qXR-Detect

qXR-Detect is a computer-assisted detection (CADe) software device that analyzes chest radiographs and highlights suspicious regions of interest (ROIs) across six anatomical categories: lung, pleura, bone, mediastinum & hila & heart, hardware, and other. It is intended for use as a concurrent reading aid by ER physicians, family medicine practitioners, and radiologists to support (not replace) clinical decision-making in adult patients.

qXR-Detect uses deep learning (CNN-based architecture) to analyze DICOM chest X-ray images, similar to the predicate device Chest CAD. Both devices employ machine learning for ROI detection and localization, support PACS integration and cloud or on-premises deployment, and are HIPAA-compliant. qXR-Detect categorizes findings into six anatomical categories versus the predicate's seven categories, but both share the same fundamental CADe software approach and intended users.

ISO 13485:2016 (quality management systems) and IEC 62304:2006+A1:2015 (medical device software lifecycle) were referenced as regulatory standards. Testing included unit, integration, regression, and user acceptance testing; non-clinical standalone performance testing with 301 chest X-ray samples; and a clinical multireader multicase study with 18 readers evaluating detection and localization using weighted alternative free-response receiver operating characteristic (wAFROC) metrics.

qXR-Detect is substantially equivalent to Chest CAD (K210666) because both are CADe software devices that analyze chest radiographs using deep learning to identify and localize suspicious ROIs for concurrent clinical use in adults. The devices share identical regulatory classification (Class II, 21 CFR 892.2070, product code MYN), similar intended purpose, overlapping technological characteristics, and comparable performance metrics. Clinical testing demonstrated significant improvement in reader performance when aided by qXR-Detect, with wAFROC improving from 0.6894 unaided to 0.7505 aided (p<0.001), confirming safety and effectiveness without raising new safety or effectiveness questions.