Search Results

VitalRhythm is a cloud-based software application for continuous and automatic analysis of cardiac arrhythmias. VitalRhythm is compatible with the "Vista Solution" platform, which includes the VitalPatch biosensor and VistaCenter web application. VitalRhythm is intended to be used for outpatient cardiac telemetry and patient monitoring in non-critical healthcare settings for non-urgent clinical decision-making. VitalRhythm provides analysis of cardiac arrhythmias using ECG data and RR-interval from the VitalPatch in patients who are 18 years of age or older. Results of the VitalRhythm are displayed within the VistaCenter web application to be reviewed and confirmed by qualified healthcare professionals and/or cardiac technicians. VitalRhythm is not intended for use in life-supporting or sustaining systems or for critical care monitoring. The arrhythmia analysis results are not intended to be the sole means of diagnosis and are offered on an advisory basis only, in conjunction with the physician's knowledge of ECG patterns, patient background, clinical history, symptoms, and other diagnostic information.

VitalRhythm is a cloud-based, arrhythmia detection software application that is compatible and intended to be used with the VitalConnect "Vista Solution" platform. The Vista Solution Platform consists of the VitalPatch biosensor, a phone (VistaPhone) or tablet (VistaTablet) relay device preloaded with the VistaPoint software application, VC Cloud and the cloud-based VistaCenter user interface.

VitalRhythm is intended to be used for outpatient cardiac telemetry and patient monitoring in non-critical healthcare settings for non-urgent clinical decision-making.

VitalRhythm analyzes ECG data and RR-interval from the VitalPatch biosensor for reporting of cardiac arrhythmias to be reviewed and adjudicated by qualified healthcare professionals and/or cardiac technicians.

The cloud-based VitalRhythm software application supports the analysis of ECG data and RR-interval using a proprietary algorithm developed using deep learning techniques. The application works in the following way:

1. VitalRhythm accepts ECG and RR-interval data transmitted from the VitalPatch via a secure, cloud-based API (Application Programming Interface).

2. ECG and RR-interval data are analyzed by VitalRhythm using a proprietary algorithm, which detects the following cardiac rhythms:

   • Atrial fibrillation/atrial flutter
   • AV Block (2nd degree, Type I and II)
   • Pause
   • Paroxysmal supraventricular tachycardia (PSVT)
   • Ventricular tachycardia/run
   • Sinus bradycardia
   • Sinus tachycardia
   • Normal sinus rhythm
   • Others (inconclusive)

3. For use with Mobile Cardiac Telemetry (MCT), i.e., outpatient cardiac telemetry: any of the above listed arrhythmias that are detected by the software algorithm are displayed in VistaCenter to be reviewed and analyzed by a qualified cardiac technician in the 24/7 attended Cardiac Monitoring Center, prior to transmitting a notifiable event consistent with the prescribed notification criteria to the prescribing physician during the monitoring period. An event report is generated by VistaCenter as a result of the analysis.

For use with patient monitoring in non-critical healthcare settings: any of the above listed arrhythmias that are detected by the software algorithm are displayed and notified in VistaCenter, in accordance with the notification criteria, during the monitoring period to be reviewed by the prescribing healthcare professional for non-urgent clinical decision-making.

The features, operating procedures, and mitigations in place for the compatible devices ensure continuous data collection and transmission to support the VitalRhythm application for the intended use. Further information is provided in the VitalRhythm Instructions for Use document.

Here's a detailed breakdown of the VitalRhythm device's acceptance criteria and the study that proves it meets those criteria, based on the provided FDA 510(k) clearance letter:

Acceptance Criteria and Device Performance

Study Details

1. Sample Size Used for the Test Set and Data Provenance:

   • Patient count: 3,309 patients (18 years of age or older).
   • Dataset count: 7,553 datasets.
   • Annotated arrhythmia episodes: 22,034.
   • Data Provenance: Retrospective, de-identified ECG and RR-interval data obtained from patients prescribed the VitalPatch biosensor across 7 clinical sites in the United States (US).

2. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications:
   The document does not explicitly state the number of experts or their specific qualifications (e.g., "radiologist with 10 years of experience") used to establish the ground truth for the test set. However, it indicates that the results are "to be reviewed and confirmed by qualified healthcare professionals and/or cardiac technicians" in the context of device usage and implies that annotation for ground truth would follow a similar expert-driven process.

3. Adjudication Method for the Test Set:
   The document does not explicitly state the adjudication method (e.g., 2+1, 3+1, none) used for establishing the ground truth of the test set. It only states that the generated "event report is generated by VistaCenter as a result of the analysis" and is "to be reviewed and analyzed by a qualified cardiac technician in the 24/7 attended Cardiac Monitoring Center, prior to transmitting a notifiable event consistent with the prescribed notification criteria to the prescribing physician during the monitoring period." This describes post-processing of algorithm results by human readers, not the method for establishing the ground truth used for algorithm validation.

4. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:
   The document does not mention a multi-reader multi-case (MRMC) comparative effectiveness study, nor does it provide an effect size of how much human readers improve with AI vs without AI assistance. The study described focuses on the standalone performance of the algorithm.

5. Standalone Performance (Algorithm Only Without Human-in-the-Loop Performance):
   Yes, a standalone performance study was conducted. The performance metrics (Sensitivity, Specificity, PPV, NPV, Accuracy) presented in the table are explicitly for the VitalRhythm algorithm "when assessed using the independent test database," indicating standalone algorithm performance against a pre-established ground truth.

6. Type of Ground Truth Used:
   The ground truth was established through annotated arrhythmia episodes. While the specific process is not detailed, it implies expert review and labeling of ECG and RR-interval data to define the 'true' presence or absence of arrhythmias. The "results are displayed within the VistaCenter web application to be reviewed and confirmed by qualified healthcare professionals and/or cardiac technicians," suggesting human consensus or expert interpretation forms the basis of the ground truth.

7. Sample Size for the Training Set:

   • Patient count: 23,587 patients.
   • Dataset count: 81,391 datasets.

8. How the Ground Truth for the Training Set Was Established:
   The document states the training database was from 354 US clinical sites, but it does not explicitly detail how the ground truth for the training set was established. It implies a process of data collection from clinical sites and subsequent processing for training the deep learning algorithm.

Ask a specific question about this device

Vista Solution 2.1 is a software user intended for use by healthcare professionals to display physiological data collected by the VitalConnect Platform wireless remote monitoring system in home and healthcare settings. Vista Solution is a secondary, adjunct patient monitor and is not intended to replace existing standard-of-care patient monitoring practices.

The VitalConnect Platform, Vista Solution is a wireless remote monitoring system consisting of the VitalPatch biosensor (K190916) and the VistaCenter and VistaPoint software (formally branded as VitalWatch in K170973). This submission is a catch-up 510(k) for changes made to the software since K170973 was cleared. The principles of operation, mechanism of action, and proposed environment of use are unchanged from the predicate device (K170973).

This document pertains to the Vista Solution 2.1, a software user interface for displaying physiological data collected by the VitalConnect Platform wireless remote monitoring system. The submission is a "catch-up 510(k)" for changes made to the software since its predicate (K170973) was cleared.

Here's the breakdown of the requested information:

1. Table of acceptance criteria and the reported device performance

The document does not explicitly state "acceptance criteria" in a quantitative performance metric table directly comparable to the predicate for all features. However, it implicitly demonstrates substantial equivalence by highlighting the similarities and changes made. The core "acceptance criteria" appear to be focused on ensuring that the modified software performs its intended function (displaying physiological data) without raising new questions of safety and effectiveness, and that new functionalities meet established standards or are extensions of existing ones.

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

The document focuses on software changes and equivalency. There is no specific mention of a "test set" in terms of patient data for evaluating clinical performance. The evaluation was done through software verification and validation, and benchtop usability testing. Therefore, details regarding sample size, country of origin, or retrospective/prospective nature of a clinical test set are not provided.

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. The submission did not involve a clinical test set requiring expert ground truth establishment in the traditional sense. The evaluation was primarily against software requirements, functional specifications, and usability standards.

