Search Results

The intended use of the CARTO™ 3 System is catheter-based cardiac electrophysiological (EP) procedures. The CARTOTM 3 System provides information about the electrical activity of the heart and about catheter location the procedure. The system can be used on patients who are eligible for a conventional electrophysiological procedure.

The system has no special contraindications.

The CARTOTM 3 EP Navigation System V8.0, is a catheter-based atrial and ventricular mapping system designed to acquire and analyze navigation catheter's location and intracardiac ECG signals and use this information to display 3D anatomical and electroanatomical maps of the human heart. The location information needed to create the cardiac maps and the local electrograms are acquired using specialized mapping catheters and reference devices. The CARTOTM 3 System uses two distinct types of location technology - magnetic sensor technology and Advanced Catheter Location (ACL) technology.

The CARTOTM 3 System V8.0 consists of the following hardware components:

• Patient Interface Unit (PIU)
• Workstation with Graphic User Interface (GUI)
• Wide-Screen monitors, keyboard, and mouse
• Intracardiac In Port
• Intracardiac Out Port
• Power Supply
• Patches Connection Box and Cables (PU)
• Pedals
• Location Pad (LP)
• Signal Processing Unit (SPU)

All hardware components of the CARTO™ 3 system V8.0 are the same as those found in the predicate device.

The provided document describes the CARTO™ 3 EP Navigation System Software V8.0 and its modifications. It details the various tests performed (software verification and validation, bench testing, animal testing, and clinical testing) to demonstrate the device's safety and effectiveness and its substantial equivalence to the predicate device. However, the document does not provide specific acceptance criteria or detailed study results in the format requested. It generally states that "All testing performed were successfully completed and met the acceptance criteria" and "All system features were found to perform according to specifications and met the tests acceptance criteria" without listing the metrics or the numerical outcomes.

Therefore, for aspects like "Table of acceptance criteria and the reported device performance," "Effect size of how much human readers improve with AI vs without AI assistance," and specific numerical results for performance, the information is not present in the provided text.

Based on the available text, here's a breakdown of what can be extracted:

Device: CARTO™ 3 EP Navigation System Software V8.0

No AI/ML Specific Performance Study (MRMC or Standalone) provided in the document for the new AI/ML features that show a quantitative acceptance criteria and the actual performance. The document broadly states that tests met acceptance criteria and performs as expected.

1. Table of Acceptance Criteria and the Reported Device Performance:

The document broadly states that all tests successfully met acceptance criteria without providing specific performance metrics or numerical thresholds for acceptance. It is a general statement of compliance.

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

• Complex Signals Identification (CSI) module:
  • Data Provenance: Medical Centers in Europe and the United States of America.
  • Data Type: Control database created using input from multiple stable atrial typical and atypical flutter cases. Included data from different catheters, with complex signals, single activations, noisy signals, and so on.
  • Sample Size: Not specified.
  • Retrospective/Prospective: "Retrospective Validation Tests Testing was performed retrospectively, on clinical recorded data from historic EP procedures performed with the CARTOTM 3 System." This applies to all new modules including CSI.
• CARTOSOUND™ FAM Module:
  • Data Provenance: Diverse locations and populations.
  • Data Type: Collected using a variety of ultrasound system settings, scanners, and Catheters.
  • Sample Size: Not specified.
  • Retrospective/Prospective: "Retrospective Validation Tests Testing was performed retrospectively, on clinical recorded data from historic EP procedures performed with the CARTOTM 3 System." This applies to all new modules including CARTOSOUND™ FAM.
• Clinical Testing (REAL AF Registry Sub-Study):
  • Sample Size: The "REAL AF Registry population" is mentioned, but the specific number of patients is not provided.
  • Data Provenance: Not explicitly stated beyond "REAL AF Registry Sub-Study."
  • Retrospective/Prospective: Not explicitly stated, but "Registry Sub-Study" often implies a prospective or retrospective observational study on existing patient data.

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

• For the Complex Signals Identification (CSI) module, it states that "a control database was created using input from multiple stable atrial typical and atypical flutter cases." It does not specify the number or qualifications of "experts" if involvement extends beyond clinical data collection.
• For the CARTOSOUND™ FAM Module, no information provided about experts involved in ground truth establishment for the test set.
• For the Clinical Testing (REAL AF Registry Sub-Study), the ground truth is clinical outcomes and safety endpoints, which are typically established by clinical professionals as part of the study protocol, but specific numbers or qualifications of adjudicators are not provided.

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

• No specific adjudication method for establishing ground truth is described for any of the modules or studies.

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:

• The document does not mention a multi-reader multi-case (MRMC) comparative effectiveness study to evaluate human reader improvement with AI assistance. The AI/ML algorithms are described as components of the system to aid specific tasks (e.g., automatically tagging fractionated IC ECG signals, reconstructing 3D volumes) rather than an AI-assisted interpretation by human readers.

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

• The document indicates that the "Complex Signals Identification (CSI) module is a machine learning locked algorithm that automatically tags fractionated IC ECG signals." Similarly, "The CARTOSOUND™ FAM Module incorporates a Deep Learning (DL) algorithm" for automatic reconstruction and segmentation.
• "Retrospective Validation Tests Testing was performed retrospectively, on clinical recorded data from historic EP procedures" for newly introduced modules. This suggests standalone performance evaluation on historic data, but specific performance metrics (e.g., accuracy, sensitivity, specificity) for these algorithms are not provided in the document. The document only states that "All testing performed were successfully completed and met the acceptance criteria."

7. The type of ground truth used:

• Complex Signals Identification (CSI) module: "Control database created using input from multiple stable atrial typical and atypical flutter cases." This implies clinical ground truth based on real patient data.
• CARTOSOUND™ FAM Module: Training and validation data collected from diverse clinical settings with ultrasound systems. Ground truth for 3D segmentation would likely be based on expert manual segmentation or a consensus method, though not explicitly stated.
• Clinical Testing (REAL AF Registry Sub-Study): Clinical outcomes (safety and acute effectiveness endpoints) based on patient data.

8. The sample size for the training set:

• Complex Signals Identification (CSI) module: Not specified.
• CARTOSOUND™ FAM Module: Not specified, only mentioned that "Data for training and validation of the DL algorithm was collected from diverse locations and populations."

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

• Complex Signals Identification (CSI) module: "a control database was created using input from multiple stable atrial typical and atypical flutter cases." This suggests the ground truth was derived from the characteristics of these cases, likely based on established clinical criteria or expert review of the EP data. The exact method (e.g., expert consensus, independent review) is not detailed.
• CARTOSOUND™ FAM Module: Not specified. For 3D segmentation, ground truth is typically established by manual annotation by trained professionals or a consensus of multiple annotators, but this is not confirmed in the document.

In summary, the provided FDA 510(k) summary focuses on demonstrating substantial equivalence through various V&V activities and broad statements of meeting acceptance criteria. It introduces new AI/ML components but does not delve into detailed quantitative performance metrics for these AI components or comparative studies with human readers.

Ask a specific question about this device