AI-Rad Companion Organs RT

AI-Rad Companion Organs RT is post-processing software that automatically contours predefined anatomical structures (organs at risk and brain metastases) from CT and MR DICOM images using deep-learning algorithms. The generated contours are intended as input for radiation therapy treatment planning workflows and must be reviewed, edited, and accepted by qualified clinicians using treatment planning systems or interactive contouring applications before clinical use.

The subject device adds new MR-brain metastases and MR-brain OAR contouring algorithms while enhancing existing CT and MR pelvis algorithms compared to predicate K242745. Both devices use deep learning, process DICOM CT/MR data, deploy via edge and cloud, output DICOM-RT structure sets, and support any DICOM-compliant treatment planning system. The subject device expands scanner model compatibility for MR beyond Siemens-only data and qualifies algorithms for adult populations with expanded validation datasets.

ISO 62366-1 (usability engineering), ISO 14971:2019 (risk management), IEC 62304 (medical device software lifecycle), DICOM PS 3.1–3.20 (digital imaging and communications), ISO 15223-1 (medical device symbols), IEC 82304-1 (health software safety), IEC 81001-5-1 (health software and IT system safety), and ISO 20417 (manufacturer information).

The subject device is substantially equivalent to predicate K242745 because both use deep-learning algorithms to automatically segment organs from DICOM images for radiation therapy planning and must be used with treatment planning systems for review and editing. Performance validation demonstrates the enhanced CT algorithm and new MR algorithms are equivalent or superior to the predicate and reference devices in segmentation metrics (Dice coefficient, ASSD). The risk analysis and verification/validation data support equivalent safety and effectiveness with no changes to intended use.