Search Results

Esprit Nova is a software-only product intended to assist a physician in the diagnostic evaluation of sleep quality and sleep disordered breathing in adults only. Esprit Nova analyzes the physiological signals and automatically scores sleep events; including the stages of sleep, microarousals, snoring, periodic limb movements, desaturations and sleep disordered breathing events apneas, hypopneas, and respiratory event related arousals). This device is to be used under the supervision of a physician. The device does not provide any diagnostic conclusion about the patient's condition.

Esprit Nova is a software application that analyzes previously recorded physiological signals obtained during sleep. The software can analyze any EDF files acquired with Trackit SleepWalker (K010460) Compumedics Somté PSG System( K072201) and/or Sandman (K040113). Automated algorithms are applied to the raw signals to interpret the raw signal information. The software automates recognition of:

• Sleep stages Rapid Eye Movement (REM), nREM (N1, N2, N3) and wake
• Heart rate
• Snoring
• Sleep/wake
• Body position
• Arousals
• EEG, ECG, EOG, EMG waveforms
• Sa02
• Airflow
• Respiratory Effort
• Apneas and Hypopneas
• Oxygen desaturations.
• Limb movements .
  The software identifies and rejects periods with poor electroencephalography signal quality. The output of the device is a comprehensive sleep study report. Medical and history information can be entered from a questionnaire. The automated analysis of physiological data is integrated with the questionnaire data, medical and history information and included in the report. The Esprit Nova analysis module is hosted on a central server, while the administrative module is a stand-alone application for use on Microsoft Windows 7 (or higher) operating system platforms. The administrative module works as a local database to keep records of patients' and to transmit data back and forth to the central server.

Here's a summary of the acceptance criteria and the study that proves the device meets them, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criterion for Esprit Nova's performance was to demonstrate equivalence to the reported performance of the predicate device (Sleep Profiler K153412 for sleep staging, and Sandman Elite K153353 for PLM detector) and to "perform equivalently to the gold standard" (panel of 3 electrophysiologists). The table above shows the reported performance and how it compares to these benchmarks.

2. Sample size used for the test set and the data provenance

• Sample Size: 60 PSG studies.
• Data Provenance: The studies were from "adult patients that were seen at a Sleep Clinic." The specific country of origin is not explicitly stated, but the submitter's address is in Dundas, Ontario, Canada, which could suggest Canadian data. The studies comprised a mix of diagnoses:
  • 40 studies with varying degrees of Obstructive Sleep Apnea (12 severe, 14 moderate, 14 mild)
  • 1 study for Restless Legs Syndrome
  • 3 studies for Restless Legs Syndrome + OSA
  • 7 Titration studies
  • 9 normal studies
• Retrospective/Prospective: Not explicitly stated, but the description "studies from adult patients that were seen at a Sleep Clinic" typically implies retrospective data collection from existing patient records.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

• Number of Experts: 3
• Qualifications of Experts: "3 EEG board certified medical professionals" (referred to as "electrophysiologists" in another section).

4. Adjudication method for the test set

• The text describes the ground truth as "a panel of 3 EEG board certified medical professionals scoring studies as done in their daily practice (i.e gold standard)". This implies that the consensus or individual scores of these three experts served as the ground truth. However, the specific adjudication method (e.g., 2+1, 3+1, majority vote, or if each expert's scoring was considered individually against the algorithm's output without inter-expert adjudication for the final ground truth) is not explicitly stated. It just says their scoring was the "gold standard."

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

• No MRMC comparative effectiveness study was done in the context of human readers with AI assistance versus without AI assistance. The study focused on the standalone performance of the algorithm against expert scores. The "gold standard" was the panel of experts, not human readers assisted by AI.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

• Yes, a standalone performance study was done. The entire clinical validation section describes the "Algorithm Performance Comparison" of Esprit Nova against the "expert scoring," which establishes its standalone performance.

7. The type of ground truth used

• Expert Consensus/Opinion: The ground truth for the sleep staging and PLM detection was established by a "panel of 3 EEG board certified medical professionals" scoring the studies. This is explicitly stated as the "gold standard."

