AutoContour (RADAC V6)

{0} **FDA** U.S. FOOD & DRUG ADMINISTRATION August 14, 2026 Radformation, Inc. Jennifer Wampler Senior Regulatory Affairs Specialist 261 Madison Ave. 9th Floor New York, New York 10016 Re: K262452 Trade/Device Name: AutoContour (RADAC V6) Regulation Number: 21 CFR 892.2050 Regulation Name: Medical Image Management And Processing System Regulatory Class: Class II Product Code: QKB Dated: July 16, 2026 Received: July 17, 2026 Dear Jennifer Wampler: We have reviewed your section 510(k) premarket notification of intent to market the device referenced above and have determined the device is substantially equivalent (for the indications for use stated in the enclosure) to legally marketed predicate devices marketed in interstate commerce prior to May 28, 1976, the enactment date of the Medical Device Amendments, or to devices that have been reclassified in accordance with the provisions of the Federal Food, Drug, and Cosmetic Act (the Act) that do not require approval of a premarket approval application (PMA). You may, therefore, market the device, subject to the general controls provisions of the Act. Although this letter refers to your product as a device, please be aware that some cleared products may instead be combination products. The 510(k) Premarket Notification Database available at https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm identifies combination product submissions. The general controls provisions of the Act include requirements for annual registration, listing of devices, good manufacturing practice, labeling, and prohibitions against misbranding and adulteration. Please note: CDRH does not evaluate information related to contract liability warranties. We remind you, however, that device labeling must be truthful and not misleading. If your device is classified (see above) into either class II (Special Controls) or class III (PMA), it may be subject to additional controls. Existing major regulations affecting your device can be found in the Code of Federal Regulations, Title 21, Parts 800 to 898. In addition, FDA may publish further announcements concerning your device in the Federal Register. U.S. Food & Drug Administration 10903 New Hampshire Avenue Silver Spring, MD 20993 www.fda.gov {1} K262452 - Jennifer Wampler Page 2 Additional information about changes that may require a new premarket notification are provided in the FDA guidance documents entitled "Deciding When to Submit a 510(k) for a Change to an Existing Device" (https://www.fda.gov/media/99812/download) and "Deciding When to Submit a 510(k) for a Software Change to an Existing Device" (https://www.fda.gov/media/99785/download). Your device is also subject to, among other requirements, the Quality Management System Regulation (QMSR) (21 CFR Part 820), which includes, but is not limited to, ISO 13485 clause 7.3 (Design controls), ISO 13485 clause 8.3 (Nonconforming product), ISO 13485 clause 8.5.2 (Corrective action), and ISO 13485 clause 8.5.3 (Preventative action). Please note that regardless of whether a change requires premarket review, the QMSR requires device manufacturers to review and approve changes to device design and production (ISO 13485 clause 7.3 and ISO 13485 clause 7.5) and document changes and approvals in the Medical Device File (ISO 13485 clause 4.2.3). Please be advised that FDA's issuance of a substantial equivalence determination does not mean that FDA has made a determination that your device complies with other requirements of the Act or any Federal statutes and regulations administered by other Federal agencies. You must comply with all the Act's requirements, including, but not limited to: registration and listing (21 CFR Part 807); labeling (21 CFR Part 801); medical device reporting (reporting of medical device-related adverse events) (21 CFR Part 803) for devices or postmarketing safety reporting (21 CFR Part 4, Subpart B) for combination products (see https://www.fda.gov/combination-products/guidance-regulatory-information/postmarketing-safety-reporting-combination-products); good manufacturing practice requirements as set forth in the Quality Management System Regulation (QMSR) (21 CFR Part 820) for devices or current good manufacturing practices (21 CFR Part 4, Subpart A) for combination products; and, if applicable, the electronic product radiation control provisions (Sections 531-542 of the Act); 21 CFR Parts 1000-1050. All medical devices, including Class I and unclassified devices and combination product device constituent parts are required to be in compliance with the final Unique Device Identification System rule ("UDI Rule"). The UDI Rule requires, among other things, that a device bear a unique device identifier (UDI) on its label and package (21 CFR 801.20(a)) unless an exception or alternative applies (21 CFR 801.20(b)) and that the dates on the device label be formatted in accordance with 21 CFR 801.18. The UDI Rule (21 CFR 830.300(a) and 830.320(b)) also requires that certain information be submitted to the Global Unique Device Identification Database (GUDID) (21 CFR Part 830 Subpart E). For additional information on these requirements, please see the UDI System webpage at https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance/unique-device-identification-system-udi-system. Also, please note the regulation entitled, "Misbranding by reference to premarket notification" (21 CFR 807.97). For questions regarding the reporting of adverse events under the MDR regulation (21 CFR Part 803), please go to https://www.fda.gov/medical-devices/medical-device-safety/medical-device-reporting-mdr-how-report-medical-device-problems. For comprehensive regulatory information about medical devices and radiation-emitting products, including information about labeling regulations, please see Device Advice (https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance) and CDRH Learn (https://www.fda.gov/training-and-continuing-education/cdrh-learn). Additionally, you may contact the Division of Industry and Consumer Education (DICE) to ask a question about a specific regulatory topic. See the DICE website (https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory- {2} K262452 - Jennifer Wampler Page 3 assistance/contact-us-division-industry-and-consumer-education-dice) for more information or contact DICE by email (DICE@fda.hhs.gov) or phone (1-800-638-2041 or 301-796-7100). Sincerely, Lora D. Weidner, Ph.D. Assistant Director Radiation Therapy Team DHT8C: Division of Radiological Imaging and Radiation Therapy Devices OHT8: Office of Radiological Health Office of Product Evaluation and Quality Center for Devices and Radiological Health Enclosure {3} | Indications for Use | | | | --- | --- | --- | | Please