JLK-SDH

The JLK-SDH is a radiological computer-aided triage and notification software that employs an artificial intelligence algorithm to identify suspected subdural hemorrhage from non-contrast head CT images, intended to assist trained radiologists. This device's methodology involves analyzing imaging data for SDH indicators and sending high-accuracy notifications, underpinned by performance validation demonstrating 97.1% sensitivity and 97.4% specificity, confirming its scientific utility as a clinical communication aid.

JLK, Inc.'s JLK-SDH is a radiological computer-aided triage and notification software designed to assist trained radiologists in identifying and communicating suspected subdural hemorrhage (SDH) from non-contrast CT images of the head. This device operates as a notification-only, parallel workflow tool, employing an artificial intelligence algorithm to analyze images and alert appropriate users. It has been determined to be substantially equivalent to a legally marketed predicate device for its intended use. The JLK-SDH system analyzes imaging data for findings suggestive of SDH and sends notifications to clinicians, recommending a review of the identified images. While it is a notification tool and not for diagnostic use, performance validation testing showed high accuracy, achieving 97.1% sensitivity and 97.4% specificity in detecting SDH across diverse patient demographics and scanner manufacturers. The system also demonstrated efficient notification times, underscoring its effectiveness as a triage and communication aid in clinical settings.

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. What specific features or patterns does the AI algorithm prioritize to achieve its high sensitivity and specificity in detecting SDH, and how does it differentiate from other intracranial pathologies?; How does the JLK-SDH system integrate into existing Picture Archiving and Communication Systems (PACS) and Hospital Information Systems (HIS), and what are the practical implications for seamless clinical adoption?; Beyond the stated sensitivity and specificity, what was the clinical impact of the system on patient management, such as reduction in turnaround time or improved patient outcomes?; Has the JLK-SDH been validated on diverse datasets representing different scanner models, image acquisition protocols, or patient populations, particularly those with subtle or atypical presentations of SDH?; What are the known limitations or potential failure modes of the AI algorithm, such as false positives or false negatives, and under what circumstances are these more likely to occur?