Search Results

The AV Cardiac CT applications are intended to assist the user in viewing, processing, analysis of CT datasets and in preparation of cardiac interventions.
The CT Coronary Analysis application is indicated to assist radiologists, cardiologists, and 3D technologists in the analysis of coronary artery anatomy for patients with suspected or diagnosed cardiac disease including coronary artery disease.
The CT Functional Analysis application is indicated to assist radiologists, cardiologists, and 3D technologists in the analysis of heart anatomy and function, for patients with suspected or diagnosed cardiac diseases.

AV Cardiac CT is a standalone, software‑only medical device intended for advanced visualization, post‑processing, and analysis of cardiac computed tomography (CT) datasets. The software assists qualified medical professionals in viewing and analyzing gated cardiac CT images to support diagnosis and follow‑up, as well as review of information in the context of preparation for cardiac interventions.

AV Cardiac CT includes CT Coronary Analysis and CT Functional Analysis applications. The device is intended for use in the analysis of coronary artery anatomy and cardiac function in patients with suspected or diagnosed coronary artery disease, using gated cardiac CT data, including conventional and spectral acquisitions.

The device provides visualization, measurement, and post‑processing tools that enable review of quantitative information derived from CT images. AV Cardiac CT does not perform automated diagnosis, clinical decision‑making, or treatment recommendations. All measurements and visualizations are intended to support physician review; the final clinical interpretation and procedural decisions remain the responsibility of the qualified healthcare professional.

As part of the CT Coronary Analysis application, AV Cardiac CT provides a dedicated planning workflow - planning module - that organizes existing visualization and measurement tools—such as vessel length, diameter, and approximate C-arm angulation—to support pre-procedural review, such as for percutaneous coronary intervention (PCI). The workflow enables users to review relevant findings with corresponding anatomical images, determine proximal and distal landing zones, and save or export measurement values. The planning workflow does not automate device selection or treatment decisions and is intended to assist review and preparation, with physician judgment retained.

The provided FDA 510(k) clearance letter describes the acceptance criteria and study data for the Philips Medical Systems' AV Cardiac CT device, focusing on three key algorithms: Coronary Artery Centerline Extraction, Major Vessel Labeling, and Lumen Segmentation.

Here's a breakdown of the requested information:

1. Table of Acceptance Criteria and Reported Device Performance

Coronary Artery Centerline Extraction

Major Vessel Labeling

Lumen Segmentation

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

Coronary Artery Centerline Extraction

• Sample Size: 80 patients
• Data Provenance: Retrospective data collected from:
  • Eight clinical sites in the US (51 patients, 63.75%)
  • Three hospitals in Europe (19 patients, 38%)
  • Two hospitals in Asia (10 patients, 12.5%)
• Patient Characteristics: Patients with suspected or confirmed coronary artery disease, encompassing various stenosis degrees, calcification scores, and plaque presence.

Major Vessel Labeling

• Sample Size: Not explicitly stated as a separate sample size. It can be inferred that the same 80 patients from the Coronary Artery Centerline Extraction study were likely used for validating major vessel labeling, as it is a closely related function.

Lumen Segmentation

• Sample Size: 80 patients
• Data Provenance: Retrospective data collected from:
  • Seven clinical sites in the US (68 patients, 85%)
  • Three hospitals in Europe (12 patients, 15%)
• Patient Characteristics: Patients with suspected or confirmed coronary artery disease, encompassing various stenosis degrees, calcification scores, and plaque presence.

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

Coronary Artery Centerline Extraction

• Number of Experts: 3 observers (1 US-board certified radiologist and 2 US-board certified cardiologists).
• Qualifications: US-board certified radiologist and US-board certified cardiologists.

Major Vessel Labeling

• Number of Experts: 3 US-board certified observers.
• Qualifications: US-board certified observers (implied to be radiologists/cardiologists given the context of cardiac imaging).

Lumen Segmentation

• Number of Experts: 3 US-board certified observers.
• Qualifications: US-board certified observers (radiologists and cardiologists).

4. Adjudication Method for the Test Set

Coronary Artery Centerline Extraction

• Adjudication Method: "per coronary artery, the three observer-annotated centerlines were averaged as the reference standard." This indicates a consensus-based approach by averaging.

Major Vessel Labeling

• Adjudication Method: "reference labels established by consensus among three US-board certified observers." This indicates a qualitative consensus method.

Lumen Segmentation

• Adjudication Method: "for each coronary artery, the three observer-annotated lumen contours were averaged as the reference standard." This indicates a consensus-based approach by averaging.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done

The provided document does not indicate that a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was done to compare human readers with and without AI assistance. The studies performed were primarily focused on the standalone performance of the algorithms against expert-established ground truth.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

Yes, standalone performance studies were done for the automated coronary artery centerline extraction, major vessel labeling, and coronary artery lumen segmentation algorithms. The performance statistics provided (OV, OT, AvgDist, Sensitivity, Precision, DSC, MSD, HD) directly reflect the algorithm's performance without human intervention in the analysis process on the test set. The document explicitly states that "All algorithm outputs are reviewable and editable by the user, and the device does not perform autonomous clinical interpretation," but the validation shown is for the automated output's accuracy against ground truth.

7. The Type of Ground Truth Used

• Coronary Artery Centerline Extraction: Expert consensus (average of three independent expert annotations).
• Major Vessel Labeling: Expert consensus (consensus among three observers).
• Lumen Segmentation: Expert consensus (average of three independent expert annotations).

8. The Sample Size for the Training Set

The document states, "The algorithm model was developed with training data collected from completely distinct clinical sites, ensuring independence between training and testing data" but does not specify the sample size for the training set.

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

The document states, "The algorithm model was developed with training data collected from completely distinct clinical sites" but does not describe how the ground truth for the training set was established.

Ask a specific question about this device