CINA CHEST

CINA CHEST is an AI-powered software designed to expedite the triage of critical findings by analyzing chest and thoraco-abdominal CT angiograms for suspected Pulmonary Embolism and Aortic Dissection. It employs distinct algorithms for PE and AD detection, providing non-diagnostic notifications and preview images to radiologists via a web application, aiming to improve workflow efficiency and prioritize urgent patient care.

CINA CHEST, developed by Avicenna.AI, is a radiological computer-aided triage and notification software designed to assist radiologists in managing workflow for urgent findings. It analyzes Chest and Thoraco-abdominal CT angiography (CTA) scans to detect suspected cases of Pulmonary Embolism (PE) and Aortic Dissection (AD). The system employs an artificial intelligence algorithm to process DICOM images, identifying potential PE or AD. These findings are then communicated through a standalone web application that operates in parallel with standard image interpretation workflows. The software provides active pop-up notifications and passive iconic flags on a worklist, along with compressed, non-diagnostic preview images, to alert clinicians to high-priority cases, thereby facilitating earlier review. Technologically, CINA CHEST runs on standard server hardware and consists of separate image processing applications for PE and AD detection, which can integrate with compatible medical image communication platforms like the CINA Platform. These applications verify DICOM properties, execute their respective algorithms, and generate notification results. Performance was evaluated in a retrospective, multicenter study using 396 cases for PE and 298 for AD. For PE detection, sensitivity was 91.1% and specificity was 91.8%. For AD detection, sensitivity was 96.4% and specificity was 97.5%. The mean time-to-notification was approximately 63 seconds for PE and 36.5 seconds for AD, demonstrating its potential to expedite the identification of critical cases.

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 algorithm differentiate between subtle PE/AD findings and mimics or artifacts in CT angiography, and what is its performance in these challenging scenarios?; What impact does the implementation of CINA CHEST have on overall patient outcomes and departmental efficiency metrics, such as report turnaround time or length of stay for PE/AD patients?; Are there specific patient populations or CT scanner parameters where the CINA CHEST algorithm's performance (sensitivity/specificity) is notably higher or lower?; How does the CINA CHEST system handle variations in image quality or contrast opacification across different institutions or scanners, and what are the minimum image quality requirements?; What is the learning curve for radiologists integrating CINA CHEST notifications into their existing workflow, and what strategies are effective for optimizing its use without inducing notification fatigue?