Obstetrics_gynecology · Intrapartum fetal heart-rate pattern analysis
PeriCALM Patterns 3.0
PeriGen, Inc.
PeriCALM Patterns is the rare AI medical device whose story is longevity rather than novelty: a fetal heart-rate pattern-recognition engine that was already being benchmarked against five-expert panels in the peer-reviewed literature in 2010, was validated against NICHD experts in a vendor-announced 2013 collaboration, and quietly powers large NICHD-funded research infrastructure today. The January 2025 Patterns 3.0 clearance (K241009) modernises the product's scope from beyond-36-weeks to 32-plus weeks of gestation. The honest limit is the evidence grade: what is published is agreement-grade — the software annotates tracings about as consistently as experts agree with each other — and no outcome-grade trial showing that Patterns' annotations improve neonatal outcomes has been published. With a two-clearance US lineage (the 2005 CALM Patterns predicate and the 2025 Patterns 3.0 clearance) and no located adverse-event record, it holds one mark on evidence restraint, not on doubt about its pedigree.
Performance Metrics
Clinical Evidence
The device-specific evidence is agreement-grade, and unusually honest about what that means. The foundational study (Parer & Hamilton, American Journal of Obstetrics & Gynecology 2010; co-authored by PeriGen's research lead, so manufacturer-affiliated) had five expert clinicians and the PeriCALM Patterns engine independently score 769 eight-minute segments from the last three hours of 30 tracings on a five-tier framework. The experts agreed exactly with their own majority opinion in 57 percent of segments and were within one tier in 89 percent, with weighted kappa averaging 0.58 (range 0.48–0.68); the computer's figures — 0.87 within one tier, kappa 0.52 — were not statistically different. In other words, the software sits inside the band of expert-to-expert disagreement, which is both the claim and the ceiling. A vendor-announced 2013 collaboration reported that NICHD experts agreed with Patterns' analysis in over 97 percent of assessments across the final hour of 100 tracings, as the first phase of a programme on computerised fetal heart-rate interpretation led by George Saade and Sean Blackwell; the press release cites an SMFM conference abstract (American Journal of Obstetrics & Gynecology 2013;208(1 Suppl):S316–S317) and no full paper was located, so those figures are recorded here as abstract-grade with that attribution. Since then the engine's most visible role has been as research infrastructure. An NICHD Maternal-Fetal Medicine Units Network secondary analysis (Chauhan et al., Obstetrics & Gynecology 2018) used PeriCALM Patterns to categorise intrapartum fetal heart-rate tracings when comparing small-for-gestational-age with appropriate-for-gestational-age neonates, and a 2024 methods paper (Kearney et al., MethodsX) built a fetal-monitoring database of more than 250,000 EFM records — linked to normal, acidosis and hypoxic-ischaemic-encephalopathy outcomes — on Patterns' feature extraction. What has not been published is outcome-grade evidence: no trial shows that monitoring with Patterns' annotations changes neonatal outcomes, and the cleared indication accordingly frames the output as secondary adjunct information rather than an alert or a diagnosis.
| Study | Design | n | Sensitivity | Specificity | AUC | Published |
|---|---|---|---|---|---|---|
| Parer JT, Hamilton EF (manufacturer-affiliated) | ExpertExpert review | 30 | not reported (agreement study) | not reported (agreement study) | — | Am J Obstet Gynecol 2010; 769 8-min segments from 30 tracings scored by 5 experts vs computer; computer κ 0.52 vs expert mean 0.58 — not statistically different |
Inside the algorithm
Editorial featureHow PeriCALM Patterns reads a fetal monitoring strip.
Five stages — from raw input to verdict — drawn from manufacturer documentation and the public regulatory record.
- INGEST
- NORMALISE
- DETECT
- LOCALISE
- VERDICT
Stage 01 · INGEST
Two channels stream in from the bedside monitor.
The software works on the electronic fetal monitoring record as it is acquired: the fetal heart-rate tracing on one channel and the uterine-pressure recording on the other. Its cleared scope is antepartum and intrapartum monitoring at 32 weeks gestation or later — widened from the prior version's beyond-36-weeks indication.
The input is the same strip the clinician watches; Patterns reads it alongside them, not after the fact.
Input
Cardiotocography (EFM)
Inference
Continuous, during monitoring
Inside the Auris+ Listing
Five more sections complete this device’s Auris+ Listing.
Decision Ledger
ProPro unlocks a private, cross-vendor log of every case you read on this device — what the AI called, what you concluded, and a one-line reason if you overrode it. Ready for the EU AI Act's deployer logging obligations when they land in 2028.
Clinical Evidence Deep Dive
ProPro unlocks the structured clinical-evidence summary — study count, target patient population, and a tabular accuracy-metrics view drawn from peer-reviewed sources.
Peer-Reviewed Publications
ProPro unlocks the curated peer-reviewed publication list with PubMed cross-links — the citation backbone of every editorial verdict.
Post-Market & Regulatory Conditions
ProPro unlocks the post-market surveillance summary, recall record, and the conditions of approval that bound real-world use.
AI Algorithm Version History
ProPro unlocks the chronological record of algorithm version changes — what changed when, drawn from manufacturer changelogs and regulatory filings.
Regulatory Approvals
K241009
Class II
Safety Record
Type II recall of PeriCALM Tracings central-monitoring software (models 03.14.00/03.18.00): failure of the lmstimesync module allowed workstation time drift relative to the server. Affects the Tracings display platform, not the Patterns analysis engine; recorded here as family context.
No adverse-event report or recall specific to PeriCALM Patterns was located in public reporting as of August 2026; the FDA MAUDE and enforcement databases could not be queried directly from this environment, so this is recorded as no located report rather than a verified-empty register. The adjacent safety record belongs to a sibling product: Health Canada logged a Type II recall of PeriCALM Tracings — the central-monitoring platform, not the Patterns analysis engine — in January 2021, for a time-synchronisation module failure allowing workstation clocks to drift relative to the server. By design Patterns' harm surface is indirect: it annotates and summarises the tracing but issues no alarm and no diagnosis, and the FDA-cleared labelling warns explicitly that evaluation of the fetal heart rate and patient-management decisions must not be based solely on its annotations or summaries.
Intended Use & Indications
PeriCALM Patterns 3.0 runs against electronic fetal monitoring (EFM) records as they are acquired in the maternity unit. It detects, labels and measures the features clinicians read on a cardiotocograph: the fetal heart-rate baseline, accelerations and decelerations on the heart-rate channel, and contractions on the uterine-pressure channel, presenting these annotations and summaries alongside the tracing. The FDA-cleared indication frames the output strictly as additional secondary information — an adjunct to qualified clinical decision-making — during antepartum or intrapartum monitoring at 32 weeks gestation or later, and the clearance carries an explicit warning that fetal heart-rate evaluation and patient-management decisions must not rest solely on the software's annotations or summaries: the clinician is expected to inspect the fetal-monitor tracing itself and weigh all pertinent clinical information. The prior cleared version of Patterns was indicated at greater than 36 weeks gestation; the 3.0 clearance of January 2025 widens that window to 32 weeks and later. Patterns is the pattern-recognition engine within PeriGen's wider perinatal software family (PeriCALM Tracings central monitoring; the PeriWatch surveillance platform), which this entry does not separately assess.