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

Not applicable. There was no clinical test set requiring adjudication.

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 was not an AI-assisted diagnostic or interpretive device, nor was it a study comparing human performance with and without AI. It is a software user interface for displaying physiological data.

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

The entire submission is for a "standalone software user interface" that displays data. The performance evaluation primarily involved software verification and validation testing (IEC 62304) and benchtop usability testing (IEC 62366). These are assessments of the algorithm/software itself in meeting its specifications and being usable, rather than a clinical performance study. The device is intended for "human-in-the-loop" (healthcare professionals) to interpret the displayed data.

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

For the software validation, the "ground truth" would be the software requirements specifications and functional specifications. For usability testing, the "ground truth" would be adherence to usability standards and user task success criteria. No pathology, outcomes data, or expert consensus on clinical data was used to establish ground truth for this software submission.

8. The sample size for the training set

Not applicable. This is not a machine learning or artificial intelligence device that requires a "training set" in the context of model development.

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

Not applicable, as there was no training set.

Ask a specific question about this device

The VitalConnect Platform is a wireless remote monitoring system intended for use by healthcare professionals for continuous collection of physiological data in home and healthcare settings. This can include heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate, body temperature, activity (including step count), and posture (body position relative to gravity including fall). Data are transmitted wirelessly from the VitalConnect Sensor for storage and analysis. The VitalConnect Platform can include the ability to notify healthcare professionals when physiological data fall outside selected parameters.

The device is intended for use on general care patients who are 18 years of age or older as a general patient monitor, to provide physiological information. The data from the VitalConnect Platform are intended for use by healthcare professionals as an aid to diagnosis and treatment. VitalPatch device can be used on patients with pacemakers that comply with ISO 14117:2012 and ANSVAAMI PC69:2000 without deviations. Heart rate, electrocardiography (ECG), heart rate variability, R-R interval, and respiratory rate are not intended for patients with pacemakers. The device is not intended for use on critical care patients.

VitalPatch is a wearable biosensor designed to measure a patient's vital signs, including heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate, body temperature, skin temperature, activity (including step count), and posture (body position relative to gravity including fall). VitalPatch was most recently cleared in K190916.

There are three device modifications included in this 510(k) pre-submission:

1. removal of the pacemaker contraindication from the VitalPatch instructions for use,
2. heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate are not intended for patients with pacemakers.
3. the patch setup procedure interface was modified to select whether the patient has a pacemaker; the response inhibits as appropriate the functions not intended for patients with pacemaker. If pacemaker is indicated as present, the device will not transmit or store heart rate, electrocardiography (ECG), heart rate variability, R-R interval, and respiratory rate.

The document provided describes a 510(k) premarket notification for the VitalConnect Biosensor, specifically addressing modifications to allow its use with patients who have pacemakers. The primary study presented focuses on ensuring that the device's functionality does not interfere with pacemakers and that certain physiological measurements are appropriately suppressed when a pacemaker is detected.

Here's an analysis of the acceptance criteria and study information based on the provided text:

Acceptance Criteria and Reported Device Performance

The document doesn't present a single, concise table of acceptance criteria and reported device performance in the format requested. Instead, it discusses performance characteristics and verification activities related to specific aspects of the device, particularly regarding its functionality with pacemakers.

Based on the "Performance Data" section, two key areas with implied acceptance criteria and reported performance are:

Table 1: Implied Acceptance Criteria and Reported Performance

Study Information:

The document describes several verification and validation activities rather than a single, comprehensive "study" with a traditional design for evaluating a diagnostic algorithm.

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

• EMI Interference Testing: Not specified.
• Pacemaker Status Check Test: Not specified.
• Heart Rate Performance (non-pacemaker subjects): Not specified.
• Clinical Studies: "Clinical testing was performed in non-pacemaker human subjects to demonstrate substantial equivalence in heart rate and respiratory rate of the subject device." The number of subjects is not specified.
• Data Provenance: Not explicitly stated (e.g., country of origin, retrospective/prospective). The mention of "non-pacemaker human subjects" for clinical studies implies prospective data collection, but details are lacking.