8. The sample size for the training set

• The document does not explicitly state 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 document does not provide information on how the ground truth for the training set was established. It only details the ground truth establishment for the clinical validation (test) set.

Ask a specific question about this device

The Datex-Ohmeda S/5™ Nellcor Compatible Saturation Module, E-NSAT, and accessories are indicated for monitoring arterial oxygen saturation of hospitalized patients. The device is indicated for use by qualified medical personnel only.

The Datex-Ohmeda S'57M E-NSAT module is a single-width plug-in parameter module for a Datex-Ohmeda S 5 modular monitoring system. E-NSAT is used for monitoring arterial oxygen saturation of hospitalized patients. The E-NSAT module provides continuous non-invasive measurement of the pulse rate and oxygen saturation. Visual and auditory alarms are given for high/low pulse rate and high/low saturation values. The user can adjust the alarm limits. The E-NSAT module is designed using the Nelleor pulse oximetry technology. The real-time pulse oximetry signal is displayed without any scale. The calculated Sp()2 value and pulse rate can be shown in a separate numberfield, or beside the pulse oximetry signal. The parameters can be trended and the trends can be printed with the Recorder module in the momitor or with a separate laser printer.

The provided text is a 510(k) Premarket Notification summary for a medical device, the Datex-Ohmeda S/5™ Nellcor Compatible Saturation Module, E-NSAT. This document focuses on demonstrating substantial equivalence to a predicate device based on technical and functional similarities, rather than presenting a detailed clinical study with acceptance criteria and performance metrics for a novel AI device.

Therefore, most of the requested information regarding acceptance criteria, specific study designs (like MRMC or standalone performance), sample sizes, ground truth establishment for AI training/testing, and expert qualifications for ground truth cannot be extracted from this document, as it describes a traditional medical device submission and not an AI/ML device.

However, I can provide what is available or inferable:

1. A table of acceptance criteria and the reported device performance

The document does not detail specific acceptance criteria for performance metrics (e.g., accuracy, precision) in a tabular format, nor does it report specific performance values for the E-NSAT module against such criteria. Instead, it states that the device has been "thoroughly tested through validation and verification of specifications" and assessed against various medical device standards. The key assertion is "The Customer and parameter specifications are the same" as the predicate device, implying that the performance targets are identical to the predicate.

2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

Not applicable. This document does not describe a clinical study with a test set of data in the context of AI/ML evaluation. The primary validation appears to be non-clinical, focusing on compliance with standards and equivalence to a predicate device.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

Not applicable. This document does not describe establishing ground truth for a test set in the context of an AI/ML device.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

Not applicable. This document does not describe a test set or adjudication method.

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

Not applicable. This document describes a pulse oximeter, not an AI-assisted diagnostic tool that would typically undergo an MRMC study.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

Not applicable. This device is a measurement module, and its performance is inherent in its design and operation; it's not an algorithm that functions without human interaction in its use context. "Standalone performance" in the AI/ML context doesn't apply here. The device itself is "standalone" in its measurement function, but that's a different concept.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

Not applicable. Ground truth, in the context of evaluating a diagnostic AI device, is not mentioned in this document. The device's performance is likely evaluated against established standards for pulse oximetry measurements.

8. The sample size for the training set

Not applicable. This is not an AI/ML device, and thus there is no "training set."

9. How the ground truth for the training set was established

Not applicable. There is no training set mentioned for this device.

Summary of available information related to performance/validation from the document:

While the document doesn't fit the AI/ML framework of your questions, it does describe the validation process for a traditional medical device:

• Validation Method: The device has been "thoroughly tested through validation and verification of specifications" against a comprehensive list of international and national standards, including IEC 60601-1 (safety), IEC 60601-1-2 (EMC), ISO 9919 (pulse oximeters for medical use safety requirements), and EN 865 (pulse oximeters particular requirements). FDA guidance documents for premarket notifications and software were also considered.
• Equivalence Basis: The primary "proof" that the device meets acceptance criteria (implicit, as being safe and effective) is its substantial equivalence to a legally marketed predicate device (Datex-Ohmeda M-NSAT Module K020479). This equivalence is based on:
  • Identical intended use and indications for use.
  • Same operating principle and fundamental scientific technology.
  • Compatibility with the same monitor software and user interface.
  • Identical customer and parameter specifications.
  • Same safety and effectiveness profile.
  • Same manufacturing processes.
• Key Change and Its Clearance: The main difference is an updated Nellcor OEM Pulse Oximetry electronic measurement board (MP100 instead of MP404). Crucially, the document states: "The MP100 is an enhanced version of the MP404, and has been previously cleared with the Sandman Express (K040113)." This prior clearance of the core measurement technology likely significantly contributes to the non-clinical validation reliance.

Ask a specific question about this device

The device is for use on individual adult and pediatric patients in hospital areas and hospital-type facilities, such as clinics. Clinical users may use the monitor during hospital transport. These indications were previously cleared under K030931 and K041434.

The purpose and function of the LIFEGARD II Family is to monitor:

• ECG
• Heart rate
• Non-invasive blood pressure (NIBP)
• Functional arterial oxygen saturation (SpO2)
• Respiration rate
• Temperature and
• End-tidal carbon dioxide (EtCO2)
• Impedance Cardiography (ICG)
• Arrhythmia with ST segment detection
• Continuous non-invasive blood pressure (CNIBP)

The Impedance Cardiography (ICG) is intended for use on individual adult patients that meet the limits specified below:

• Height: 4 ft - 7 ft - 6 in (122 - 229 cm)
• Weight: 67 - 350 lbs (30 - 159 Kg)

The CNIBP is a non-invasive blood pressure monitor which uses a pressure sensor placed on the wrist over the radial artery. This device is intended to be used on patients by trained medical personnel to continuously monitor systolic, diastolic and mean blood pressure and pulse rate. This pressure information is intended to guide clinicians in the therapeutic management of their patients by providing accurate and frequent updated blood pressure information in a safe, non-invasive, easily obtained and comfortable manner.

The LIFEGARD II Family is a compact, lightweight device for measuring, processing, printing, and displaying information derived from nine physiological measurements:

• Electrocardiogram (ECG). A three lead or five lead ECG is acquired and a waveform can be displayed real-time on the LCD screen or permanently recorded on the optional printer. The design of the ECG function is derived directly from the predicate devices, the parent C3 Family of patient monitors.
• Respiration. Waveform and numeric respiration rate value from ECG. Airway respiration rate from EtCO2 if available. The design of the respiration parameter is derived directly from the predicate devices, the parent C3 Family of patient monitors.
• Arrhythmia with ST segment analysis software performs five distinct operations: beat detection, lead selection, beat classification, ventricular tachyarrhythmia detection, and ST segment measurement. The design of arrhythmia with ST segment function provides VF and ST segment deviation functionality similar to the predicate device, M3046A (M2/M3/M4) Compact Portable Patient Monitor
• Pulse Oximetry (SpO2). Functional Oxygen Saturation is calculated from the ratio of light transmissivity through the capillary bed at two wavelengths. The SpO2 subsystem uses software and firmware that is used in the OxiMAX Pulse Oximetry System and Sandman SD20 Amplifier patient monitor and system.
• The temperature is measured using thermistor probes for continuous temperature measurements. These are pre-amendment devices.
• Blood pressure is measured non-invasively (NIBP) by the oscillometric method. This is the same as in the predicate devices, parent C3 Family of patient monitors.
• Continuous non-invasive blood pressure (CNIBP) is a non-invasive blood pressure monitor which uses a pressure sensor placed on the wrist over the radial artery. This is the same technology as in the predicate device, Vasotrac APM205A.
• End-tidal CO2 (EtCO2) This method pulls a constant sample flow of exhaled breath from the patient, and analyzes it with a remote CO2 sensor built into the measurement system. This is the same as in the predicate devices, parent C3 Family of patient monitors.
• Impedance Cardiograph (ICG). This mode uses Thoracic Electrical Bioimpedance (TEB). This is identical in all aspects to the predicate device, the C3 ICG.
• An internal thermal printer records waveforms, hemodynamic parameters and tabular trends on a 50-mm wide strip chart.
• External USB printing identical to the parent C3 Family.

