horse-owners welcome Read more →

CARLON EQUINE

Heart Science

Research-grade HRV analytics for horses · Polar H10 · GDPR · Local.

RR intervals, autonomic balance, vagal tone.
Made visible — for vets, owners, research.

1000 HzPolar H10 · ECG-grade

25 metricsResearch-grade export

24 h+Offline recording

GDPRData stays local

Analyze your session here

ZERO UPLOAD · DATA STAYS LOCAL · CARLON_EQUINE / OPEN_HRV_JSON

— or —

1000 Hz Sampling

Polar H10 — der einzige Consumer-Brustgurt mit EKG-grade R-Peak-Sampling, peer-reviewed gegen 12-Kanal-EKG.

Gilgen-Ammann 2019 · Schaffarczyk 2022

25 Metrics

Zeit-Domain, Frequenz, Nichtlinear, Equine-spezifisch.

Task Force 1996 · Visser 2002 · Stucke 2015

24 h+ Offline

Polar H10 zeichnet am Gurt auf — kein iPhone in Reichweite nötig. Sync später per Ein-Klick.

PSFTP-Protokoll · Mehrtages-fähig

0 KB Upload

Alle Roh-Daten und Auswertungen bleiben auf deinem Gerät. Kein Account, kein Cloud-Default.

DSGVO Art. 25 Privacy-by-Design

SHA-256 Integrity

Jeder JSON-Export mit Hash + Provenance-Chain — audit-fest.

W3C PROV-O · FAIR · Open-Schema

Method transparency · equine-specific

PSD bands (equine) VLF 0–0.01 · LF 0.01–0.07 · HF 0.07–0.6 Hz — angepasst an niedrigere Ruhe-HF und Atemfrequenz des Pferdes (nicht Human-TF-1996). Visser 2002 · Equine Vet J 34:280-288

RMSSD reference range (equine, at rest) 80–250 ms typisch für Warmblut/Sportpferd in Ruhe — Human-Cutoffs (20–60 ms) sind nicht übertragbar. Stucke 2015 · Physiol Behav 138:95-105

AV-block classification Detektiert werden Mobitz-Typ-I-Pausen (2°-AV-Block) — beim ruhenden Pferd physiologisch (vagaler Reflex), verschwinden bei Belastung. Keine Detektion höhergradiger Blocks. Reef 1998 · Vet Clin Equine 14:21-43

AFib screening · RR-only Polar H10 liefert RR-Intervalle, keine P-Wellen. Screening erfolgt über extreme RR-Irregularität (RMSSD > 302 ms Schwelle). Kein Ersatz für 12-Kanal-EKG bei Verdacht. Mitchell 2018 · J Vet Intern Med 32:1232-1240

Live-Demo · was hinten rauskommt

Lege eine Aufnahme rein — sieh die Tiefe.

Hier ist die Demo-Auswertung von Hidalgo (12 J., Warmblut, 6:28 min Ruhe-EKG via Polar H10). Kein Login, kein Upload — die echten Charts und Metriken, die auch dein Klinik-Team sehen würde.

CARLON Equine Demo — Hidalgo · Ruhe-EKG mit Executive Summary, KPI-Tafel, Tachogramm mit vagale AV-Blocks, Poincaré-Plot mit SD1/SD2, Power-Spectral-Density Welch

DEMO-DATEN HIDALGO · RUHE · 6:28 MIN POLAR H10 · 1000 HZ

Executive Summary · 3 Perspektiven Toggle zwischen Halter-/Tierarzt-/Forschungs-Sprache mit einem Klick.

Tachogramm mit AV-Block-Markern Beat-für-Beat RR-Verlauf, vagale 2°-AV-Blocks (Mobitz I) automatisch flagged nach Reef 1998.

Poincaré + Power-Spektrum SD1/SD2-Ellipse · Welch-PSD mit equine VLF/LF/HF-Bändern · LF/HF-Ratio mit Cross-Validation (Welch vs Lomb-Scargle).

Volle interaktive Demo öffnen Lädt im Browser · kein Login

Warum HRV beim Pferd zählt

Belastung und Erholung werden oft zu spät sichtbar.