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

Not applicable. The described tests are primarily performance evaluations of device functionality and safety, not diagnostic accuracy requiring expert ground truth for interpretation.

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

Not applicable. This is not a study requiring adjudication of expert interpretations for a test set.

4. If a multi-reader multi-case (MRMC) comparative effectiveness study was done:

No, an MRMC comparative effectiveness study was not done. The document focuses on device performance and safety, particularly concerning pacemaker compatibility, not on human reader performance with or without AI assistance.

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

Yes, the tests described are primarily standalone performance evaluations of the device's firmware and hardware functionalities, independent of human interpretation or assistance.

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

• EMI Interference: Ground truth would be established by regulatory standards (ISO 14117:2012 and ANSI/AAMI PC69:2000) and possibly measurements from calibrated equipment.
• Pacemaker Status Check Test: Ground truth for pacemaker presence would be the actual state of the patient (e.g., presence of a pacemaker) and the expected device response (suppression of certain measurements).
• Heart Rate Performance (non-pacemaker subjects): For heart rate and respiratory rate, ground truth would likely come from a reference standard device or method, although not explicitly stated.
• Battery Current Draw: Ground truth is measured current consumption.

7. The sample size for the training set:

Not applicable. This device is not an AI/ML algorithm that requires a training set in the conventional sense for learning patterns from data to make predictions or classifications. The "firmware" being modified refers to programmed logic, not a learned model.

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

Not applicable, as there is no training set mentioned for an AI/ML algorithm.

Ask a specific question about this device

The VitalConnect Platform is a wireless remote monitoring system intended for use by healthcare professionals for continuous collection of physiological data in home and healthcare settings. This can include heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate, body temperature, activity (including step count), and posture (body position relative to gravity including fall). Data are transmitted wirelessly from the VitalConnect Sensor for storage and analysis. The VitalConnect Platform can include the ability to notify healthcare professionals when physiological data fall outside selected parameters.

The device is intended for use on general care patients who are 18 years of age or older as a general patient monitor, to provide physiological information. The data from the VitalConnect Platform are intended for use by healthcare professionals as an aid to diagnosis and treatment. The device is not intended for use on critical care patients.

VitalPatch is a wearable biosensor designed to measure a patient's vital signs, including heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate, skin temperature, activity (including step count), and posture (body position relative to gravity including fall). VitalPatch was most recently cleared in K183078.

The following modifications are proposed for the VitalPatch wearable biosensors:

• . Increasing the duration for use from 5 days (120 hours) to 7 days (168 hours). This includes a minor software change and minor changes to the IFU.
  All other vital signs, including heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate, body temperature, skin temperature, activity (including step count), and posture (body position relative to gravity including fall) cleared in K183078 remain the same.

Below is the information regarding the acceptance criteria and the study proving the device meets those criteria, extracted from the provided text.

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

   The document does not explicitly present a table of "acceptance criteria" with numerical targets for each physiological measurement (e.g., heart rate accuracy within X BPM). Instead, it states that the "Performance Characteristics" of the proposed device are "Identical" to the predicate device (K183078) for all listed physiological data types and other features. The "Performance Characteristic" table (Table 4) lists the capabilities and ranges for each measured parameter. Therefore, the "acceptance criteria" are implied to be the same as the established performance of the predicate device.

