AVIEW CAC

AVIEW CAC is a software application that performs quantitative analysis of coronary artery calcification from non-contrast, non-gated chest CT scans by employing deep learning for automated artery segmentation and Agatston score calculation. It is intended to assist healthcare professionals in risk stratification by providing calcium scores and percentile-based risk categories, with current validated use limited to CT scans from General Electric equipment.

AVIEW CAC is a medical image processing software designed for healthcare professionals, such as radiologists and cardiologists, to conduct quantitative analysis of calcified plaques within the coronary arteries. It utilizes non-contrast, non-gated Chest CT scans from adult patients aged 40 and older. The system primarily calculates the Agatston score for coronary artery calcification by segmenting and evaluating the right and left coronary arteries. Furthermore, it offers risk stratification based on the calculated calcium score, patient gender, and age, referencing established medical guidelines. This tool aims to support clinical assessment by providing quantitative data derived from existing CT studies. The software operates on a personal computer and employs the DICOM 3.0 standard for managing medical images, including functions for reading, manipulating, analyzing, post-processing, storing, and transferring data. It facilitates access through mobile devices and Chrome browsers for on-premises data sets. The core technology involves automatic segmentation of coronary arteries and calcified regions, reportedly using deep learning algorithms, to determine the Agatston score. Performance evaluations, based on non-clinical testing with 150 non-gated chest CT cases, demonstrated high intraclass correlation coefficients (ICC) for Agatston scores when compared to ground truth (Total ICC: 0.896, LCA: 0.927, RCA: 0.840) and a predicate device. However, the device's use is specifically validated for CT scans acquired using General Electric (GE) equipment.

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. How does the deep learning algorithm specifically identify and segment calcified plaques in non-gated chest CT scans, and what is its performance compared to manual segmentation or other automated methods?; What are the specific sensitivity, specificity, and accuracy of AVIEW CAC for detecting various Agatston score categories (e.g., 0, 1-10, 11-100, 101-400, >400) on non-contrast, non-gated chest CTs?; Given the device's limitation to GE CT scanners, what are the technical challenges and validation requirements for expanding its use to CT scanners from other manufacturers?; How does AVIEW CAC's performance in Agatston scoring vary with different CT acquisition parameters (e.g., slice thickness, radiation dose) within the GE scanner ecosystem?; What impact does the use of AVIEW CAC for opportunistic screening on non-gated chest CTs have on downstream clinical decision-making and patient outcomes compared to traditional cardiac-gated CT calcium scoring?