Dataset inspector
Open Analyze → Signal Explorer, pick a dataset from the dropdown, and the inspector loads inline — the main workspace for one recording. (It's also reachable directly at /datasets/{id}, e.g. from a run detail page's Open dataset link.)
Tabs
Explorer
Interactive signal view:
- ECG / PPG — downsampled waveform with pan/zoom; R-peak or pulse peak markers
- RR CSV — RR interval tachogram
- Quality overlay when preprocess has run (low-SQI segments highlighted)
Requires a completed preprocess run (or RR upload) for full detail.
Overview
Pipeline actions and latest metrics:
| Signal type | Typical sequence |
|---|---|
| ECG | Preprocess ECG → Compute HRV → Arrhythmia burden (optional) |
| RR CSV | Compute HRV directly |
| PPG | Preprocess PPG → Compute PRV |
| BP | Analyze BP |
Panels show:
- Preprocess — SQI, peak count, artifact rate; run/re-run button. For a cohort-wide view of preprocessing status across every dataset, use Analyze → Signal Quality instead of opening each dataset individually.
- HRV / PRV — metric cards with units; links to source run
- Arrhythmia burden — irregularity % and pause metrics when RR exists
- Signal metadata — format, sample rate, duration, channels
Annotations
Add time-range labels and notes on the recording timeline. Annotations persist with the dataset for collaboration and QC notes.
Runs
Full pipeline history for this dataset:
- Status badges per run
- Links to run detail
- Select two completed runs → Compare selected
Running pipelines
- Click the pipeline action (for example Run preprocessing)
- The job enters
pendingthenrunning— watch Jobs in the navbar - On completion, metric panels refresh automatically
- Failed runs show error messages on the run detail page
Use Full ECG analysis (via generic job API) to chain preprocess + HRV in one run when calling the analysis service directly.
Interpret from dataset context
When HRV completes, use Interpret from the linked run detail page. The narrative cites RMSSD, LF/HF, SQI, and arrhythmia burden when available.