BriefCase

BriefCase is an AI-powered software that analyzes CTPA scans from adult patients to identify suspected Pulmonary Embolism, aiming to expedite triage by notifying clinicians of critical findings. Its methodology relies on a deep learning algorithm to process DICOM images and flag urgent cases for prioritized review, working alongside standard radiological workflows without altering diagnostic images or making diagnoses.

BriefCase is a radiological computer-assisted triage and notification software designed to analyze CT Pulmonary Angiography (CTPA) images in adults and transitional adolescents aged 18 and older for the detection of suspected Pulmonary Embolism (PE). The system utilizes an artificial intelligence (AI) deep-learning algorithm within a three-module architecture: an Aidoc Hospital Server for image acquisition, an Aidoc Cloud Server for image processing, and an Aidoc Desktop Application for workflow integration. It processes DICOM images, identifies potential PE findings, and alerts medical specialists via notifications that include compressed, non-diagnostic preview images. This version of BriefCase is capable of analyzing both single and dual-energy CTPA exams and allows users to filter notifications, though it does not alter original medical images or replace standard diagnostic procedures, functioning in parallel to the existing standard of care. A pivotal retrospective, multicenter study involving 499 CTPA cases demonstrated BriefCase's efficacy. The software achieved a sensitivity of 94.86% and a specificity of 94.04% for PE detection, surpassing the predefined 80% performance goal. The mean time-to-notification for positive cases was 78.0 seconds, significantly faster than its predicate. Additional reported metrics include a Positive Predictive Value (PPV) of 73.7% and a Negative Predictive Value (NPV) of 99.0%. The study also validated two additional operating points, one optimized for higher sensitivity (98.60%) and another for higher specificity (98.25%), underscoring its role in expediting the triage process for suspected PE by providing timely notifications to clinicians.

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 algorithm's performance vary between single-energy and dual-energy CTPA exams, given its expanded capability to process both?; What is the quantifiable impact on overall patient outcomes, such as time-to-treatment or length of hospital stay, when BriefCase is integrated into a clinical workflow for PE detection, beyond the reported time-to-notification?; Does the AI algorithm differentiate between acute and chronic pulmonary emboli, and if so, how does this distinction affect its sensitivity, specificity, or the nature of the notification provided?; What are the specific computational and infrastructure requirements for the Aidoc Hospital Server and Aidoc Cloud Server components, particularly under high-volume caseloads typical in large hospital networks?; Considering the Positive Predictive Value of 73.7%, what user training protocols or interface features are in place or recommended to help clinicians effectively manage potential false-positive notifications and mitigate alert fatigue?