   Performance Characteristic	Predicate (K183078) Performance (Implied Acceptance Criteria)	Proposed Device Performance (Reported)
   Wireless Transmission	Bluetooth Low Energy (BT4.1) technology	Identical
   Data encryption	Advanced Encryption Standard-CCM mode	Identical
   Radio Frequency	2.4GHz ISM band, FCC Part 15 Complied	Identical
   ECG Dynamic Range	Range: -10mV to +10mV	Identical
   Heart Rate (stationary & ambulatory)	Range: 30 - 200 BPM	Identical
   Respiratory Rate	Range: 10-30 breaths per minute	Identical
   Skin Temperature	Range: 15°C - 50 °C (61°F - 113°F)	Identical
   Ambient Temperature	Range: 15°C - 50 °C (61°F - 113°F)	Identical
   Body Temperature	Range: 32°C - 42°C	Identical
   Fall Detection	Range: Fall or No Fall	Identical
   Step Count	Range: 0 - 65535 steps	Identical
   Posture Detection	Postures detected include lying down, upright, walking, running and leaning.	Identical
   R-R Interval	Measurement of the interval from the R wave peak of one QRS complex to the next R-wave peak on the electrocardiogram, enhanced QRS performance during motion.	Identical
   Heart Rate Variability	Quantifies the variation in the beat-to-beat interval time series obtained from the ECG waveform. Detects each QRS peak with high temporal accuracy. Measurement of the R wave to R wave (RR) interval allows short and long-term variability analysis to determine analytics such as mean, median, standard deviation, frequency content, etc.	Identical
   Low Power Mode	When patch is off the body, processor goes into a low power mode, resulting in reduced current draw.	Identical

                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                              |                                        |
   

   | New Feature/Modification | Predicate (K183078) Performance | Proposed Device Performance |
   | Duration of Use | 5 days (120 hours) | 7 days (168 hours) |

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

   • The document states that substantial equivalence is based upon "benchtop performance testing" and lists several tests: "Wear Duration Testing", "Battery operating life testing", and "EOL Detection after 168 hours".
   • The sample size for these tests is not specified in the provided text.
   • The data provenance is not specified (e.g., country of origin, retrospective or prospective).

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

   • This information is not applicable as the provided documentation describes benchtop performance testing, not a study involving human interpretation or ground truth established by experts for a test set.

4. Adjudication method for the test set

   • This information is not applicable as the provided documentation describes benchtop performance testing, not a study requiring adjudication of expert interpretations.

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

   • An MRMC comparative effectiveness study was not done. The study described is a benchtop performance validation for an extended wear duration, not an AI-assisted diagnostic device.

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

   • The document refers to "benchtop performance testing" which is essentially a standalone evaluation of the device's technical specifications and extended wear duration. It is not an "algorithm only" performance study in the context of AI, but rather a direct measurement of the device's physical and functional performance (e.g., battery life, EOL detection).

7. The type of ground truth used

   • For the benchtop tests (Wear Duration, Battery operating life, EOL Detection), the ground truth would be objective pre-defined technical specifications and direct measurements (e.g., a timer measuring 168 hours, battery voltage measurements, sensor output at end-of-life).

8. The sample size for the training set

   • This information is not applicable. The device (VitalPatch Biosensor) measures physiological parameters through sensors and firmware. It's not an AI/ML device that requires a training set in the typical sense for classification or prediction from complex data (like images). The firmware updates are minor changes to support extended duration and bug fixes.

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

   • This information is not applicable for the reasons stated above. The device is not based on machine learning from a training set with established ground truth.

Ask a specific question about this device

The VitalConnect Platform is a wireless remote monitoring system intended for use by healthcare professionals for continuous collection of physiological data in home and healthcare settings. This can include heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate, body temperature, activity (including step count), and posture (body position relative to gravity including fall). Data are transmitted wirelessly from the VitalConnect Biosensor for storage and analysis. The VitalConnect Platform can include the ability to notify healthcare professionals when physiological data fall outside selected parameters.

The device is intended for use on general care patients who are 18 years of age or older as a general patient monitor, to provide physiological information. The data from the VitalConnect Platform are intended for use by healthcare professionals as an aid to diagnosis and treatment. The device is not intended for use on critical care patients.

VitalPatch is a wearable biosensor designed to measure a patient's vital signs, including heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate, skin temperature, activity (including step count), and posture (body position relative to gravity including fall). VitalPatch was most recently cleared in K163453.

The following modifications are proposed for the VitalPatch wearable biosensors:

• . An ambient temperature sensor is added to the original VitalPatch, which was most recently cleared in K163453, to measure ambient temperature. The measured ambient temperature along with the measured skin temperature cleared in K163453 are used to calculate the body temperature.
• . An updated IFU for VitalPatch has been created to reflect the temperature sensor changes.

