ARTAssistant

ARTAssistant is a standalone software system designed to assist in adaptive radiation therapy by processing medical images through registration, synthetic CT generation, and contouring. It enables radiation oncologists to design daily adaptive radiotherapy plans based on processed imaging from CT, MRI, PET, and CBCT modalities.

The device uses intensity-based algorithms for rigid and deformable image registration (matching the predicate AccuContour K221706), but employs deep learning for synthetic CT generation rather than the predicate's absence of this feature. It supports multi-modality registration, CBCT image enhancement, manual contouring with quality assurance, and generates synthetic CT from both MR and CBCT sources.

AAPM TG-119 methodology was used to validate image conversion accuracy through gamma analysis. Normalized Mutual Information (NMI) and Hausdorff Distance (HD) metrics were used to evaluate registration performance. Software verification and validation followed FDA's Guideline for Industry and FDA Staff on Content of Premarket Submission for Device Software Functions.

ARTAssistant is substantially equivalent to AccuContour (K221706) because both are Class II medical image management systems with the same intended use in adaptive radiotherapy, identical regulatory classification and product codes, and comparable core functions including automatic rigid/deformable registration and manual contouring. Performance testing demonstrated non-inferiority of registration algorithms compared to the predicate, and image conversion accuracy met established clinical standards.