Beat detection
- QRS detection
- Total beat count across the full recording.
Holter ECG Analytics
AI analysis for 1- to 3-lead Holter recordings.
Cardiolyse analyses 1- to 3-channel Holter recordings beat by beat. It counts every beat, classifies rhythms and ectopic beats, flags tachycardia and bradycardia runs, measures intervals and scores signal quality — across the full recording.
3-lead Holter · every beat analysed
1–3 leads
Single- to 3-channel Holter recordings
0
Rhythm and beat findings classified
0
Intervals measured: P-wave, PR/PQ, QRS, QT
Per beat
Signal-quality and confidence scoring
What the analysis returns
Hours of signal in. A structured Holter read out.
Each 1- to 3-lead recording is processed from start to finish. Results come back as counts, findings, events and measurements that a qualified healthcare professional can review and confirm.
Rhythm classification
- Normal sinus rhythm (NSR)
- Baseline sinus rhythm identification.
- Atrial fibrillation (AFib)
- Irregular atrial rhythm detection.
- 1st-degree AV block
- Prolonged AV conduction detection.
- Premature ventricular contraction (PVC)
- Ectopic ventricular beat detection.
- Premature atrial contraction (PAC)
- Ectopic atrial beat detection.
- Tachy-brady pattern
- Alternating tachycardia and bradycardia episodes.
- Pause
- Sinus pause detection.
Rate-based events
- Tachycardia run
- Tachycardia episode · rate above configured threshold
- Bradycardia run
- Bradycardia episode · rate below configured threshold
Heart rate summary
- HR min / max / average
- Summary heart rate statistics across the recording.
Interval analysis
- P-wave duration
- Atrial depolarisation duration.
- PR / PQ interval
- Atrioventricular conduction time.
- QRS duration
- Ventricular depolarisation duration.
- QT interval
- Ventricular depolarisation and repolarisation duration.
Signal quality
- Artifact / noise detection
- Identification of non-physiological signal segments.
- Beat quality / signal quality score
- Per-beat confidence scoring.
From recording to insights
Every beat counted. Every run flagged.
Long recordings capture the full picture. The analysis walks the whole signal, labels each beat and marks the stretches that meet event criteria — while noisy segments are identified rather than counted as rhythm.
How it works
From a Holter recording to a structured result.
-
01
Record
The patient wears a 1- to 3-lead Holter recorder or ECG patch for the monitoring period your service defines.
-
02
Upload
The recording is sent to Cardiolyse through the REST API or a partner integration.
-
03
Analyse
Beats are detected and classified, rate-based events flagged, intervals measured and signal quality scored across the full recording.
-
04
Results
Findings return as structured data and a report, ready for review by a qualified healthcare professional in your own system.
Technical fit
Built for Holter services and device partners.
Designed to sit behind recorders, patches and monitoring services, so analysis happens in the background and results land where your team already works.
Inputs
1–3 lead ECG
- Single-, 2- and 3-channel Holter recordings
- Ambulatory recorders and ECG patches
- Device-agnostic digital signal input
Outputs
Structured Holter read
- Beat count and beat classification
- Rhythm findings and rate-based events
- HR min / max / average and intervals
- Artifact segments and per-beat quality score
- JSON via REST API and PDF report
Deployment
API-first
- EU-hosted managed service
- OEM and white-label options
- Batch processing of recordings
- Documented ECG API & Integration
Who uses it
For teams that analyse long ECG recordings at scale.
Monitoring services
Holter and remote-monitoring providers
Give every recording a consistent first analysis, so reviewers start from counted beats, flagged events and measured intervals.
Device manufacturers
Recorder, patch and wearable makers
Add beat-level Holter analytics for 1- to 3-lead devices through one API, under your brand.
Research
Clinical studies and research groups
Analyse ambulatory ECG data with the same criteria across every participant and site, with results as structured data.
Frequently asked
Holter ECG analytics, answered.
Which Holter recordings can Cardiolyse analyse?
Holter ECG analytics is built for 1- to 3-lead recordings — single-channel patches and 2- or 3-channel Holter recorders. For 12-lead resting ECGs, see Resting ECG Analytics.
What does the analysis detect?
It counts beats (QRS detection) and identifies normal sinus rhythm, atrial fibrillation, 1st-degree AV block, premature ventricular and atrial contractions, tachy-brady patterns and sinus pauses. It also measures P-wave duration, PR/PQ, QRS and QT intervals and summarises heart rate.
How does it handle noise and poor signal?
Artifact and noise detection identifies non-physiological signal segments, and every beat receives a quality and confidence score, so unreliable stretches are visible instead of being counted as rhythm.
How are results delivered?
As structured JSON through the REST API and as a PDF report, so they can be shown in your own platform or monitoring service. Results support qualified healthcare professionals, who review and confirm the findings.
Can we test it on our own recordings?
Yes. Send us a small set of anonymised 1- to 3-lead recordings and we will walk you through the results. Contact us to set it up.
For everything else, see the full FAQ.
See your Holter recordings analysed beat by beat.
Send us a few anonymised 1- to 3-lead recordings — we will analyse them and review the results with you.