Brainomix 360 e-Lung

Brainomix 360 e-Lung is a software package that analyzes CT chest scans to provide quantitative measurements of lung tissue, including 3D segmentation, volumetric measurements, and density evaluations. It is intended as a workflow enhancement tool to support physicians in examining pulmonary and thoracic tissue and documenting radiological findings that may indicate chest disease.

The proposed device uses an AI/ML image processing algorithm for lung segmentation (versus the predicate's non-AI approach), adds a longitudinal assessment feature that automatically groups and compares scans from the same patient, and outputs results to PACS and cloud platforms (versus the predicate's web UI only). Minor UI improvements include adjustable slab thickness, original resolution viewing, and dynamic MPR with cross hair.

DICOM (Digital Imaging and Communications in Medicine) compliance per NEMA PS 3.1–3.20; ISO 14971:2019 for risk management; FDA Cybersecurity Guidance and IEC 81001-5-1 for cybersecurity; 21 CFR Part 820.30 design controls. Validation included head-to-head comparison of lung segmentation accuracy using Dice Similarity Coefficient (DSC) against a ground-truth consensus mask from three board-certified radiologists.

Both devices perform the same intended function—quantitative CT analysis of lung tissue using Hounsfield Units to distinguish tissue types, with identical indications for use, product code (JAK), intended 2D comparative review, and reporting capabilities. Although the proposed device improves the segmentation algorithm with AI/ML (demonstrating higher DSC values) and adds longitudinal assessment and PACS/cloud output features, these enhancements do not raise different safety or effectiveness questions because they automate and improve existing capabilities without changing the fundamental principles of operation or expanding the scope of use.