HeartFlow has announced it has submitted a 510(k) premarket application to the US Food and Drug Administration (FDA) to add advanced anatomic assessment and plaque evaluation to the HeartFlow FFRct Analysis.
The PreRead anatomic assessment will help identify the presence and location of narrowings or stenoses in the coronary arteries based on coronary computed tomography (CT) scans. The HeartFlow Plaque technology will provide plaque volume and characterise the type of plaque present. By adding the anatomic assessment and plaque evaluation to the physiological information currently provided by the HeartFlow Analysis, physicians will gain a more comprehensive understanding of a patient’s coronary disease burden and support efficient risk stratification of patients who may be at high risk of death from a heart attack, the company said in a press release.
“HeartFlow is committed to becoming an indispensable partner to physicians in delivering precision heart care by providing a broad portfolio of game-changing innovations that leverages our core deep learning technology,” said John H Stevens, president, CEO and co-founder, HeartFlow. “The FDA submission is an important step towards delivering a solution that we believe supports complete cardiovascular care by providing clinicians with advanced insights about anatomy, physiology and plaque for their patients with coronary artery disease.”
PreRead is designed to provide a rich anatomic assessment of each coronary CT that, when considered by physicians in conjunction with other patient clinical information, helps physicians to quickly identify where coronary disease is present and understand the severity of the narrowing.
The PreRead assessment highlights areas of modelled stenosis greater than 30% for vessels 1.8mm or larger in diameter. This information will be provided as a 2D image of the coronary arteries, with areas of concern colour coded by severity. In addition, CT images of each of the three main arteries are provided so physicians can more easily confirm the PreRead color-coded narrowings.
The PreRead assessment is intended to support clinical workflow efficiency, diagnostic accuracy and CT reader consistency as measured by repeatability and reproducibility. In an internal study, the PreRead assessment was 35% more repeatable and 47% more reproducible in classifying percent stenosis than expert CT readers’ (Level 3) classification of the same cases.
In addition, the HeartFlow Plaque feature is based on a fully automated deep-learning (a form of AI) algorithm for characterising and quantifying plaque. In an internal study, the HeartFlow Plaque technology was found to be more reliable than expert CT readers in identifying different types of plaque and quantifying total plaque volume.