1Problem

Klinische Untersuchung erfasst nur den Messpunkt.

Überbelastung, schleichender Stress, beginnende Regenerations-Defizite zeigen sich häufig erst, wenn sie systemisch werden — verspätet, teuer, manchmal irreversibel.

2Indikator

HRV reagiert auf autonome Veränderungen früh.

Die Herzraten-Variabilität spiegelt vagalen Tonus, sympathische Aktivierung und Erholungs-Status — oft Stunden bis Tage bevor klinische Symptome auftreten. Bei Pferden peer-reviewed seit den 2000ern.

Visser 2002 · Stucke 2015 · Younes 2017

3Lösung

CARLON macht den Mess-Workflow alltagstauglich.

Polar H10 anlegen, iPhone-App starten oder am Gurt aufzeichnen, Auswertung im Browser ansehen — 5 min Ruhe oder 24 h Langzeit. Tierarzt sieht klinische Sprache, Halter sieht laienverständliche Sprache.

Beta · join our growing cohort

Help shape an equine HRV tool built for daily clinical use.

Who we're looking for

Equine vets, clinics, sport-horse owners — anyone with a Polar H10 (or open to using one) and a handful of horses to record from.

What you get

Full Pro access during beta. iOS app + this web viewer. Direct line to Jan for every question, bug or feature wish.

What we ask in return

Use it on real horses. One 15-minute feedback call after your first recordings — what worked, what didn't, what's missing.

Email Jan to join the beta — Jan Rohwedder · Founder

Built for

For people who want answers, not promises.

Veterinarians

Reproducible cardiology workups

Every JSON export carries a SHA-256 integrity hash and a W3C-PROV provenance chain. Defensible against any audit.

Researchers

Field-grade multi-horse studies

Roster + batch-trigger + batch-sync for 5–10+ horses simultaneously. Welfare assessment, training load, recovery monitoring, transport stress.

Sport-horse trainers

Autonomic recovery tracking

Compare RMSSD, DFA-α₁ and Baevsky Stress Index across sessions. Know when your horse is ready for the next load.

Engaged owners

Share with your vet

Record on iPhone, get a one-tap referral PDF, send it to your vet. They open this browser viewer, see everything.

Scientific foundation

64+ peer-reviewed sources. Every metric, every cutoff.

No black-box algorithms. Every number traces back to a published study, equine-specific where it matters.

Task Force 1996 · time- and frequency-domain HRV standards

Lipponen-Tarvainen 2019 · beat classifier + smoothness-priors detrending

Visser 2002 · equine HRV power-spectrum bands

Reef 1998 · vagal AV-blocks physiological in resting horses

Mitchell 2018 · equine AFib screening cutoffs

Stucke 2015 · equine DFA-α₁ training zones

Brennan 2001 · Poincaré SD1/SD2

Peng 1995 · DFA scaling

Costa 2002 · Multi-Scale Entropy

Porta 2001 · Symbolic Dynamics

Baevsky 1984 · autonomic stress index

Gehlen 2020 · welfare HRV markers

Positioning: CARLON Equine is a research and observation tool. It is not a medical device under MDR or FDA. All on-screen flags are screening indicators requiring veterinarian follow-up — never diagnoses.

CARLON EQUINE

Heart Science

HRV Report

—

Mean HR

—

RMSSD

—

SDNN

—

DFA α₁

—

Recording

—

Duration

—

Device

—

Session ID

—

Quality

—

Schema

—

Detected research indicators

—

Notice: This report represents automated indicator markers from peer-reviewed studies (Mitchell 2018, Vitale 2020, Gehlen 2020, Reef 1998 a. o.). It does not replace clinical assessment. Diagnostic interpretation and therapeutic decisions belong exclusively to the treating veterinarian.

