Bunkerhill AVC

The Bunkerhill AVC is a deep-learning powered software device designed to automatically detect and quantify aortic valve calcification from non-gated, non-contrast chest CT scans. Its scientific basis relies on advanced image analysis algorithms validated to provide accurate calcium localization and quantification, serving as an adjunctive tool for clinicians in patient assessment.

BunkerHill Health's Bunkerhill AVC is a software device cleared for its intended use, designed to automatically detect and quantify aortic valve calcification (AVC) in adult patients aged 40 and above using non-gated, non-contrast chest computed tomography (CT) images. It provides physicians with crucial informational output, including a region of interest and quantification of detected calcium, serving as an adjunctive tool to assist in patient case review without replacing original patient reports or CT scans.As a Class II medical device, Bunkerhill AVC received substantial equivalence approval, permitting its market entry contingent on compliance with general and additional controls such as annual registration, good manufacturing practices, and the Quality System (QS) regulation. The device leverages deep-learning algorithms, akin to its predicate iCAC Device, for image analysis, with its safety and performance validated through a retrospective study demonstrating effective localization and agreement with established criteria.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. What is the device's inter-scanner and intra-scanner variability when quantifying aortic valve calcification across different CT machine models and acquisition protocols?; How does the accuracy of Bunkerhill AVC compare against gold-standard methods (e.g., echocardiography, gated CT) for detecting and quantifying clinically significant AVC?; Are there specific patient populations or image characteristics (e.g., severe motion artifact, extensive diffuse calcification) where the device's performance may be compromised?; What is the long-term clinical utility and impact of using Bunkerhill AVC's output on patient management decisions or cardiovascular risk stratification?; Has the deep-learning algorithm been validated on diverse datasets representative of various demographics and disease severities to ensure generalizability?