Bunkerhill Contrast AVC

Bunkerhill Contrast AVC is a software device that analyzes non-gated, contrast-enhanced chest CT scans to automatically detect, localize, and quantify aortic valve calcification in patients aged 40 and above. The device outputs a calcification volume score in cubic millimeters and a visual region of interest to assist physicians in reviewing patient cases; results are intended as adjunctive information only and do not replace the original CT scan or clinical judgment.

Both the subject device and predicate use deep-learning convolutional neural networks to detect calcifications on CT volumes with slice thickness up to 5 mm. The key difference is that the subject device analyzes contrast-enhanced non-gated CT images whereas the predicate analyzes non-contrast non-gated images; the subject device outputs volume in mm³ while the predicate outputs Agatston Units, both using deterministic mathematical formulas on segmentation maps.

Not stated in this summary.

Bunkerhill Contrast AVC is substantially equivalent to the predicate Bunkerhill AVC (K243229) because both devices share identical principles of operation, the same intended use of detecting and quantifying aortic valve calcification, and similar technological characteristics. A pivotal retrospective study and bridging study demonstrated the subject device met co-primary endpoints for sensitivity, specificity, precision, and recall with acceptable limits of agreement comparable to the predicate, establishing that minor differences in imaging modality (contrast-enhanced versus non-contrast) do not raise new safety or effectiveness concerns.