The LIFEGARD II Family is powered by internal sealed lead-acid batteries or from the mains supply via a battery eliminator. A fully charged battery will power the monitor for three hours.

The provided text describes a 510(k) summary for the Analogic Corporation LIFEGARD II Family of patient monitors. It extensively details the device's characteristics, intended use, and comparison to predicate devices, focusing on regulatory equivalency rather than specific detailed clinical study results to demonstrate acceptance criteria performance.

Therefore, much of the requested information regarding a detailed study proving the device meets acceptance criteria, including sample sizes, expert qualifications, adjudication methods, and ground truth establishment for AI performance, is not present in this document. This document primarily outlines non-clinical tests and regulatory compliance to establish substantial equivalence with predicate devices.

However, based on the information available, I can construct the table of acceptance criteria (as inferred from the non-clinical tests) and indicate the reported device performance in a general sense where mentioned.

Here's the breakdown of what can be extracted and what cannot:

1. Table of acceptance criteria and the reported device performance

The document lists various non-clinical tests and standards that the device will be subjected to. These standards implicitly define the acceptance criteria for various aspects of the device's safety and performance. The "reported device performance" is generally stated as "will be conducted" or "demonstrates that the performance... is substantially equivalent."

2. Sample size used for the test set and the data provenance:

• Not provided. The document focuses on non-clinical testing and regulatory compliance rather than detailed clinical study results with specific sample sizes. The "AAMI EC 57 tests already performed" implies some test data exists, but details are absent.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• Not provided. This information would typically be found in a clinical study report, which is not part of this 510(k) summary.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

• Not provided. This information is absent.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

• Not applicable / Not provided. The LIFEGARD II Family device is a patient monitor, not an AI-assisted diagnostic tool that would typically involve human readers interpreting results in a comparative effectiveness study. The "Arrhythmia with ST segment analysis software" is an algorithm within the device, not an AI for human interpretation enhancement. The document primarily focuses on the device's technical specifications and equivalency, not its impact on human reader performance.

6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:

• Yes, implicitly. The context of the "Arrhythmia with ST segment analysis software" and other measurement parameters (SpO2, NIBP, EtCO2, etc.) indicates these perform their functions as standalone algorithms within the device. The document states: "Arrhythmia with ST segment analysis software performs five distinct operations: beat detection, lead selection, beat classification, ventricular tachyarrhythmia detection, and ST segment measurement." Similarly, SpO2 calculates saturation, and NIBP measures blood pressure. These are inherent functions of the device's algorithms.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

• Not explicitly detailed, but inferred from testing standards. For parameters like ECG, NIBP, SpO2, and EtCO2, ground truth would typically be established by established reference methods or highly accurate (often invasive or laboratory-grade) measurement devices during testing against the standards mentioned (e.g., AAMI EC 57 for ECG, ANSI/AAMI SP10 for NIBP, EN 865 for SpO2, EN 864 for EtCO2). The document suggests that the performance of these integrated modules is evaluated against these industry-accepted standards, implying that the "ground truth" is defined by such reference methods.

8. The sample size for the training set:

• Not applicable / Not provided. This device pre-dates the widespread use of deep learning and large-scale training sets for medical devices, particularly for a summary from 2005. The mention of "design of the ECG function is derived directly from the predicate devices" and "SpO2 subsystem uses software and firmware that is used in the OxiMAX Pulse Oximetry System" points to established, possibly rule-based or conventional signal processing algorithms, rather than models requiring large training sets in the modern AI sense.

9. How the ground truth for the training set was established:

• Not applicable / Not provided. As mentioned above, the concept of a "training set" with ground truth in the current AI/ML context is likely not relevant to this device's development as described in a 2005 510(k) summary. The algorithms' development would have relied on engineering principles, signal processing, and testing against known physiological signals and reference measurements, rather than ground truth established for a statistical learning model.

Ask a specific question about this device