Generated with CARLON · EQUINE Heart Science · carlon-equine.app · —

SESSION · —

—

Data quality

—%

Executive Summary

Section 1

Time- and Frequency-Domain

Tachogram

RR-interval over time · Drag = pan · Trackpad-pinch or + / − = zoom · Double-click = fullscreen

Task Force 1996 · Brennan 2001 · Lipponen-Tarvainen 2019

Poincaré Plot

RR_n × RR_n+1 · SD1/SD2 ellipse · Drag = pan · Trackpad-pinch or + / − = zoom · Double-click = fullscreen

Brennan 2001

Power Spectral Density

Welch · 4 Hz cubic spline · VLF/LF/HF bands (equine)

Task Force 1996

Baevsky Stress Index

Autonomic balance tachometer · Mode + AMo + MxDMn → SI

Baevsky 1984 · CARLON normalized √SI ×6

DFA-α₁ · Detrended Fluctuation Analysis

Short-term scaling exponent · Training-intensity marker

Peng 1995 · Gronwald-Hoos 2020

Symbolic Dynamics

Pattern distribution · 0V / 1V / 2LV / 2ULV

Porta 2001

Multi-Scale Entropy

Complexity index across τ scales

Costa 2002

Section 2

Non-Linear HRV + Equine Specifics

Section 3

Findings · Notes · Colleague Context

Section 4

Scientific Integrity · Limitations & Provenance

Section 5

Reproducibility · FAIR · Standards

Prepare AI consultation

Copy prompt + JSON → ChatGPT / Claude / Perplexity → AI first opinion in 30s

Owner tool · NOT a substitute for vet assessment

⚠ Important: AI answers are brainstorming aid — not a diagnosis. Final judgment belongs in the hands of a licensed veterinarian.

1 · Prompt template

2 · Compact data JSON

→ Open ChatGPT → Open Claude → Open Perplexity

Findings · Auto-synthesis

Appendix · Methodology

Algorithms & Scientific References

CARLON · v1.7.6

Beat detection & classification

Lipponen-Tarvainen (2019) — adaptive RR-correction, ectopic-detection. J Med Eng Technol 43:173-181
Sliding-median + cubic-spline artifact interpolation (Kubios-style)
Vagal AV-block detection — RR ≥1.8× moving-median, equine-physiological. Reef 1998 · Vet Clin Equine 14:21-43

Time-domain HRV

RMSSD, SDNN, pNN50, NN50 — Task Force standards. Task Force 1996 · Circulation 93:1043-1065
Triangular Index (TI) + RR-histogram bin 7.8125 ms
Bootstrap-CI (n=1000) for all primary metrics

Frequency-domain & nonlinear

Welch-PSD · 4 Hz cubic-spline-resampling · Hann-window
Equine PSD bands VLF 0–0.01 / LF 0.01–0.07 / HF 0.07–0.6 Hz. Visser 2002 · Equine Vet J 34:280-288
DFA-α₁ Detrended Fluctuation Analysis (Peng 1995)
Multi-Scale Entropy, Symbolic Dynamics (Porta 2007), Sample Entropy

Veterinary indicator markers

Acute colic marker — RMSSD-drop + HR-rise. Younes 2017 · Equine Vet J 49:212-217
Respiratory marker — HF-power (indirect, not diagnostic). Gehlen 2020 · BMC Vet Res 16:127
AFib hint — RMSSD > 302 ms threshold. Mitchell 2018 · J Vet Intern Med 32:1232-1240
Disease-risk cutoffs — SD1 < 30, sEA-RF. Vitale 2020 · Nyerges-Bohák 2025
DFA equine zones <0.5 / 0.5-0.75 / 0.75-1.15 / 1.15-1.5 / >1.5. Stucke 2015 · Physiol Behav 138:95-105

Data provenance & integrity

W3C PROV-O compliant — every value carries origin agent/activity
SHA-256 hash over canonical JSON for reproducibility
Cross-validation — primary metrics verified against secondary method
Schema versioning — carlon_equine_v1 · open-source

What this report is & is not

Is: automated computation of peer-reviewed HRV markers from an RR-interval recording
Is: a structured handoff document for the treating veterinarian
Is not: a diagnosis or replacement for clinical assessment
Is not: an MDR/FDA-cleared medical device

⚠ Clinical decisions are the veterinarian's responsibility. CARLON EQUINE provides automated marker computation from validated peer-reviewed studies. Final diagnostic interpretation, differential consideration and therapeutic decisions belong exclusively to the licensed treating veterinarian. This report is not a medical device under MDR/FDA jurisdiction.