EFAI Chestsuite XR Malpositioned ETT Assessment System (ETT-XR-100)

EFAI ETTXR is a radiological computer-aided triage and notification software that analyzes chest X-ray images in adults to identify suspected malpositioned endotracheal tubes (ETTs) relative to the carina. It flags cases when the ETT distal tip is assessed as more than 7 cm above the carina, less than 3 cm above it, or below the carina, using deep learning algorithms to provide case-level notifications to PACS/workstations for workflow prioritization. The device is not intended for stand-alone clinical decision-making or to rule out malpositioned ETT.

Both the proposed device and predicate (BriefCase, K221330) are AI-based triage software analyzing chest X-rays for malpositioned ETT detection with similar intended uses, user populations, anatomical focus, DICOM image format, and workflow integration. The minor technical difference is that EFAI ETTXR delivers notifications as text-based JSON files rather than with embedded preview images, though both provide case-level outputs for triage purposes without removing or deprioritizing cases.

IEC 62304:2006/A1:2016 (Medical device software – Software life cycle processes), ISO 14971:2019 (Application of risk management to medical devices), and FDA Guidance documents on premarket submissions for device software functions and cybersecurity in medical devices. Compliance with 21 CFR Part 820.30 design control requirements was demonstrated.

EFAI ETTXR and BriefCase share the same intended use (triaging chest radiographs for suspected malpositioned ETTs), identical user populations and anatomical focus, equivalent AI algorithmic approaches, and comparable clinical workflow integration without case removal or deprioritization. Performance validation demonstrated sensitivity and specificity of 0.890 and 0.935 respectively, meeting the 80% performance goal and matching the predicate's performance. The sole technical difference—JSON notification format versus embedded preview images—does not introduce new safety or effectiveness risks since both provide case-level outputs for use with full clinical images.