type in the marketing application/submission number, if it is known. This textbox will be left blank for original applications/submissions. | | K262452 | | Please provide the device trade name(s). | | ? | | AutoContour (RADAC V6) | | | | Please provide your Indications for Use below. | | ? | | AutoContour is intended to assist radiation treatment planners in contouring and reviewing structures within medical images in preparation for radiation therapy treatment planning. | | | | Please select the types of uses (select one or both, as applicable). | ☑ Prescription Use (21 CFR 801 Subpart D) ☐ Over-The-Counter Use (21 CFR 801 Subpart C) | ? | | Please select the age group(s) for which the device(s) is to be used. | ☐ Neonates/Newborns (Birth to < 29 days old) ☐ Infants (29 days old to < 2 years old) ☐ Children (2 years old to < 12 years old) ☐ Adolescents (12 years old to < 22 years old) ☑ Adults (22 years old and greater) | ? | {4} RAD formation K262452 # 510(k) Summary - AutoContour Model RADAC V6 This 510(k) Summary has been created per the requirements of the Safe Medical Device Act (SMDA) of 1990, and the content is provided in conformance with 21 CFR Part 807.92. # 1. Submitter's Information | Table 1: Submitter's Information | | | --- | --- | | Submitter's Name: | Kevin Robinson | | Company: | Radformation, Inc. | | Address: | 261 Madison Avenue, 9th Floor New York, NY 10016 | | Contact Person: | Jennifer Wampler Sr. Regulatory Specialist, Radformation | | Phone: | 844-723-3675 | | Fax: | — | | Email: | regulatory@radformation.com | | Date of Summary Preparation | 8/5/2026 | # 2. Device Information | Table 2 : Device Information | | | --- | --- | | Trade Name: | AutoContour | | Model Name: | RADAC V6 | | Common Names: | AutoContour, AutoContouring, AutoContour Agent, AutoContour Cloud Server | | Classification Name: | Medical image management and processing system | | Classification: | Class II | | Regulation Number: | 892.2050 | | Product Code: | QKB | | Classification Panel: | Radiology | AutoContour Model RADAC V6 Special 510(k) - Summary 1 {5} ### **3. Predicate Device Information** AutoContour Model RADAC V6 (Subject Device) makes use of its prior submissions - AutoContour Model RADAC V5 (K260509) as the Predicate Device. The Indications for Use, patient population, functionality, and technical components of this Predicate Device remain unchanged in AutoContour Model RADAC V6. This submission is intended to build on the functionality and technological components of the 510(k) cleared AutoContour Model RADAC V5. ### **4. Device Description** The AutoContour Model RADAC V6 device is software that uses DICOM-compliant image data (CT or MR) as input to: (1) automatically contour various structures of interest for radiation therapy treatment planning using machine learning-based contouring. The deep learning-based structure models are trained using multiple imaging datasets consisting of anatomical organs of the head, neck, thorax, abdomen, and pelvis of adult male and female patients. Model performance is validated against ground truth contours to ensure model quality. Users are allowed to review and modify the resulting contours and generate DICOM-compliant structure set data that can be imported into a radiation therapy treatment planning system (TPS). AutoContour Model RADAC V6 consists of 3 main components: 1. A .NET client application designed to run on the Windows Operating System, allowing the user to load image and structure sets for upload to the cloud-based server for automatic contouring, perform registration with other image sets, as well as review, edit, and export the structure set. 2. A local “agent” service designed to run on the Windows Operating System that is configured by the user to monitor a network storage location for new CT and MR datasets that are to be automatically contoured. 3. A cloud-based automatic contouring service that produces initial contours based on image sets sent by the user from the .NET client application. ### **5. Indications for Use** AutoContour is intended to assist radiation treatment planners in contouring and reviewing structures within medical images in preparation for radiation therapy treatment planning. ### **6. Technological Characteristics** AutoContour RADAC V6 (Subject Device) makes use of its prior submission - AutoContour RADAC V5 (K260509), a Predicate Device. The Indications for Use, patient population, functionality, and technical components of this Predicate Device remain unchanged in AutoContour RADAC V6. The main UI outputs are equivalent to the Predicate Device as well, allowing the user to properly visualize and analyze the calculations. This submission is intended to build on the functionality and technological components of the 510(k) cleared AutoContour RADAC V5. AutoContour Model RADAC V6 extends the existing automatic Planning Structure generation to the Zero-Click workflow. The Predicate Device supports Planning Structure generation through the AutoContour user interface and templates, where users can add automatically generated AutoContour Model RADAC V6 Special 510(k) 2 {6} Planning Structures by applying predefined operations (crop, boolean, etc.) to the initial contours. Additional Zero-Click updates include workflow, configuration, and user interface enhancements to support Planning Structure processing, job management, template matching rules, and DICOM data handling. These enhancements allow the user to extend manual workflow operations to the Zero-Click workflow. These changes do not modify the underlying contour generation algorithms, intended use, technological characteristics or risk controls of the Predicate Device. These updates to AutoContour Model RADAC V6 enhancements do not introduce new cybersecurity functionality or modify the device's cybersecurity controls, network communications, authentication, or external interfaces. Verification testing was performed to confirm the correct operation of the Zero-Click workflow, including Planning Structure generation, workflow processing, DICOM data handling, configuration updates, and regression testing of existing AutoContour functionality. The verification and validation results summarized in the respective test records demonstrated that the Subject Device performs as intended and remains substantially equivalent to the Predicate Device. | **Table 3: Substantial Equivalence AutoContour Model RADAC V6 vs. AutoContour Model RADAC V5 (K260509)** | | | | --- | --- | --- | | Characteristic | Subject Device: AutoContour Model RADAC V6 | Predicate Device: AutoContour Model RADAC V5 (K260509) | | **AutoContour vs. Predicate Device: Technological Characteristics** | | | | Indications for Use | AutoContour is intended to