Investor brief · DBDP feature engineering
Pause-Health.ai composes with the Digital Biomarker Discovery Pipeline (Duke University) for clinically grounded feature engineering from wearable signals. Today the FLIRT and DBDP HRV math is integrated in pause_ingest with 20 passing unit tests including a closed-form RMSSD correctness check; persistence back into JupyterHealth Exchange as derivedFrom FHIR Observations is Phase 2.
Why DBDP, why now
Pills: Wired in prototype wired in pause_ingest today (and asserted by tests where applicable) · Designed committed choice, activates with Phase 2 or first design partner.
FLIRT — the only PyPI-published DBDP-affiliated package — installs cleanly on Python 3.13 and computes the features our clinical advisors already trust (RMSSD, SDNN, HF/LF power, Poincaré SD1/SD2) in one call. Integrated today in pause_ingest.features.
The DBDP HRV calculator was benchmarked against Kubios, the clinical HRV reference tool. We have ported its time-domain math as a deterministic reference inside our test suite — a closed-form RMSSD correctness test asserts the exact expected value against a strict-alternation IBI series.
Each computed feature will land in JupyterHealth Exchange as a FHIR Observation with a derivedFrom link to the raw IBI window it was computed over. Auditors and security reviewers see the full chain without us building one. Today the lineage is materialized in the FHIR Bundle served at /api/mulesoft/patient/[id]/timeline; the JHE persistence path is Phase 2.
DBDP is a Duke University–led open ecosystem with active contributors across wearables research. Our upstream contributions raise our credibility with academic medical center customers — and give us pull requests instead of integrations. The planned contributions (skin-temperature schema, menopause feature module) are on /proposal/integration.
The DBDP pieces we use
MIT · on PyPI
Feature generation toolkit for wearable data. Sliding-window HRV, EDA, and accelerometer features from Empatica E4, Holter ECG, and other consumer-grade devices.
Apache-2.0
Time-domain HRV metrics from RR / IBI intervals, validated against Kubios — the clinical HRV reference.
Apache-2.0
Curated sample wearable datasets maintained by the DBDP community.
MIT
Reading + de-identifying data from Empatica E4, Bittium Faros, Biovotion Everion, Shimmer, and Muse.
Feature catalog · per-row status
All rows below are produced by the same hrv_features_flirt call in pause_ingest. The distinction the pill draws is whether Pause has its own assertion on the output — which matters for an investor or a clinical advisor reading the math. Persistence into JupyterHealth Exchange as derivedFrom FHIR Observations is Phase 2 for every row.
Pills: Wired in prototype verified in pause_ingest (closed-form or happy-path tests) · Designed produced by the same call but not yet asserted by Pause tests.
| Status | Feature | Domain | Source | Why it matters for menopause |
|---|---|---|---|---|
| Wired in prototype | RMSSD, SDNN, pNN50 | Time-domain HRV | FLIRT + DBDP HRV calculator (Pause fallback) | Autonomic dysregulation tracking. Drops in HRV correlate with vasomotor severity and sleep disruption. Closed-form unit-tested in pause_ingest. |
| Designed | HF / LF power, LF:HF ratio | Frequency-domain HRV | FLIRT | Sympathetic vs parasympathetic balance. Useful in stratifying patients for HRT vs non-hormonal pathways. Produced by the FLIRT call today but not asserted by Pause tests. |
| Designed | Non-linear: SD1, SD2, CSI, CVI | Poincaré + chaotic HRV | FLIRT | Sensitive to overall autonomic load. Early indicator for cardiovascular risk shifts post-menopause. Same FLIRT pathway as above; not yet asserted. |
| Designed | Sleep fragmentation, IBI entropy | Statistical HRV | FLIRT | Direct proxy for night sweats and disrupted sleep — the #1 patient-reported symptom in our research. |
| Designed | EDA tonic / phasic decomposition | Electrodermal activity | FLIRT (Empatica E4 archive, Phase 2) | Hot-flash detection from skin conductance peaks. |
| Designed | Activity counts, sedentary bouts | Accelerometer | FLIRT | Fatigue and activity regression — a quality-of-life signal payers care about. |
Touch the architecture
Every claim on this page that says "today" or "tested" resolves to code in pause_ingest/. The closed-form RMSSD test is at pause_ingest/tests/test_features.py · test_fallback_alternating_ibi_has_closed_form_rmssd.
Status by phase
Today
FLIRT-backed sliding-window HRV, dependency-light HRV fallback ported from DBDP, DBDP-derived test fixture committed, 20 unit tests passing across pause_ingest.convert + pause_ingest.features (including a closed-form RMSSD correctness check).
2–3 weeks
Wire flirt.with_.empatica for Empatica E4 archive ingestion. Re-evaluate devicely once numpy 2.x support lands. Persist computed feature windows back to JupyterHealth Exchange as derivedFrom FHIR Observations.
Ongoing
Propose an Open mHealth schema for skin temperature (hot-flash signal). Contribute a menopause-specific feature module to DBDP. Publish HRV-and-menopause feature-importance benchmarks against pilot cohorts.
Why this earns trust
hrv_time_domain_fallback in pause_ingest with explicit unit annotations and a closed-form correctness test. Clinical advisors can read ~50 lines and verify the math.test_empatica_ingestion_raises_phase2_error. No silent gaps.Read deeper
Continuing the investor brief