(101 days)
The monitor is indicated for use by healthcare professionals whenever there is a need for monitoring the physiological parameters of patients. The monitor is intended to be used for monitoring and recording of, and to generate alarms, for, multiple physiological parameters of adults, pediatrics, and neonates. The monitor is intended for use by trained healthcare professionals in a hospital environment. The monitor is also intended for use during patient transport inside and outside of the hospital environment. The monitor is only for use on one patient at a time. It is not intended for home use. Not a therapeutic device. The monitor is for prescription use only. The ECG measurement is intended to be used for diagnostic recording of rhythm and detailed morphology of complex cardiac complexes (according to AAMI EC 11). ST segment monitoring is intended for use with adult patients only and is not clinically validated for use with neonatal and pediatric patients. The Predictive Temperature unit is intended for use with adult and pediatric patients in a hospital environment. The SSC Sepsis Protocol, in the ProtocolWatch clinical decision support tool, is intended for use with adult patients only. The derived measurement Pulse Pressure Variation (PPV) is intended for use with sedated patients receiving controlled mechanical ventilation and mainly free from cardiac arrhythmia. The PPV measurement has been validated only for adult patients. The transcutaneous gas measurement (tcGas) is restricted to neonatal patients only. BIS is intended for use under the direct supervision of a licensed health care practitioner or by personnel trained in its proper use. It is intended for use on adult and pediatric patients within a hospital or medical facility providing patient care to monitor the state of the brain by data acquisition of EEG signals. The BIS may be used as an aid in monitoring the effects of certain anesthetic agents. Use of BIS monitoring to help guide anesthetic administration may be associated with the reduction of the incidence of awareness with recall in adults during general anesthesia and sedation.
The Philips IntelliVue Patient Monitor family comprises the multi-parameter patient monitor models: MP2, X2, MP5, MP5T, MP5SC, MP20, MP30, MP40, MP50, MP60, MP70, MP80, MX600, MX700 and MX800 IntelliVue Patient Monitors that consist of display units including built-in or separate flat panel displays and central processing units (CPU) and physiological measurement modules. All monitors share the same system architecture and exactly the same software is executed on each monitor. The IntelliVue Patient Monitors measure multiple physiological parameters such as surface ECG, invasive and non-invasive pressure, etc., generate alarms, record physiological signals, store derived data, and communicate derived data and alarms to central stations via the IntelliVue Clinical Network. The subject modification is enhancement of ProtocolWatch SCC Sepsis protocol by an introduction of configurable thresholds, modification of sepsis criteria and free text fields within the Sepsis Resuscitation Bundle and Sepsis Management Bundle phase of SSC Sepsis protocol. Additionally the software revision J.03 is made available for the entire IntelliVue Patient Monitors family.
The provided document is a 510(k) summary for the Philips IntelliVue patient monitors. It describes the modified devices, their intended use, and indicates that verification, validation, and testing activities were performed to establish performance, functionality, and reliability. However, it does not provide specific details on acceptance criteria or the study that proves the device meets those criteria with numerical performance metrics for any AI/ML component.
The primary context of this submission is an enhancement to the "ProtocolWatch SCC Sepsis protocol" by introducing configurable thresholds, modifying sepsis criteria, and adding free text fields. It also mentions a software revision J.03 for the entire IntelliVue Patient Monitor family. This suggests a software update to an existing device rather than a new AI/ML-driven diagnostic or prognostic tool.
Based on the provided text, I can extract the following:
1. A table of acceptance criteria and the reported device performance
The document states: "Verification, validation, and testing activities establish the performance, functionality, and reliability characteristics of the modified devices with respect to the predicate. Testing involved system level and as well as testing from the hazard analysis. Pass/Fail criteria were based on the specifications cleared for the predicate devices and test results showed substantial equivalence. The results demonstrate that the Philips IntelliVue patient monitors meet all reliability requirements and performance claims."
This is a general statement. Specific acceptance criteria (e.g., minimum sensitivity, specificity, accuracy, or other performance metrics) for the Sepsis Protocol enhancement or other features are not detailed in this 510(k) summary. Therefore, a table cannot be constructed with specific numerical acceptance criteria and reported performance.
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
This information is not provided in the 510(k) summary. The document does not describe specific clinical studies or data sets used for testing the modified features.
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)
This information is not provided in the 510(k) summary. Given the nature of a patient monitor software update, it's unlikely that "experts" (in the typical sense of radiologists or clinicians reviewing cases for ground truth in an AI/ML context) were involved in a stated ground truth establishment process for this type of submission. The modifications relate to a "Sepsis protocol" which would likely be based on established clinical guidelines.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
This information is not provided in the 510(k) summary.
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
A multi-reader multi-case (MRMC) comparative effectiveness study with human readers and AI assistance is typical for AI-driven diagnostic tools. This submission describes a "Sepsis protocol" enhancement within a patient monitor, which is a clinical decision support tool rather than a standalone diagnostic AI. Therefore, an MRMC study is not applicable and was not mentioned. The document does not describe effects on human reader improvement.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
The document mentions "Verification, validation, and testing activities" which "establish the performance, functionality, and reliability characteristics of the modified devices." This implies various forms of testing. However, it does not specifically refer to a "standalone" algorithm performance study in the context of AI/ML, nor does it provide metrics for such a study. The "Sepsis protocol" is described as a "clinical decision support tool," implying a human-in-the-loop context.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The document does not detail the specific ground truth used. For the "Sepsis Protocol," the ground truth would likely be based on established clinical guidelines and definitions for sepsis, rather than individual expert consensus on specific cases for an AI model.
8. The sample size for the training set
This information is not provided in the 510(k) summary. Given that the submission focuses on an "enhancement of ProtocolWatch SCC Sepsis protocol by an introduction of configurable thresholds, modification of sepsis criteria and free text fields," it's highly probable that this change involves logic and configuration updates based on medical knowledge, rather than a machine learning model that requires a dedicated "training set" in the conventional sense.
9. How the ground truth for the training set was established
This information is not provided. As mentioned above, it's unlikely a "training set" in the AI/ML sense was used for this type of modification.
§ 870.1025 Arrhythmia detector and alarm (including ST-segment measurement and alarm).
(a)
Identification. The arrhythmia detector and alarm device monitors an electrocardiogram and is designed to produce a visible or audible signal or alarm when atrial or ventricular arrhythmia, such as premature contraction or ventricular fibrillation, occurs.(b)
Classification. Class II (special controls). The guidance document entitled “Class II Special Controls Guidance Document: Arrhythmia Detector and Alarm” will serve as the special control. See § 870.1 for the availability of this guidance document.