All other vital signs, including heart rate, electrocardiography (ECG), heart rate variability, R-R interval, respiratory rate, skin temperature, activity (including step count), and posture (body position relative to gravity including fall) cleared in K163453 remain the same.

The provided document is a 510(k) summary for the VitalConnect Platform and VitalPatch Biosensor. It details the device's characteristics, indications for use, and comparison to a predicate device (K163453). However, it does not contain the specific acceptance criteria or a detailed study proving the device meets those criteria, especially in the context of an AI/human-in-the-loop system as implied by some of the questions.

The document states that the new VitalPatch Biosensor adds an ambient temperature sensor to calculate body temperature. It focuses on demonstrating substantial equivalence to the previously cleared predicate device for general physiological monitoring (heart rate, ECG, respiratory rate, skin temperature, activity, posture, fall detection).

Here's an analysis based on the available information and a direct answer to your questions, noting what cannot be answered from the provided text:

Key Takeaway: This document describes a medical device clearance related to a hardware modification (adding an ambient temperature sensor) and a derived vital sign (body temperature). It explicitly states that clinical studies were not required for this clearance, and the substantial equivalence was based on benchtop performance testing. Therefore, many of the questions related to MRMC studies, expert consensus, and large test/training datasets for AI models are not applicable or answerable from this specific document.

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

The document summarizes the performance characteristics of the proposed device by comparing them to the predicate device in Table 4. It treats "Identical" as meeting the acceptance criteria set by the predicate. For new features, it lists the range.

For "Body Temperature," the criteria are implied to be that the proposed device can derive this value within the specified range (32°C - 42°C). The document confirms that "Paired readings, using VitalPatch and hand-held oral thermometers simultaneously, were analyzed to verify accuracy when compared to the hand-held oral thermometers." This suggests the acceptance criterion for body temperature derivation would be accuracy against a gold standard (oral thermometer), though the specific accuracy metric (e.g., mean absolute difference, standard deviation of differences, clinical equivalence) and its threshold are not provided in this summary.

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

The document primarily relies on benchtop and simulated use testing for the new modifications. It explicitly states: "Clinical testing was not required to demonstrate the safety and effectiveness of the VitalPatch. Instead, substantial equivalence is based upon benchtop performance testing."

For the new "Body Temperature" calculation: "Paired readings, using VitalPatch and hand-held oral thermometers simultaneously, were analyzed to verify accuracy when compared to the hand-held oral thermometers."

• Sample Size: Not specified for the paired readings.
• Data Provenance: Not specified (e.g., country). It was likely prospective, as it involved simultaneous measurements.

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 is not applicable in the context of this 510(k) clearance document. The ground truth for the new body temperature parameter was established via a hand-held oral thermometer, which serves as a widely accepted clinical standard for temperature measurement. There is no mention of human experts interpreting physiological waveforms or images to establish ground truth.

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

Not applicable. This type of adjudication is typically used in studies where human readers are interpreting complex medical data (like images) and consensus is needed. The performance assessment here involved direct comparison to an established measurement device (oral thermometer) and verification of engineering specifications.

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, an MRMC comparative effectiveness study was not done. The device described is a physiological data sensor, not an AI-assisted diagnostic tool that would typically involve human readers. The clearance was based on substantial equivalence and benchtop testing, not on improving human reader performance.

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

For the new body temperature feature, the "algorithm only" performance was evaluated by comparing the VitalPatch's calculated body temperature to that of a hand-held oral thermometer. The exact methodology and results (e.g., specific accuracy metrics) are not detailed beyond "analyzed to verify accuracy."

For the other parameters (heart rate, respiratory rate, etc.), they remain "Identical" to the predicate, meaning their standalone performance was previously established and is considered sufficient.

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

• For the new "Body Temperature" calculation, the ground truth was "paired readings, using VitalPatch and hand-held oral thermometers simultaneously." This represents a clinical measurement reference standard.
• For other physiological parameters, the ground truth would have been established during the clearance of the predicate device (K163453) using appropriate reference methods for each parameter (e.g., ECG for heart rate, spirometry for respiratory rate, etc.).

