Search Results

Bunkerhill Contrast AVC is a software device intended for use in detecting presence and estimating quantity of aortic valve calcification for adult patients aged 40 years and above. The device automatically analyzes non-gated, contrast-enhanced chest computed tomography (CT) images collected during clinical care and outputs the region of interest (intended for informational purposes only) and quantification of detected calcium.

The output of the subject device is made available to the physician on-demand as part of his or her standard workflow. The device-generated quantification can be viewed in the patient report at the discretion of the physician, and the physician also has the option of viewing the device- generated calcium region of interest in a diagnostic image viewer. The subject device output in no way replaces the original patient report or the original non-gated, contrast-enhanced CT scan; both are still available to be viewed and used at the discretion of the physician.

The device is intended to provide information to the physician to provide assistance during review of the patient's case. Results of the subject device are not intended to be used on a stand- alone basis and are solely intended to aid and provide information to the physician. In all cases, further action taken on a patient should only come at the recommendation of the physician after further reviewing the patient's results.

Bunkerhill Contrast AVC is a software as a medical device (SaMD) product that interfaces with compatible and commercially available computed tomography (CT) systems. Bunkerhill Contrast AVC detects localizes, and quantifies aortic valve calcification in non-gated, contrast-enhanced chest CT studies. The core features of the product are:

• Detection of aortic valve calcification volume at a threshold of 0 mm³.
• Quantification of the overall aortic valve calcification burden in the form of an estimated volume score (mm³).
• Localization of estimated calcium burden in the form of AVC region of interest applied to a copy of the original CT scan.

The device integrates into the clinician's PACS and does not include a built-in viewer. It works in parallel with and runs in the background of the physician's workflow. When chest CT scans are captured, the images are automatically sent via the DICOM protocol to an on-premises server ("Bunkerhill Edge Server") running the Bunkerhill software. Images are de-identified on the Bunkerhill Edge Server and sent to a Bunkerhill Cloud Server for processing, where the artificial intelligence algorithm (Bunkerhill Contrast AVC) is applied to the image to detect, estimate, and localize aortic valve calcium. The results from the algorithm are then relayed back to the on-premises Bunkerhill Edge Server, where the associated chest CT images are re-identified, and results are paired with the appropriate images and routed to be made available in the physician's workflow.

Here's a breakdown of the acceptance criteria and study details for the Bunkerhill Contrast AVC device, based on the provided FDA 510(k) Clearance Letter:

Bunkerhill Contrast AVC Study Analysis

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample Size and Data Provenance for Test Set

• Pivotal Study:
  • Sample Size: Not explicitly stated. The text mentions the dataset was "curated from multiple sites across three geographical regions in the United States."
  • Data Provenance: Multiple sites across three geographical regions in the United States. Retrospective.
• Bridging Study:
  • Sample Size: 65 cases.
  • Data Provenance: Not explicitly stated, but "comparing the volume generated by the Contrast AVC device on a contrast-enhanced image to the calcification volume on the same patient calculated on a non-contrast image." Implies retrospective data from patients who underwent both types of scans.

3. Number of Experts and Qualifications for Ground Truth

• Number of Experts: Not specified.
• Qualifications of Experts: Not specified. The acceptance criteria for the pivotal study's agreement metrics (bias, LOA) were derived from "clinical literature in high impact journals that investigate the inter-reader agreement of manual segmentation," implying expert-based ground truth.

4. Adjudication Method for Test Set

• Adjudication Method: Not explicitly stated. The reliance on "inter-reader agreement of manual segmentation" from clinical literature suggests that the ground truth methodology likely involved multiple readers, but the specific adjudication method (e.g., 2+1, 3+1) is not detailed.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• Was an MRMC study done? No. The studies described are standalone performance evaluations of the algorithm, not comparative effectiveness studies with or without AI assistance for human readers.

6. Standalone Performance

• Was a standalone study done? Yes. Both the pivotal study and the bridging study evaluated the "Contrast AVC device" performance in a "stand-alone retrospective study" for "detection, localization and agreement of the device output compared to the established ground truth."

7. Type of Ground Truth Used

• Pivotal Study: Implied to be expert consensus or manual segmentation by experts, given the reference to "inter-reader agreement of manual segmentation" in clinical literature.
• Bridging Study: Ground truth was established by "calcification volume on the same patient calculated on a non-contrast image" — which itself would have been established by a ground truth methodology (likely expert manual segmentation) for the non-contrast images.

8. Sample Size for Training Set

• Sample Size: Not provided in the document.

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

• Ground Truth Establishment: Not provided in the document. However, for a deep learning model, it would typically involve expert manual annotation or segmentation of calcifications on a large dataset of CT images.

Ask a specific question about this device