Experiment workflow
Cardiology follows a left-to-right research flow. You can revisit earlier steps at any time; downstream cohort views depend on completed HRV runs for linked subjects.
Create study → Upload datasets → Run pipelines → Link subjects → Compare cohorts → Interpret
Overview — study home
The study overview (/experiments/{id}) shows dataset count, completed runs, recent activity, and quick links to upload.
Readiness signal: at least one dataset uploaded and one completed preprocess or HRV run.
Upload — bring data in
Add WFDB, CSV, EDF, RR, or BP files. Format and signal type are detected automatically. Validation catches corrupt files, missing WFDB pairs, and invalid CSV columns before import.
See Data upload for format details.
Dataset inspector — analyze and explore
Each recording opens in the dataset inspector with Explorer, Overview, Annotations, and Runs tabs. Run preprocessing and HRV here; inspect waveforms; compare runs.
See Dataset inspector.
Subjects — cohort metadata
Import a subjects CSV, assign datasets to participants, and open the subject hub for multimodal views.
See Subjects and cohorts.
Cohort — group comparison
Filter by treatment group or demographics, batch HRV across datasets, compare metrics with statistical tests, and drill into outliers.
Runs and reproducibility
Every pipeline execution gets a run detail page with outputs, artifacts, inputs, and logs. Track all jobs from the navbar Jobs panel.
Typical paths
| Study type | Path |
|---|---|
| Single Holter review | Upload → preprocess → HRV → explorer → interpret |
| RR wearable cohort | Import subjects → upload RR CSVs → batch HRV → cohort compare |
| Multimodal subject | Link RR + BP → HRV + analyze BP → baroreflex → mechanistic interpret |