Search Results

PeriCALM Patterns is intended for use to provide additional secondary information as an adjunct to qualified clinical decision-making during antepartum or intrapartum obstetrical monitoring at ≥32 weeks gestation for annotation and summary of the fetal heart rate recording for baseline, accelerations and the uterine pressure recording for contractions.

WARNING: Evaluation of FHR during labor and patient management decisions should not be based solely on PeriCALM Patterns annotations or summaries and should include inspection of the fetal monitor tracing and consideration of all pertinent clinical information.

PeriCALM Patterns 3.0 is a software device to be used with fetal/maternal monitoring systems. The subject device is a software algorithm to detect, label and measure features (accelerations, decelerations, baseline, and contractions) in electronic fetal monitoring (EFM) records. PeriCALM Patterns 3.0 uses fetal monitor data imported through an interface with an external source or with a third-party clinical information system. PeriCALM Patterns can function in a networked environment or as a standalone workstation.

The subject device includes present-day Long and Short-Term Memory (LSTM) neural networks to identify segments of a fetal heart rate tracing corresponding to accelerations, decelerations, baseline as well as uninterpretable segments where is missing tracing. Contraction detection is achieved using the same processes as the predicate device.

Here's a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample Size Used for the Test Set and Data Provenance

• Sample Size: A total of 70 subjects were used in the study, with one tracing per subject. This included 30 preterm subjects and 40 term subjects.
• Data Provenance: Tracings were obtained from hospitals using a variety of fetal monitor models and manufacturers (Corometrics 170, 250 series by GE HealthCare, Avalon FM50 and FM40 by Philips Medical, and S1 from Neoventa). The study was retrospective. The specific country of origin is not explicitly stated, but PeriGen, Inc. is located in Israel, however, the data provenance is through collaborations with US Hospitals through various monitor OEMs.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

• Number of Experts: 3 experts (referred to as "Truthers") were used.
• Qualifications: They were specified as "a panel of 3 experts." While the specific medical specialty (e.g., Obstetrician) is implied by the context of "fetal heart rate tracings" and "obstetrical monitoring," their exact qualifications (e.g., years of experience, specific board certifications) are not detailed in the provided text.

4. Adjudication Method for the Test Set Ground Truth

• Adjudication Method: A majority opinion approach was used to resolve inter-observer variation among the 3 experts.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• Yes, an MRMC study was done. The study design is described as "a retrospective study with multi-reader/multi-case technique."
• Effect Size of Human Readers' Improvement with AI vs. Without AI Assistance: The study's primary objective was to demonstrate the non-inferiority of the device (PeriCALM Patterns 3.0) to a set of qualified and experienced Obstetrician Gynecologists (Clinician Readers) in terms of pattern detection and baseline measurement. It does not describe a study where human readers were assisted by AI and then compared to human readers without AI assistance. Therefore, an effect size of human improvement with AI assistance is not reported in this text.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study

• Yes, implicitly. The study compared the "performance of PeriCALM Patterns 3.0" directly to "Ground Truth" and "Clinician Readers." This suggests the algorithm's performance was evaluated in isolation (without human interaction) in comparison to both human consensus and individual human expert performance. The metrics (Sensitivity, Specificity, PPV, Bias, LoA) are typical of standalone algorithm evaluations.

7. Type of Ground Truth Used

• Expert Consensus: The ground truth for this study was established by a panel of 3 experts, with a majority opinion approach used to resolve inter-observer variation.

8. Sample Size for the Training Set

• The provided text does not specify the sample size used for the training set. It only describes the test set.

9. How the Ground Truth for the Training Set Was Established

• The provided text does not specify how the ground truth for the training set was established. It only describes the ground truth establishment for the test set.

Ask a specific question about this device