assist radiation treatment planners in contouring and reviewing structures within medical images in preparation for radiation therapy treatment planning *(Equivalent to Predicate)* | AutoContour is intended to assist radiation treatment planners in contouring and reviewing structures within medical images in preparation for radiation therapy treatment planning | | Target Population | Any patient type for whom relevant modality scan data is available. *(Equivalent to Predicate)* | Any patient type for whom relevant modality scan data is available. | | Energy Used and/or Delivered | None – software-only application. The software application does not deliver or depend on energy delivered to or from patients *(Equivalent to Predicate)* | None – software-only application. The software application does not deliver or depend on energy delivered to or from patients | | Intended users | Trained radiation oncology personnel *(Equivalent to Predicate)* | Trained radiation oncology personnel | | Design: Data Visualization/Graphical User Interface | Contains both an automated processing component and Data Visualization / Graphical User Interface | Contains both an automated processing component and Data Visualization / Graphical User Interface | AutoContour Model RADAC V6 Special 510(k) 3 {7} | | *(Equivalent to Predicate)* | | | --- | --- | --- | | Design: View manipulation and Volume rendering | Window and level, pan, zoom, cross-hairs, slice navigation, fused views. *(Equivalent to Predicate)* | Window and level, pan, zoom, cross-hairs, slice navigation, fused views. | | Design: Workflow | Interaction within a User Interface as well as automation via a Zero-Click, widening availability of options across workflows *(Substantially Equivalent to Predicate)* | Interaction within a User Interface as well as automation via a Zero-Click operation | | Design: Image registration | Manual and Automatic Rigid registration. Automatic Deformable Registration *(Equivalent to Predicate)* | Manual and Automatic Rigid registration. Automatic Deformable Registration | | **AutoContour vs. Predicate Device: Model Comparison** | | | | Regions and Volumes of interest (ROI) | CT or MR input for contouring of anatomical regions: Head and Neck, Thorax, Abdomen and Pelvis. Machine learning based contouring of **442 CT-based** and **81 MR-based models** and manual ROI manipulation CT Models: • A_Aorta* • A_Aorta_Asc* • A_Aorta_Dsc* • A_Brachiocephals* • A_Carotid_L • A_Carotid_R • A_Celiac • A_Circumflex_L • A_Coronary_2d_R • A_Coronary_L • A_Coronary_R • A_LAD* • A_Mesenteric_S • A_Pulmonary • A_Subclavian_L* • A_Subclavian_R* • Atrium_L • Atrium_R • AV_Node • Barrigel™ • BileDuct_Common* • Bladder • Bladder_CBCT • Bladder_F • Body • Body+Mask • Bone_Hyoid | CT or MR input for contouring of anatomical regions: Head and Neck, Thorax, Abdomen and Pelvis. Machine learning based contouring of 420 CT-based and 62 MR-based models and manual ROI manipulation CT Models: • A_Aorta • A_Aorta_Asc • A_Aorta_Dsc • A_Brachiocephals • A_Carotid_L • A_Carotid_R • A_Celiac • A_Circumflex_L • A_Coronary_2d_R • A_Coronary_L • A_Coronary_R • A_LAD • A_Mesenteric_S • A_Pulmonary • A_Subclavian_L • A_Subclavian_R • Atrium_L • Atrium_R • AV_Node • Barrigel™ • BileDuct_Common • Bladder • Bladder_CBCT • Bladder_F • Body • Body+Mask • Bone_Hyoid • Bone_Ilium • Bone_Ilium_L • Bone_Ilium_R • Bone_Ischium_L • Bone_Ischium_R | AutoContour Model RADAC V6 Special 510(k) 4 {8} | | • Bone_Ilium • Bone_Ilium_L • Bone_Ilium_R • Bone_Ischium_L • Bone_Ischium_R • Bone_Mandible* • Bone_Pelvic • Bone_Pterygoid_L • Bone_Pterygoid_R • Bone_PubicSymphys • Bone_Sacrum • Bone_Skull • Bone_Sternum • Bone_Teeth* • Bowel • Bowel_Bag* • Bowel_Bag_F* • Bowel_F • Bowel_Large • Bowel_Large_F • Bowel_Small • Bowel_Small_F • BrachialPlex_L • BrachialPlex_R • BrachialPlexs • Brain • Brainstem • Breast_Implant_L • Breast_Implant_R • Breast_L • Breast_Prone • **Breast_Prone_RTOG** • Breast_R • Breast_RTOG_L • Breast_RTOG_R • Breast_Wire_L • Breast_Wire_R • Breasts_RTOG • Bronchus • BuccalMucosa* • Canal_Anal • Canal_Anal_F • Carina • CaudaEquina • Cavity_Oral • Cavity_Oral_Ext • Cerebellum • Chestwall_2cm_L • Chestwall_2cm_R • Chestwall_Anat • Chestwall_Anat_L • Chestwall_Anat_R • Chestwall_ESTRO_L • Chestwall_ESTRO_R • Chestwall_L* | • Bone_Mandible • Bone_Pelvic • Bone_Pterygoid_L • Bone_Pterygoid_R • Bone_PubicSymphys • Bone_Sacrum • Bone_Skull • Bone_Sternum • Bone_Teeth • Bowel • Bowel_Bag • Bowel_Bag_F • Bowel_F • Bowel_Large • Bowel_Large_F • Bowel_Small • Bowel_Small_F • BrachialPlex_L • BrachialPlex_R • BrachialPlexs • Brain • Brainstem • Breast_Implant_L • Breast_Implant_R • Breast_L • Breast_Prone • Breast_R • Breast_RTOG_L • Breast_RTOG_R • Breast_Wire_L • Breast_Wire_R • Breasts_RTOG • Bronchus • BuccalMucosa • Canal_Anal • Canal_Anal_F • Carina • CaudaEquina • Cavity_Oral • Cavity_Oral_Ext • Cerebellum • Chestwall_2cm_L • Chestwall_2cm_R • Chestwall_Anat • Chestwall_Anat_L • Chestwall_Anat_R • Chestwall_ESTRO_L • Chestwall_ESTRO_R • Chestwall_L • Chestwall_OAR • Chestwall_R • Chestwall_RC_L • Chestwall_RC_R • Clavicle_L • Clavicle_R • Cochlea_L • Cochlea_R • Colon_Sigmoid • Cornea_2d_L • Cornea_2d_R • Cornea_L • Cornea_R • CorpusCallosum | | --- | --- | --- | AutoContour Model RADAC V6 Special 510(k) 5 {9} | | • Chestwall_OAR* • Chestwall_R* • Chestwall_RC_L* • Chestwall_RC_R* • Clavicle_L • Clavicle_R • Cochlea_L • Cochlea_R • Colon_Sigmoid • Cornea_2d_L • Cornea_2d_R • Cornea_L • Cornea_R • CorpusCallosum • Cribriform • Dental_Artifact* • Duodenum • Ear_Internal_L • Ear_Internal_R • Esophagus • External • Eye_L • Eye_R • Eyes • Falx • Femur_Head_L • Femur_Head_R • Femur_L • Femur_L_CBCT • Femur_R • Femur_R_CBCT • Femur_RTOG_L • Femur_RTOG_R • Femurs_RTOG • Foley_Balloon • GallBladder* • Genitals_M • Genitals_F • Glnd_Lacrimal_L • Glnd_Lacrimal_R • Glnd_Submand_L • Glnd_Submand_R • Glnd_Thyroid • GreatVessels* • HDR_Bladder • HDR_Bowel* • HDR_Canal_Anal* • HDR_Colon_Sigmoid* • HDR_Cylinder • HDR_Rectum* • HDR_Ring • HDR_Urethra • HDR_UteroCervix • Heart • Heart_Prone* | • Cribriform • Dental_Artifact • Duodenum • Ear_Internal_L • Ear_Internal_R • Esophagus • External • Eye_L • Eye_R • Eyes • Falx • Femur_Head_L • Femur_Head_R • Femur_L • Femur_L_CBCT • Femur_R • Femur_R_CBCT • Femur_RTOG_L • Femur_RTOG_R • Femurs_RTOG • Foley_Balloon • GallBladder • Genitals_F • Genitals_M • Glnd_Lacrimal_L • Glnd_Lacrimal_R • Glnd_Submand_L • Glnd_Submand_R • Glnd_Thyroid • GreatVessels • HDR_Bladder • HDR_Bowel • HDR_Canal_Anal • HDR_Colon_Sigmoid • HDR_Cylinder • HDR_Rectum • HDR_Ring • HDR_Urethra • Heart • Heart_Prone • Heart+A_Pulm • Hippocampus_L • Hippocampus_R • Humerus_L • Humerus_R • Iliac_Int_L • Iliac_Int_R • Iliac_L • Iliac_R • InternalAuditoryCanal_L • InternalAuditoryCanal_R • Kidney_L • Kidney_Outer_L • Kidney_Outer_R • Kidney_R • Kidneys • Kidneys_Outer • Larynx • Larynx_Glottic • Larynx_NRG • Larynx_SG • Lens_L • Lens_R | | --- | --- | --- | AutoContour Model RADAC V6 Special 510(k) 6 {10} | | Heart+A_PulmHippocampus_LHippocampus_RHumerus_LHumerus_RIliac_Int_LIliac_Int_RIliac_LIliac_RInternalAuditoryCanal_LInternalAuditoryCanal_RKidney_LKidney_Outer_LKidney_Outer_RKidney_RKidneysKidneys_OuterLarynxLarynx_GlotticLarynx_NRGLarynx_SGLens_LLens_RLips*Liver*LN_Ax_LLN_Ax_L1_2d_LLN_Ax_L1_2d_RLN_Ax_L1_ESTRO_2d_LLN_Ax_L1_ESTRO_2d_RLN_Ax_L1_ESTRO_LLN_Ax_L1_ESTRO_RLN_Ax_L1_LLN_Ax_L1_RLN_Ax_L2_2d_LLN_Ax_L2_2d_RLN_Ax_L2_ESTRO_2d_LLN_Ax_L2_ESTRO_2d_RLN_Ax_L2_ESTRO_LLN_Ax_L2_ESTRO_RLN_Ax_L2_InPec_ESTRO_2d_LLN_Ax_L2_InPec_ESTRO_2d_RLN_Ax_L2_LLN_Ax_L2_L3_LLN_Ax_L2_L3_RLN_Ax_L2_RLN_Ax_L3_2d_LLN_Ax_L3_2d_RLN_Ax_L3_ESTRO_2d_LLN_Ax_L3_ESTRO_2d_RLN_Ax_L3_ESTRO_LLN_Ax_L3_ESTRO_RLN_Ax_L3_LLN_Ax_L3_RLN_Ax_R | LipsLiverLN_Ax_LLN_Ax_L1_2d_LLN_Ax_L1_2d_RLN_Ax_L1_ESTRO_2d_LLN_Ax_L1_ESTRO_2d_RLN_Ax_L1_ESTRO_LLN_Ax_L1_ESTRO_RLN_Ax_L1_LLN_Ax_L1_RLN_Ax_L2_2d_LLN_Ax_L2_2d_RLN_Ax_L2_ESTRO_2d_LLN_Ax_L2_ESTRO_2d_RLN_Ax_L2_ESTRO_LLN_Ax_L2_ESTRO_RLN_Ax_L2_InPec_ESTRO_2d_LLN_Ax_L2_InPec_ESTRO_2d_RLN_Ax_L2_LLN_Ax_L2_L3_LLN_Ax_L2_L3_RLN_Ax_L2_RLN_Ax_L3_2d_LLN_Ax_L3_2d_RLN_Ax_L3_ESTRO_2d_LLN_Ax_L3_ESTRO_2d_RLN_Ax_L3_ESTRO_LLN_Ax_L3_ESTRO_RLN_Ax_L3_LLN_Ax_L3_RLN_Ax_RLN_Ax_Sclav_2d_LLN_Ax_Sclav_2d_RLN_IMN_2d_LLN_IMN_2d_RLN_IMN_ESTRO_2d_LLN_IMN_ESTRO_2d_RLN_IMN_Expand_2d_LLN_IMN_Expand_2d_RLN_IMN_Expand_ESTRO_2d_LLN_IMN_Expand_ESTRO_2d_RLN_IMN_LLN_IMN_RLN_IMN_RC_LLN_IMN_RC_RLN_Inguinofem_LLN_Inguinofem_RLN_InPec_ESTRO_2d_LLN_InPec_ESTRO_2d_RLN_InPec_ESTRO_LLN_InPec_ESTRO_RLN_MesorectumLN_Neck_2d_LLN_Neck_2d_RLN_Neck_CerVII_2d_LLN_Neck_CerVII_2d_RLN_Neck_IALN_Neck_IA_2dLN_Neck_IA_VI_2dLN_Neck_IB_2d_LLN_Neck_IB_2d_RLN_Neck_IB_L | | --- | --- | --- | AutoContour Model RADAC V6 Special 510(k) 7 {11} | | • LN_Ax_Sclav_2d_L • LN_Ax_Sclav_2d_R • LN_IMN_2d_L • LN_IMN_2d_R • LN_IMN_ESTRO_2d_L • LN_IMN_ESTRO_2d_R • LN_IMN_Expand_2d_L • LN_IMN_Expand_2d_R • LN_IMN_Expand_ESTRO_2d_L • LN_IMN_Expand_ESTRO_2d_R • LN_IMN_L • LN_IMN_R • LN_IMN_RC_L • LN_IMN_RC_R • LN_Inguinofem_L • LN_Inguinofem_R • LN_InPec_ESTRO_2d_L • LN_InPec_ESTRO_2d_R • LN_InPec_ESTRO_L • LN_InPec_ESTRO_R • LN_Mesorectum • LN_Mediast_1L • LN_Mediast_1R • LN_Mediast_2L • LN_Mediast_2R • LN_Mediast_3A • LN_Mediast_3P • LN_Mediast_4L • LN_Mediast_4R • LN_Mediast_5 • LN_Mediast_6 • LN_Mediast_7 • LN_Mediast_8 • LN_Mediast_10 • LN_Mediast_11 • LN_Neck_2d_L • LN_Neck_2d_R • LN_Neck_CerVII_2d_L* • LN_Neck_CerVII_2d_R* • LN_Neck_IA • LN_Neck_IA_2d • LN_Neck_IA_VI_2d • LN_Neck_IB_2d_L • LN_Neck_IB_2d_R • LN_Neck_IB_L • LN_Neck_IB_R • LN_Neck_IB-V_L • LN_Neck_IB-V_R • LN_Neck_II_2d_L • LN_Neck_II_2d_R • LN_Neck_II_L • LN_Neck_II_R • LN_Neck_III_2d_L • LN_Neck_III_2d_R • LN_Neck_III_L • LN_Neck_III_R • LN_Neck_II-IV_2d_L • LN_Neck_II-IV_2d_R • LN_Neck_II-IV_L • LN_Neck_II-IV_R • LN_Neck_II-V_L • LN_Neck_II-V_R • LN_Neck_IV_2d_L • LN_Neck_IV_2d_R • LN_Neck_IV_L • LN_Neck_IV_R • LN_Neck_V_2d_L • LN_Neck_V_2d_R • LN_Neck_V_L • LN_Neck_V_R • LN_Neck_VI_2d • LN_Neck_VIA • LN_Neck_VII_2d_L • LN_Neck_VII_2d_R • LN_Neck_VIIA_2d_L • LN_Neck_VIIA_2d_R • LN_Neck_VIIA_L • LN_Neck_VIIA_R • LN_Neck_VIIB_2d_L • LN_Neck_VIIB_2d_R • LN_Neck_VIIB_L • LN_Neck_VIIB_R • LN_Paraaortic • LN_Pelvics • LN_Pelvics_CBCT • LN_Pelvics_F • LN_Pelvics_NRG • LN_Post_Neck_L • LN_Post_Neck_R • LN_Presacral • LN_Sclav_2d_L • LN_Sclav_2d_R • LN_Sclav_ESTRO_2d_L • LN_Sclav_ESTRO_2d_R • LN_Sclav_ESTRO_L • LN_Sclav_ESTRO_R • LN_Sclav_L • LN_Sclav_R • LN_Sclav_RC_L • LN_Sclav_RC_R • Lobe_Temporal_L • Lobe_Temporal_R • Lung_L • Lung_R • Lungs • Macula_L • Macula_R • Marrow_Ilium_L | | --- | --- | AutoContour Model RADAC V6 Special 510(k) 8 {12} | | • LN_Neck_III_R • LN_Neck_II-IV_2d_L* • LN_Neck_II-IV_2d_R* • LN_Neck_II-IV_L • LN_Neck_II-IV_R • LN_Neck_II-V_L • LN_Neck_II-V_R • LN_Neck_IV_2d_L* • LN_Neck_IV_2d_R* • LN_Neck_IVA_2d_L • LN_Neck_IVA_2d_R • LN_Neck_IVB_2d_L • LN_Neck_IVB_2d_R • LN_Neck_IV_L • LN_Neck_IV_R • LN_Neck_V_2d_L • LN_Neck_V_2d_R • LN_Neck_VC_2d_L • LN_Neck_VC_2d_R • LN_Neck_V_L • LN_Neck_V_R • LN_Neck_VI_2d • LN_Neck_VIA • LN_Neck_VII_2d_L • LN_Neck_VII_2d_R • LN_Neck_VIIA_2d_L • LN_Neck_VIIA_2d_R • LN_Neck_VIIA_L • LN_Neck_VIIA_R • LN_Neck_VIIB_2d_L • LN_Neck_VIIB_2d_R • LN_Neck_VIIB_L • LN_Neck_VIIB_R • LN_Paraaortic* • LN_Pelvics • LN_Pelvics_CBCT • LN_Pelvics_F • LN_Pelvics_NRG • LN_Post_Neck_L • LN_Post_Neck_R • LN_Presacral • LN_Rectum_EIN_L • LN_Rectum_EIN_R • LN_Sclav_2d_L • LN_Sclav_2d_R • LN_Sclav_ESTRO_2d_L • LN_Sclav_ESTRO_2d_R • LN_Sclav_ESTRO_L • LN_Sclav_ESTRO_R • LN_Sclav_L • LN_Sclav_R • LN_Sclav_RC_L • LN_Sclav_RC_R • Lobe_Temporal_L • Lobe_Temporal_R | • Marrow_Ilium_R • Marrow_Pelvis • Medulla • Midbrain • Musc_Constrict_2d • Musc_Constrict • Musc_Iliopsoas_L • Musc_Iliopsoas_R • Musc_PecMinor_L • Musc_PecMinor_R • Musc_Sclmast_L • Musc_Sclmast_R • Myocardium • Nipple_L • Nipple_Prone • Nipple_R • OpticChiasm • OpticNrv_L • OpticNrv_R • Optics • Pancreas • Parametrium • Parotid_Glnd_L • Parotid_Glnd_R • Parotid_L • Parotid_R • PenileBulb • Pericardium • Pericardium_Inf • Pericardium_Inf+A_Pulm • Pharynx • Pituitary • Pituitary_3d • Pons • Prostate • Prostate_CBCT • Prostate+SeminalVes • Prostate+SV-Spacergel • ProstateBed • ProstateFiducials • Prostate-Gel_Fiducials • Prostate-Spacergel_Vue • Rectum • Rectum_CBCT • Rectum_F • Rectum-Spacergel_Vue • Retina_2d_L • Retina_2d_R • Retina_L • Retina_R • Rib • Rib_L • Rib_R • Rib01_L • Rib01_R • Rib02_L • Rib02_R • Rib03_L • Rib03_R • Rib04_L • Rib04_R • Rib05_L • Rib05_R | | --- | --- | --- | AutoContour Model RADAC V6 Special 510(k) 9 {13} | | • Lung_L* • Lung_R* • Lungs* • Macula_L • Macula_R • Marrow_Ilium_L • Marrow_Ilium_R • Marrow_Pelvis • Medulla • Midbrain • Musc_Constrict_2d • Musc_Constrict • Musc_Iliopsoas_L • Musc_Iliopsoas_R • Musc_PecMinor_L • Musc_PecMinor_R • Musc_Sclmast_L • Musc_Sclmast_R • Myocardium • Nipple_L • Nipple_Prone • Nipple_R • OpticChiasm • OpticNrv_L • OpticNrv_R • Optics • Pancreas • Parametrium • Parotid_Glnd_L • Parotid_Glnd_R • Parotid_L • Parotid_R • PenileBulb • Pericardium* • Pericardium_Inf • Pericardium_Inf+A_Pulm • Pharynx • Pituitary • Pituitary_3d • Pons • Prostate* • Prostate_CBCT • Prostate+SeminalVes • Prostate+SV-Spacergel • ProstateBed • ProstateFiducials • Prostate-Gel_Fiducials • Prostate-Spacergel_Vue • Rectum • Rectum_CBCT • Rectum_F • Rectum-Spacergel_Vue • Retina_2d_L • Retina_2d_R • Retina_L | • Rib06_L • Rib06_R • Rib07_L • Rib07_R • Rib08_L • Rib08_R • Rib09_L • Rib09_R • Rib10_L • Rib10_R • Rib11_L • Rib11_R • Rib12_L • Rib12_R • SacralPlex_L • SacralPlex_R • SeminalVes • SeminalVes_CBCT • Sinuses • Skin • SpaceOAR™Vue • SpinalCanal • SpinalCord • Spleen • Stomach • Tentorium • Trachea • UteroCervix • V_Brachioceph_L • V_Brachioceph_R • V_Jugular_L • V_Jugular_R • V_Pulmonary • V_Venacava_I • V_Venacava_I_UKSABR • V_Venacava_S • V_Venacava_S_UKSABR • Vagina* • Valve_Aortic • Valve_Mitral • Valve_Pulmonic • Valve_Tricuspid • VB • VB_C1 • VB_C2 • VB_C3 • VB_C4 • VB_C5 • VB_C6 • VB_C7 • VB_L1 • VB_L2 • VB_L3 • VB_L4 • VB_L5 • VB_T01 • VB_T02 • VB_T03 • VB_T04 • VB_T05 • VB_T06 • VB_T07 • VB_T08 | | --- | --- | --- | AutoContour Model RADAC V6 Special 510(k) 10 {14} | | Retina_RRib*Rib_LRib_RRib01_LRib01_RRib02_LRib02_RRib03_LRib03_RRib04_LRib04_RRib05_LRib05_RRib06_LRib06_RRib07_LRib07_RRib08_LRib08_RRib09_LRib09_RRib10_LRib10_RRib11_LRib11_RRib12_LRib12_RSacralPlex_LSacralPlex_RSeminalVesSeminalVes_CBCTSinusesSkinSpaceOAR™VueSpinalCanalSpinalCordSpleenStomachTentoriumTracheaUteroCervixV_Brachioceph_L*V_Brachioceph_R*V_Jugular_LV_Jugular_RV_PulmonaryV_Venacava_IV_Venacava_I_UKSABRV_Venacava_SV_Venacava_S_UKSABRVaginaValve_AorticValve_MitralValve_Pulmonic | VB_T09VB_T10VB_T11VB_T12Ventricle_BrainVentricle_LVentricle_RVessels_PelvisMR Models:A_Pud_Int_LA_Pud_Int_RAmygdala_LAmygdala_RBladderBladder_TrigoneBladder_TRUFIBone_PubicSymphysBone_SacrumBrainBrainstemCerebellumColon_SigmoidCornea_LCornea_RCorpusCallosumExternal_PelvisEye_LEye_RFalxFemur_LFemur_RGlnd_ProstateHDR_BladderHDR_BowelHDR_Canal_AnalHDR_Colon_SigmoidHDR_Rec-CanalHDR_RectumHDR_UrethraHippocampus_LHippocampus_RHypo_TrueHypothalamusLens_LLens_RMedullaMidbrainNVB_LNVB_ROpticChiasmOpticNrv_LOpticNrv_ROpticsOpticTract_LOpticTract_RPenileBulbPenileBulb_TRUFIPituitaryPonsProstateRectal_Spacer | | --- | --- | --- | AutoContour Model RADAC V6 Special 510(k) 11 {15} | | • Valve_Tricuspid • VB • VB_C1 • VB_C2 • VB_C3 • VB_C4 • VB_C5 • VB_C6 • VB_C7 • VB_L1 • VB_L2 • VB_L3 • VB_L4 • VB_L5 • VB_T01 • VB_T02 • VB_T03 • VB_T04 • VB_T05 • VB_T06 • VB_T07 • VB_T08 • VB_T09 • VB_T10 • VB_T11 • VB_T12 • Ventricle_Brain • Ventricle_L • Ventricle_R • Vessels_Pelvis MR Models: • A_Pud_Int_L • A_Pud_Int_R • Amygdala_L • Amygdala_R • Bladder • Bladder_Trigone • Bladder_TRUFI • Bone_PubicSymphys • Bone_Sacrum • Bowel • Brain • Brainstem • CaudateNucleus_L • CaudateNucleus_R • Cerebellum • Colon_Sigmoid • Cornea_L • Cornea_R • CorpusCallosum • Duodenum • Esophagus • External_Pelvis | • Rectum • Retina_L • Retina_R • SeminalVes • Sinuses • SpinalCord_Cerv • Tentorium • Thalamus • Urethra • Ventricle_Brain | | --- | --- | --- | AutoContour Model RADAC V6 Special 510(k) 12 {16} | | • Eye_L • Eye_R • Falx • Femur_L • Femur_R • Fornix_L • Fornix_R • Gallbladder • Glnd_Prostate • Glnd_Lacrimal_L • Glnd_Lacrimal_R • HDR_Bladder • HDR_Bowel • HDR_Canal_Anal • HDR_Colon_Sigmoid • HDR_Rec-Canal • HDR_Rectum • HDR_Urethra • Hippocampus_L • Hippocampus_R • Hypo_True • Hypothalamus • Kidney_L • Kidney_R • Lens_L • Lens_R • Liver • Medulla • Midbrain • NVB_L • NVB_R • OpticChiasm • OpticNrv_L • OpticNrv_R • Optics • OpticTract_L • OpticTract_R • Pancreas • PenileBulb • PenileBulb_TRUFI • Periventricular_Space • Pituitary • Pineal • Pons • Prostate • Rectal_Spacer • Rectum • Retina_L • Retina_R • SeminalVes • Sinuses • SpinalCord_Cerv • Spleen • Stomach • Tentorium | | | --- | --- | --- | AutoContour Model RADAC V6 Special 510(k) 13 {17} | | - Thalamus - **ThecalSac** - Urethra - Ventricle_Brain Bold = new; * = updated *(Substantially Equivalent)* | | | --- | --- | --- | | Design: Region/volume of interest measurements and size measurements | None – not applicable *(Equivalent to Predicate)* | None – not applicable | | Design: Region/Volume Quantification | None – not applicable *(Equivalent to Predicate)* | None – not applicable | | Design: Supported modalities | CT or MR input for contouring or registration/fusion. PET/CT input for registration/fusion only. DICOM RTSTRUCT and REGISTRATION for input *(Equivalent to Predicate)* | CT or MR input for contouring or registration/fusion. PET/CT input for registration/fusion only. DICOM RTSTRUCT and REGISTRATION for input | | Design: Reporting and data routing | No built-in reporting; supports exporting DICOM RTSTRUCT, REGISTRATION and DOSE files for output. *(Equivalent to Predicate)* | No built-in reporting; supports exporting DICOM RTSTRUCT, REGISTRATION and DOSE files for output. | | Compatibility with the environment and other devices | Compatible with data from any DICOM-compliant scanners for the applicable modalities. Agent Uploader component compatible with Microsoft Windows. Cloud-based automatic contouring service compatible with Linux. Web application server-based application compatible with Linux. *(Equivalent to Predicate)* | Compatible with data from any DICOM-compliant scanners for the applicable modalities. Agent Uploader component compatible with Microsoft Windows. Cloud-based automatic contouring service compatible with Linux. Web application server-based application compatible with Linux. | | Communications/ Networking | TCP/IP *(Equivalent to Predicate)* | TCP/IP | | Computer platform & Operating System | Windows Operating System *(Equivalent to Predicate)* | Windows Operating System | AutoContour Model RADAC V6 Special 510(k) 14 {18} ## 7. Performance Verification testing confirmed that AutoContour Model RADAC V6 features and functionality met predefined acceptance criteria consistent with the predicate device. AutoContour Model RADAC V6 models were divided into 3 categories and tested using protocols consistent with the category, while applying consistent acceptance criteria across each. Models were classified into 3 categories: 1) New Models that did not exist in the Predicate Device, 2) Models that previously existed in the Predicate Device but were updated in AutoContour Model RADAC V6, and 3) Models that previously existed in the Predicate Device and are unchanged. These classifications are noted in Table 6. ### Performance Data Testing demonstrates that AutoContour Model RADAC V6 performs as intended per its indications for use. Further tests were performed on independent datasets from those included in training and validation sets in order to validate the generalizability of the machine learning model. ### Description of Changes to Test Protocol Changes to the test protocol include an update to the new and existing structure model tests to include 95th Percentile Hausdorff Distance (HD95) as an additional metric for testing of new and updated model output changes. There were no changes to the regression testing protocol between Subject Device AutoContour RADAC V6 and Predicate Device AutoContour Model RADAC V5 for unchanged structure models, including the test methodology, acceptance criteria, and use of the HD95 