Rapid Aneurysm Triage and Notification

The Rapid Aneurysm Triage and Notification (ANRTN) device is a radiological software utilizing an artificial intelligence algorithm to analyze head CTA images for identifying and prioritizing suspected saccular aneurysms of at least 4mm in adults. Its methodology involves processing DICOM CTA data within the Rapid Platform to generate non-diagnostic notifications and preview images, aiming to assist clinicians in efficient workflow management for potential aneurysm cases, without altering original images or providing diagnoses.

iSchemaView's Rapid Aneurysm Triage and Notification (ANRTN) is a radiological computer-assisted software designed to enhance workflow triage for medical professionals, particularly radiologists, by analyzing CT Angiography (CTA) images of the head. The device's primary purpose is to flag and prioritize studies indicating suspected saccular aneurysms, specifically those 4mm in diameter or larger, in adult patients during routine care. It integrates into the Rapid Platform and processes CTA images provided in DICOM format, assisting hospital networks and trained radiologists in managing their caseload more effectively by drawing attention to potentially critical findings. The ANRTN software operates in parallel with the standard of care and is intended for informational purposes to aid triage, not for direct diagnosis or to replace a full patient evaluation. The ANRTN system employs an artificial intelligence and machine learning (AI/ML) algorithm to analyze the input CTA images and identify potential saccular aneurysms within specific anatomical regions of the brain's vasculature, such as the Internal Carotid, Middle Cerebral, and Anterior Cerebral Arteries. Upon detection, it provides notifications and generates non-diagnostic, compressed preview images accessible via PACS, email, or a mobile application, which are solely for informational review and triage prioritization; the original medical images remain unaltered. Performance validation on 266 CTA cases, with ground truth established by three experts, demonstrated an Area Under the Curve (AUC) of ≥ 0.95, an overall sensitivity of 0.933, and specificity of 0.868. The device has defined inclusion criteria for CTA datasets it can process, related to image acquisition parameters, and exclusion criteria for cases with significant motion or imaging artifacts.

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 AI/ML algorithm's performance vary with CTA acquisition protocols or image quality variations beyond the specified inclusion/exclusion criteria, such as contrast timing or patient motion artifacts?; What is the clinical impact of the 4mm minimum detection size threshold on potentially missing smaller, yet clinically relevant, saccular aneurysms that might benefit from earlier monitoring or intervention?; How does the device's performance specifically compare in detecting aneurysms in complex anatomical locations or in the presence of co-existing vascular pathologies that might mimic or obscure aneurysms?; What are the key image features or biomarkers the AI/ML model has learned to prioritize for saccular aneurysm detection, and how robust is this detection against various imaging artifacts or anatomical variants?; Beyond sensitivity and specificity, what is the quantified impact of Rapid ANRTN on radiologist workflow efficiency, such as mean time-to-interpretation for positive cases or reduction in missed findings, in a real-world clinical setting?