Holter & Ambulatory ECG

What it does

Holter & Ambulatory ECG reads a continuous recording from a Holter monitor, an adhesive patch or an event recorder, from 24 hours up to 14 days. A single day holds around 100,000 beats; the run classifies every one of them, finds the episodes that matter, and cuts a 10-second evidence strip for each finding.

The read returns heart-rate trends (minimum, mean and maximum), beat counts and ectopy burden, ventricular and supraventricular runs, AF episodes with total burden and longest episode, pauses with the longest RR interval, heart rate variability, and the correlation of each finding with the patient's diary entries.

Intended use

Decision support for physicians who order or receive ambulatory ECGs for palpitations, syncope, AF detection after cryptogenic stroke, PVC burden before ablation or in cardiomyopathy, and pacemaker follow-up. It turns a raw recording into a structured Holter summary with the strips to verify each statement, which matters most where only a vendor PDF would otherwise reach the reader.

Who it is for

Cardiologists and electrophysiologists, general practitioners and internists who receive Holter results, and Holter technicians who prepare the scan. Developers can add ambulatory ECG analysis to patch or remote-monitoring services.

Accepted input

• DICOM: Ambulatory ECG Waveform objects.
• Files: ISHNE Holter, EDF/EDF+, WFDB and vendor exports of the raw multi-channel signal.
• Diary: an optional symptom diary with times, for symptom–rhythm correlation.

A vendor PDF summary on its own does not contain the signal and cannot be read.

Requirements

Requirement	Why
Raw signal with a known sampling rate	Beat detection and RR timing depend on it
Amplitude scale and channel labels	Beat morphology and per-channel quality
Recording start time	Episode times, day/night split and diary correlation
A high sampling rate in paced patients	Pacemaker spikes are lost at low sampling rates

Large recordings are processed in chunks. Motion artefact and lead-off periods are flagged in the quality summary with their start and end times.

Optional context

Age, sex, the indication, current rate- or rhythm-control drugs and the presence of a pacemaker or defibrillator can be passed in clinical_context.

The result

One structured result per recording: summary tables, an episode list with onset, offset, duration and maximum heart rate, burden percentages, and a strip reference for every episode.

Section	Content
Arrhythmia and ectopy	Per-beat labels in AAMI classes (N, S, V, F, Q), episodes of VT, NSVT, SVT, bigeminy and AV block, PVC and PAC burden
AF episodes and burden	AF and flutter episodes of 30 s or longer with onset, offset and duration, AF burden and longest episode
Pauses and bradyarrhythmia	Pauses of 2 s and 3 s or more, high-grade AV block episodes, day/night split and symptom correlation
Heart rate variability	SDNN, RMSSD, pNN50 and LF/HF from the normal-to-normal series
Draft report	Opt-in (options.report): an English narrative that interprets the result for the reader (findings, impression, limitations); every number is checked against the findings

Standards

• DICOM: Waveform Annotation SR for beat labels and episodes; Encapsulated PDF for the printable Holter report.
• Reporting: 2017 ISHNE-HRS expert consensus on ambulatory ECG monitoring.
• Definitions: AF episodes per the 2024 ESC and 2023 ACC/AHA/ACCP/HRS atrial fibrillation guidelines; pauses and conduction disease per the 2018 ACC/AHA/HRS bradycardia guideline; HRV per the 1996 ESC/NASPE Task Force standards.

Reading limits

AV-block detection relies on P waves, which are small on single-channel recordings; every AV-block episode comes with its strip.