metric. ### Testing Summary Mean Dice Similarity Coefficient (DSC) was used to validate the accuracy of structure model outputs when tested on image data sequestered from the original training data population. The test datasets were independent from those used for training and consisted of approximately 10% of the number of training image sets used as input for the model. For CT structure models, there were an average of 453 training and 61 testing image sets. Training and testing sets typically contain one scan per patient, but multiple scans per patient may be included if they are available and pertinent for the use case. CT training images were gathered from several institutions, in several different countries including the United States and Australia. Ground truthing of each test CT data set were generated manually using consensus (NRG/RTOG/ESTRO) guidelines as appropriate by six clinically experienced experts consisting of 2 radiation therapy physicists, 1 radiation dosimetrist, and 3 radiation therapists with specialized training in radiation therapy contouring. CT testing data spanned the most common radiation therapy treatment sites, including prostate, breast, lung, and head and neck cancers. Datasets were acquired across multiple scanner platforms (Philips Big Bore, GE Optima, GE Lightspeed, Siemens SOMATOM, and others) from numerous institutions in the United States, Canada, and Europe. Across all datasets, slice thicknesses ranged from 1–2.5 mm, in-plane resolutions from 0.3–1.5 mm, and acquisition voltages were predominantly 120 kVp. Imaging configurations reflect the range of equipment and protocols typically encountered in clinical AutoContour Model RADAC V6 Special 510(k) 15 {19} radiotherapy practice. The MR training dataset used for initial testing of the Brain models had an average of 362 training image sets and 69 testing image sets and were acquired from several different institutions in several countries. The MR training data used for initial testing of the MR Abdomen and Pelvis models had an average of 714 training image sets and 99 testing image sets and were taken from several institutions within several countries, such as the United States, the Netherlands, Canada, Australia, and France. Ground truthing of each test MR dataset was generated manually using consensus (NRG/RTOG) guidelines as appropriate by six clinically experienced experts consisting of 2 radiation therapy physicists, 1 radiation dosimetrist, and 3 radiation therapists with specialized training in radiation therapy contouring. For the Brain models, datasets were acquired as 20 MR T1 Ax post (BRAVO) image scans acquired with a GE MR750w scanner. Images had an average slice thickness of 1.6mm, In-plane resolution between 0.94 mm, and acquisition parameters of TR=5.98ms, TE=96.8s. Data for testing of the MR Pelvis structure models were acquired from 2 publicly available datasets, which contained images of patients with prostate or rectal cancer, as well as 1 dataset shared from 2 institutions utilizing an MR Linac. Various scanner models and acquisition settings were used. Data for testing of the MR Pelvis HDR structure models were acquired from 1 institution in Canada using two different slice thicknesses, 1mm and 4mm, and two different in-plane resolutions, 1mm and .72mm. Data for testing of the MR Abdomen structure models was acquired from 1 institution in the United States using a slice thickness of 3mm, and an in-plane resolution of 1.5mm. For all models, datasets used for testing were removed from the training dataset pool before model training began and used exclusively for testing to preserve independence and remove the possibility of bias. Datasets designated for testing are labeled with an exclusion tag within the data management system, which prevents these scans from being accessed or included during model training. This control ensures separation between training and testing datasets throughout the development process. ## Subgroup Analysis A subgroup analysis was performed on the external testing datasets for all new and updated structures across CT and MR modalities. Performance metrics were stratified by sex, patient position, and slice thickness. This analysis was performed exclusively on external testing data, as subgroup comparison on training data would not yield clinically meaningful insight into real-world algorithm performance. Not all subgroup categories are represented for every structure, as dataset availability and anatomical relevance determine which cohorts apply. Subgroup sample sizes are reported alongside all metrics to enable direct comparison of algorithm performance across patient demographics and imaging acquisition parameters. The following tables present a subgroup analysis of DSC test results for the new and updated structures. AutoContour Model RADAC V6 Special 510(k) 16 {20} | Table 4: Testing Dataset Characteristics - New Models | | | | | --- | --- | --- | --- | | # Datasets | 1215 | | | | Characteristic | Subgroup | Subgroup DSC | | | | | Mean | Std | | Gender | Male | 0.7650 | 0.1152 | | | Female | 0.7313 | 0.1401 | | | Unknown | 0.7622 | 0.1643 | | Position | Prone | 0.8807 | 0.1633 | | | Supine | 0.7537 | 0.1378 | | Slice Thickness (mm) | 0 - 1 | 0.7624 | 0.0833 | | | 1 - 2 | 0.7094 | 0.1635 | | | 2 - 3 | 0.7635 | 0.1581 | | | 3 - 4 | 0.7633 | 0.1310 | | | 7 - 8 | 0.4532 | 0.2447 | | Table 5: Testing Dataset Characteristics - Updated Models | | | | | --- | --- | --- | --- | | # Datasets | 1220 | | | | Characteristic | Subgroup | Subgroup DSC | | | | | Mean | Std | | Gender | Male | 0.9271 | 0.0498 | | | Female | 0.8679 | 0.0983 | | | Unknown | 0.8530 | 0.1121 | | Position | Prone | 0.7751 | 0.1885 | | | Supine | 0.8846 | 0.0890 | | Slice Thickness (mm) | 1 - 2 | 0.8368 | 0.1001 | | | 2 - 3 | 0.8582 | 0.0919 | | | 3 - 4 | 0.8880 | 0.0977 | AutoContour Model RADAC V6 Special 510(k) 17 {21} # Validation Results Summary All new and updated models in the Subject Device were validated using the same testing protocol (Dice Similarity Coefficient (DSC), Hausdorff distance (HD) and Likert Qualitative Review) with the addition of 95th Percentile Hausdorff Distance as an additional metric for regression testing of new and updated model output changes. There were no changes to the existing structure model testing protocol, and the same risk mitigations were enforced as utilized in the Predicate Device. Common acceptance criteria is used for validating model performance, although specific quantitative acceptance criteria vary depending on the volumetric size of the organ being evaluated. Also, the models in the subject device can be categorized into three different classifications. Each of these three classifications utilized a protocol appropriate for its historical attributes as follows: 1. There are new models with no prior reference, 2. models that are unchanged from the predicate device, 3. models that were modified. The following table provides a high-level description of the protocol used for each of the three model categories. | Table 6: Model Verification Categories and Acceptance Criteria | | | | --- | --- | --- | | New Models - not previously in AC | Updated models - previously in AC | Unchanged models from AC | | New Structure Model Validation: • Training DSC Evaluation* • External Dataset DSC Evaluation* • Internal Likert Qualitative Review** • External Likert Qualitative Review** • 95th Percentile Hausdorff Distance*** | Existing Structure Model Tests: • DSC Comparison between AutoContour v2.8 and AutoContour v2.7.