Annalise Enterprise

Annalise Enterprise is an AI-powered software tool that analyzes chest X-ray images to identify suspected findings including pneumothorax, tension pneumothorax, pleural effusion, pneumoperitoneum, and vertebral compression fracture. The device interfaces with medical imaging systems (PACS/RIS) to prioritize studies in the clinical worklist based on detected findings, enabling radiologists to review potentially critical cases earlier while never downgrading existing study priorities.

The subject device differs from its predicate (Annalise Enterprise CXR Triage Trauma, K222179) in two technological areas: the set of findings detected and the underlying AI algorithm models. Both devices use convolutional neural networks trained on deep-learning techniques with over 750,000 chest X-ray studies and interface with PACS/RIS systems to provide worklist prioritization notifications. The differences represent different clinical conditions of interest but employ the same operational principles and workflow integration.

Not stated in this summary.

Substantial equivalence is established through two pathways: First, both devices are software-only packages using identical AI-based principles of operation, requiring the same inputs (DICOM chest X-ray images) and providing the same outputs (worklist prioritization notifications). Second, the applicant conducted standalone performance testing on 3,252 cases from four independent hospital sites with ABR-certified radiologist ground truth, demonstrating high sensitivity and specificity (AUC 0.972–0.989) across all five findings. Triage turn-around time of 42.3 seconds matched the predicate's published performance, and the technological differences do not raise new safety or effectiveness questions since both devices operate in parallel to standard clinical workflow without replacing radiologist interpretation.