Wearable RR cohort

Process RR interval exports from wearables or Holter summary files across a cohort, batch HRV computation, and compare metrics between groups.

Research question

How do HRV metrics differ across participants, treatment arms, or diagnostic groups when only RR interval data is available?

Who this is for

Digital health and wearable companies processing large RR exports at scale
Research programs with Holter RR summary files rather than full waveforms
Cohort studies comparing autonomic metrics without ECG preprocessing overhead

Data requirements

Data	Required	Purpose
RR interval CSV per subject	Yes	Direct HRV computation (`format=rr_csv`)
Subjects CSV with group labels	Yes (for cohort)	Treatment group, diagnosis, demographics
Dataset-to-subject links	Yes (for cohort)	Connect recordings to participant metadata

RR CSVs skip waveform preprocessing — HRV runs directly on the interval series.

Workflow

Create study → Import subjects → Upload RR CSVs → Link datasets → Batch HRV → Cohort compare

Step 1 — Import subjects

Navigate to Studies → Subjects and upload a subjects CSV:

external_id,age,sex,treatment_group,diagnosis
P001,58,M,control,hypertension
P002,62,F,treated,hypertension
P003,55,M,treated,hypertension
P004,60,F,control,hypertension

Step 2 — Upload RR datasets

Upload one RR CSV per participant from the study overview. Link each dataset to the corresponding subject.

Step 3 — Batch HRV

From the Cohort page or batch job endpoint, run HRV across all linked RR datasets:

curl -X POST http://localhost:8001/jobs/batch \
  -H "Content-Type: application/json" \
  -d '{
    "pipeline_name": "compute_hrv",
    "dataset_ids": ["ds-001", "ds-002", "ds-003", "ds-004"]
  }'

Step 4 — Compare cohorts

Navigate to Studies → Cohort:

Choose group by field (treatment group, sex, diagnosis)
Pick HRV metric (RMSSD, SDNN, LF/HF)
Select statistical test
Click Compare cohorts

Review box/violin charts, group means, SDs, p-values, and the outlier panel.

Step 5 — Interpret (optional)

Use Interpret on the cohort page for a narrative citing group-level statistics.

Expected outputs

Per-subject HRV metrics from RR-only pipeline
Cohort comparison chart with statistical test results
Outlier panel highlighting subjects >2 SD from group mean
Exportable chart (PNG/PDF)
AI cohort summary with cited metric values

Typical analyses

Analysis	Group by	Question
Treatment effect	treated vs. control	Did RMSSD improve after intervention?
Sex differences	male vs. female	Are frequency-domain metrics sex-dependent in this cohort?
Diagnosis stratification	hypertension vs. healthy	How does autonomic balance differ by diagnosis?
Outlier drill-down	outlier subject	Is the outlier driven by artifact or genuine physiology?

Click an outlier to open the subject hub and inspect individual HRV runs.

Menu

Documentation