AutoContour

Health Canada

None

Vendor neutral. Supports integration via DICOM workflows or ESAPI integration with Eclipse. No specific CT vendor dependency.

Not applicable

Not applicable

Adaptive RT

Adult

3D, multi-slice, DICOM, CT, MR, PET, CBCT, RTIMAGE, Frame of Reference (FOR), volumetric imaging, radiotherapy imaging datasets

DICOM, Frame of Reference (FOR), RTIMAGE

DICOM, RTSTRUCT, structure sets, image registrations, deformable image registration, rigid image registration, DICOM registration objects

Deformable registration, DICOM Registration objects, DICOM RTSTRUCT, Rigid registration

Integrated in TPS, Stand-alone application, Web-based application

Integration via vendor marketplace/platform

Not available

All TPS vendors

Cloud-based, Hybrid solution, Locally on dedicated hardware

Automatically, right after the image acquisition, On demand, triggered by a user through e.g. a button click, image upload

Not applicable

Models are developed within a structured model atlas aligned to established international contouring guidelines, including DAHANCA, CTG, EORTC, ESTRO, GORTEC, HKNPCSG, NCIC CTG, NCRI, NRG Oncology, RADCOMP, RTOG, and TROG, together with AAPM TG-263 nomenclature, to support standardized and consistent contour generation across clinicians, sites, and radiotherapy workflows. Designed to improve contouring consistency, support workflow efficiency, and reduce manual contouring workload. Processing time depends on the product configuration and deployment environment. For cloud-based processing, total processing time is influenced by dataset size and the number of requested contours; contours typically begin returning within seconds, with most datasets completing in under one minute. For on-premise deployments, performance depends on the installed hardware configuration and may vary accordingly.

Performance may be affected by image quality, imaging artifacts, atypical anatomy, prior surgery, implanted devices, or anatomy outside the scope of the trained model atlas. Accuracy may vary across imaging modalities, acquisition protocols, and disease presentations. Model training and validation were performed primarily using adult patient imaging data; therefore, performance on paediatric cases may be reduced and results may not be optimal in paediatric patient populations. Generated contours and registrations require review and verification by qualified clinical users prior to clinical use.