8. The sample size for the training set

This document does not describe the development or training of a complex AI model requiring a large training set. The "firmware" update (Version 3.1.0.1) mentioned for the new device includes "Body temperature derivation," which implies a calculation or algorithm rather than a machine learning model that undergoes "training." Therefore, the concept of a "training set" in the context of machine learning is not applicable here.

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

As there's no "training set" for an AI model described, this question is not applicable. The body temperature derivation is listed as a "calculated value as a function of the skin and ambient temperatures," indicating a deterministic algorithm rather than a learned model.

Ask a specific question about this device

VitalWatch is a software user intended for use by healthcare professionals to display physiological data collected by the VitalConnect Platform wireless remote monitoring system in home and healthcare is a secondary, adjunct patient monitor and is not intended to replace existing standard-of-care patient monitoring practices.

VitalWatch is a software user interface designed to be manufacturer-installed and user-executed on a validated Relay device, which serves as a graphical user interface that displays physiological data provided by the VitalConnect wireless remote monitoring system.

The provided text describes the 510(k) premarket notification for the VitalWatch Software User Interface. However, it does not include detailed information about acceptance criteria or a study proving the device meets those criteria, as typically found in clinical validation studies for AI/ML medical devices.

The document is a regulatory submission for a software user interface (VitalWatch) that displays physiological data collected by an existing wireless remote monitoring system (VitalConnect Platform). The core of this submission revolves around demonstrating substantial equivalence to a predicate device, not necessarily a new clinical performance study.

Therefore, many of the requested details about acceptance criteria, test set characteristics, expert involvement, and ground truth establishment are not present in this regulatory document, as it pertains to a user interface rather than an algorithm performing a diagnostic task.

Here's what can be extracted and what is missing:

Acceptance Criteria and Device Performance (Based on provided text, limited information)

Since this submission focuses on "substantial equivalence" for a user interface, rather than a new diagnostic algorithm, explicit performance metrics and acceptance criteria for clinical efficacy (e.g., sensitivity, specificity, accuracy) are not detailed. The performance discussed is related to software verification and validation.

Study Details:

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

  • See the table above. The focus is on software quality and functional equivalence, not clinical performance metrics like diagnostic accuracy.

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

  • Not specified. The document mentions "software verification and validation testing," but it does not describe a clinical test set with patient data for performance evaluation in the context of an AI/ML algorithm. This is a user interface for displaying pre-existing data.

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

  • Not applicable/Not specified. Ground truth establishment by experts is typically relevant for diagnostic AI algorithms. For a user interface displaying data, the "ground truth" would be the accurate transmission and display of the physiological data collected by the established VitalConnect platform, which is presumed to be accurate from its own prior clearance (K152139).

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

  • Not applicable/Not specified. Adjudication is used to resolve discrepancies in expert readings for diagnostic or screening tasks. This is not a diagnostic device.

• 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. This is not an AI-assisted diagnostic device. It is a user interface for displaying data.

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

  • Not applicable. This device is a user interface, part of a system where a human professional interprets the displayed data. It's not a standalone diagnostic algorithm.

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

  • Not applicable/Not specified in detail. For a user interface, the "ground truth" would be the integrity and accuracy of the data displayed against the raw data from the sensors. The document relies on the prior clearance of the VitalConnect Platform and the software's functional verification.

• 8. The sample size for the training set:

  • Not applicable. This is not a machine learning model that requires a training set.

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

  • Not applicable. This is not a machine learning model.

Summary:

The provided document (K170973) is a 510(k) submission for a software user interface (VitalWatch), not a new diagnostic or AI/ML algorithm. Its clearance relies on demonstrating substantial equivalence to a previously cleared predicate device (VitalConnect Platform, K152139) by showing that it has the same intended use and technological characteristics, and that software verification and validation protocols have been followed. Therefore, the details requested about clinical validation studies, expert involvement, and ground truth establishment for diagnostic performance are largely not applicable or not provided in this specific type of regulatory submission.

Ask a specific question about this device