*** • Hausdorff Distance between AutoContour v2.8 and AutoContour v2.7.*** New Structure Model Validation: • Training DSC Evaluation* • External Dataset DSC Evaluation* • Internal Likert Qualitative Review** • External Likert Qualitative Review** • 95th Percentile Hausdorff Distance*** | Existing Structure Model Tests: • DSC Comparison between AutoContour v2.8 and AutoContour v2.7.*** • Hausdorff Distance between AutoContour v2.8 and AutoContour v2.7.*** | The acceptance criteria for these protocols include the following values and ranges: * Training DSC and External Dataset DSC Evaluations: Mean DSC threshold for passing Large Structures: 0.80, Medium Structures: 0.65, and Small Structures: 0.50 ** Internal and External Likert Qualitative Review: Each structure model was determined to pass if the average grade exceeded 3 across all external image sets reviewed. AutoContour Model RADAC V6 Special 510(k) 18 {22} *** Existing Structure Model DSC Evaluation: The DSC threshold for passing Large Structures: >0.99, Medium Structures:>0.98, and Small Structures: >0.95. *** Existing Structure Model Hausdorff Distance Evaluation: The Hausdorff Distance threshold for passing structures is ≤ 3mm. *** 95th Percentile Hausdorff Distance Evaluation: The passing threshold is structure dependent and is based on structure variability. The following is the testing summary for AutoContour RADAC V6. Table 7: CT Training Data Results for AutoContour Model RADAC V6 | CT Structure | Size | Pass Criteria | # of Training Sets | # of Testing Sets | DSC (Avg) | DSC Std Dev | Lower Bound 95% Confidence Interval | | --- | --- | --- | --- | --- | --- | --- | --- | | A_Aorta** (Update) | Large | 0.80 | N/A | N/A | N/A* | N/A* | N/A* | | A_Aorta_Asc** (Update) | Medium | 0.65 | 1302 | 325 | 0.93 | 0.12 | 0.74 | | A_Aorta_Dsc** (Update) | Medium | 0.65 | 1524 | 388 | 0.93 | 0.11 | 0.73 | | A_Brachiocephls** (Update) | Small | 0.50 | 388 | 97 | 0.88 | 0.16 | 0.61 | | A_LAD** (Update) | Small | 0.50 | 640 | 161 | 0.60 | 0.14 | 0.38 | | A_Subclavian_L** (Update) | Small | 0.50 | 388 | 97 | 0.86 | 0.17 | 0.59 | | A_Subclavian_R** (Update) | Small | 0.50 | 388 | 97 | 0.89 | 0.14 | 0.66 | | BileDuct_Common** (Update) | Small | 0.50 | 643 | 162 | 0.57 | 0.20 | 0.24 | | Bone_Mandible** (Update) | Medium | 0.65 | 976 | 100 | 0.90 | 0.04 | 0.84 | | Bone_Teeth** (Update) | Medium | 0.65 | 340 | 76 | 0.88 | 0.02 | 0.84 | | Bowel_Bag** (Update) | Large | 0.80 | 454 | 48 | 0.95 | 0.05 | 0.87 | | Bowel_Bag_F** (Update) | Large | 0.80 | 454 | 48 | 0.95 | 0.05 | 0.87 | | Breast_Prone_RTOG | Large | 0.80 | 125 | 30 | 0.96 | 0.02 | 0.93 | | BuccalMucosa** (Update) | Medium | 0.65 | 908 | 50 | 0.78 | 0.24 | 0.38 | | Chestwall_L** (Update) | Large | 0.80 | 79 | 20 | 0.90 | 0.03 | 0.84 | | Chestwall_OAR** (Update) | Large | 0.80 | 118 | 30 | 0.90 | 0.03 | 0.86 | | Chestwall_R** (Update) | Large | 0.80 | 79 | 20 | 0.90 | 0.03 | 0.84 | | Chestwall_RC_L** (Update) | Large | 0.80 | 80 | 20 | 0.91 | 0.04 | 0.84 | | Chestwall_RC_R** (Update) | Large | 0.80 | 80 | 20 | 0.91 | 0.04 | 0.84 | | Dental_Artifact** (Update) | Medium | 0.65 | 342 | 86 | 0.76 | 0.10 | 0.600 | | Gallbladder** (Update) | Medium | 0.65 | 284 | 71 | 0.95 | 0.11 | 0.77 | | GreatVessels** (Update) | Large | 0.80 | N/A | N/A | N/A* | N/A* | N/A* | | HDR_Bowel** (Update) | Large | 0.80 | 226 | 57 | 0.90 | 0.03 | 0.86 | | HDR_Canal_Anal** (Update) | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | | HDR_Colon_Sigmoid** (Update) | Medium | 0.65 | 927 | 98 | 0.85 | 0.16 | 0.58 | AutoContour Model RADAC V6 Special 510(k) 19 {23} | HDR_Rectum** (Update) | Medium | 0.65 | 1022 | 50 | 0.83 | 0.18 | 0.54 | | --- | --- | --- | --- | --- | --- | --- | --- | | HDR_UteroCervix | Medium | 0.65 | 642 | 79 | 0.93 | 0.21 | 0.60 | | Heart_Prone** (Update) | Large | 0.80 | 308 | 78 | 0.97 | 0.01 | 0.95 | | Liver** (Update) | Large | 0.80 | 511 | 168 | 0.89 | 0.06 | 0.79 | | Lips** (Update) | Medium | 0.65 | 908 | 50 | 0.82 | 0.23 | 0.43 | | LN_Mediast_10 | Small | 0.50 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_11 | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_1L | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_1R | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_2L | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_2R | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_3A | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_3P | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_4L | Small | 0.50 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_4R | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_5 | Small | 0.50 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_6 | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_7 | Small | 0.50 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Mediast_8 | Medium | 0.65 | 470 | 44 | 0.88 | 0.09 | 0.73 | | LN_Neck_CerVII_2d_L** (Update) | Large | 0.80 | 672 | 50 | 0.85 | 0.18 | 0.55 | | LN_Neck_CerVII_2d_R** (Update) | Large | 0.80 | 672 | 50 | 0.85 | 0.18 | 0.55 | | LN_Neck_II-IV_2d_L** (Update) | Large | 0.80 | 672 | 50 | 0.85 | 0.18 | 0.55 | | LN_Neck_II-IV_2d_R** (Update) | Large | 0.80 | 672 | 50 | 0.85 | 0.18 | 0.55 | | LN_Neck_IV_2d_L** (Update) | Medium | 0.65 | 672 | 50 | 0.85 | 0.18 | 0.55 | | LN_Neck_IV_2d_R** (Update) | Medium | 0.65 | 672 | 50 | 0.85 | 0.18 | 0.55 | | LN_Neck_IVA_2d_L | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | | LN_Neck_IVA_2d_R | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | | LN_Neck_IVB_2d_L | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | | LN_Neck_IVB_2d_R | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | | LN_Neck_VC_2d_L | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | | LN_Neck_VC_2d_R | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | | LN_Paraaortic (Update) | Large | 0.80 | 200 | 50 | 0.89 | 0.04 | 0.82 | | LN_Rectum_EIN_L | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | | LN_Rectum_EIN_R | Medium | 0.65 | N/A | N/A | N/A* | N/A* | N/A* | AutoContour Model RADAC V6 Special 510(k) 20 {24} | Lung_L** (Update) | Large | 0.80 | 1082 | 65 | 0.97 | 0.05 | 0.89 | | --- | --- | --- | --- | --- | --- | --- | --- | | Lung_R** (Update) | Large | 0.80 | 1082 | 65 | 0.98 | 0.03 | 0.93 | | Lungs** (Update) | Large | 0.80 | N/A | N/A | N/A* | N/A* | N/A* | | Pericardium** (Update) | Large | 0.80 | 160 | 41 | 0.94 | 0.02 | 0.91 | | V_Brachioceph_L** (Update) | Medium | 0.65 | 388 | 97 | 0.91 | 0.10 | 0.74 | | V_Brachioceph_R** (Update) | Small | 0.50 | 388 | 97 | 0.86 | 0.19 | 0.55 | *N/A: Structures are generated based on a post-processing/boolean operation from previously released structure models (Eye, Rib) rather than generated from a CNN model. Quantitative and Qualitative testing for these structures are still performed in the following sections in order to validate appropriate contour generation and clinical acceptability. ** A_Aorta, A_Aorta_Asc, A_Aorta_Dsc, A_Brachiocephls, A_LAD, A_Subclavian_L, A_Subclavian_R, BileDuct_Common, Bone_Mandible, Bone_Teeth, Bowel_Bag, Bowel_Bag_F, BuccalMucosa, Chestwall_L, Chestwall_R, Chestwall_OAR, Chestwall_RC_L, Chestwall_RC_R, Dental_Artifact, Gallbladder, GreatVessels, HDR_Bowel, HDR_Canal_Anal, HDR_Colon_Sigmoid, HDR_Rectum, Heart_Prone, Liver, Lips, LN_Neck_CerVII_2d_L, LN_Neck_CerVII_2d_R, LN_Neck_II-IV_2d_L, LN_Neck_II-IV_2d_R, LN_Neck_IV_2d_L, LN_Neck_IV_2d_R, LN_Paraortic, Lung_L, Lung_R, Lungs, Pericardium, V_Brachioceph_L, and V_Brachioceph_R models were previously released/tested, but updates to the training datasets and contour output necessitated additional release testing for these models. Additional external clinical testing was performed in order to validate the accuracy of the models on image sets acquired that were unique to the training datasets. Both AutoContour and manually added ground truth contours following the same structure guidelines used for structure model training were added to the image sets. | Table 8: CT External Clinical Dataset References | | | | --- | --- | --- | | Model Group | Data Source ID | Data Citation | | CT Pelvis | TCIA - Pelvic-Ref | Afua A. Yorke, Gary C. McDonald, David Solis Jr., Thomas Guerrero. (2019) Pelvic Reference Data. The Cancer Imaging Archive. DOI: 10.7937/TCIA.2019.woskq5oo | | CT Head and Neck | TCIA - Head-Neck-PET-CT | Martin Vallières, Emily Kay-Rivest, Léo Jean Perrin, Xavier Liem, Christophe Furstoss, Nader Khaouam, Phuc Félix Nguyen-Tan, Chang-Shu Wang, Khalil Sultanem. (2017). Data from Head-Neck-PET-CT. The Cancer Imaging Archive. doi: 10.7937/K9/TCIA.2017.8oje5q00 | | CT Abdomen | TCIA - Pancreas-CT-CB | Hong, J., Reyngold, M., Crane, C., Cuaron, J., Hajj, C., Mann, J., Zinovoy, M., Yorke, E., LoCastro, E., Apte, A. P., & Mageras, G. (2021). Breath-hold CT and cone-beam CT images with expert manual organ-at-risk segmentations from radiation treatments of locally advanced pancreatic cancer [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/TCIA.ESHQ-4D90 | | CT Thorax: | TCIA - NSCLC | Aerts, H. J. W. L., Wee, L., Rios Velazquez, E., Leijenaar, R. T. H., Parmar, C., Grossmann, P., Carvalho, S., Bussink, J., Monshouwer, R., Haibe-Kains, B., Rietveld, D., Hoebers, F., Rietbergen, M. M., Leemans, C. R., Dekker, A., | AutoContour Model RADAC V6 Special 510(k) 21 {25} | | | Quackenbush, J., Gillies, R. J., Lambin, P. (2019). Data From NSCLC-Radiomics [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/K9/TCIA.2015.PF0M9REI | | --- | --- | --- | | CT Thorax | TCIA - LCTSC | Yang, J., Sharp, G., Veeraraghavan, H., Van Elmpt, W., Dekker, A., Lustberg, T., & Gooding, M. (2017). Data from Lung CT Segmentation Challenge (Version 3) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/K9/TCIA.2017.3R3FVZ08 | | CT Thorax Prone Female | TCIA- QIN-BREAST and Prone Thorax | Li, X., Abramson, R. G., Arlinghaus, L. R., Chakravarthy, A. B., Abramson, V. G., Sanders, M., & Yankeelov, T. E. (2016). Data From QIN-BREAST (Version 2) [Data set]. The Cancer Imaging Archive. https://doi.org/10.7937/K9/TCIA.2016.21JUEBH0N/A- Testing data was shared from several institutions | | CT HDR Female | Female HDR Pelvis | N/A- Testing data was shared from 2 different institutions based in the United States. | | CT Prostatectomy | Pelvis ProstateBed | N/A- Testing data was shared from 1 institution based in the United States | | CT Pelvis Barrigel | BarrigelTM | N/A- Testing data was shared from several institutions in Australia | | CT Pelvis SpaceOARVUE | SpaceOARTM_VUE | N/A- Testing data was shared from several institutions | DSC values were calculated between ground truth contour data and AutoContour structures and rated on the same DSC passing criteria used for the Training DSC validation. All structures passed the minimum DSC criteria for small, medium, and large structures with a mean DSC of 0.75+/-0.09, 0.80+/-0.09, and 0.92+/-0.04, respectively. In addition to the DSC value calculations, the 95th Percentile Hausdorff Distance was also calculated on the ground truth contour data, results in Table 5. The qualitative clinical appropriateness of AutoContour structures generated on these scans was graded by clinical experts. Autocontour structures were graded on a scale from 1 to 5, where 5 refers to a contour requiring no additional edits, and 1 refers to a score in which full manual re-contour of the structure would be required. An average score >= 3 was used to determine whether a structure model would ultimately be beneficial clinically. An average rating of 4.37 was found across all CT structure models, demonstrating that only minor edits would be required in order to make the structure models acceptable for clinical use. Three structures, LN_Neck_VC_2d_L, LN_Neck_VC_2d_R, and LN_Rectum_EIN_R were passed according to predefined qualitative clinical review criteria, with scores of 4.4, 4.4 and 4.45, respectively. The remaining structures were accepted based on their HD95 verification results. AutoContour Model RADAC V6 Special 510(k) 22 {26} **Table 9: CT External Reviewer Results for AutoContour Model RADAC V6** | CT Structure | Size | Pass Criteria | #Testing Sets | Average DSC | Average DSC Std. Dev | Lower Bound 95% Confidence Interval | External Reviewer Average Rating (1-5) | | --- | --- | --- | --- | --- | --- | --- | --- | | A_Aorta (Update) | Large | 0.80 | 40 | 0.92 | 0.01 | 0.90 | 5 | | A_Aorta_Asc (Update) | Medium | 0.65 | 40 | 0.92 | 0.02 | 0.89 | 5 | | A_Aorta_Dsc (Update) | Medium | 0.65 | 40 | 0.92 | 0.02 | 0.89 | 5 | | A_Brachiocephls (Update) | Small | 0.50 | 40 | 0.86 | 0.12 | 0.67 | 4.5 | | A_LAD (Update) | Small | 0.50 | 41 | 0.63 | 0.09 | 0.49 | 4.5 | | A_Subclavian_L (Update) | Small | 0.50 | 40 | 0.85 | 0.07 | 0.74 | 4.5 | | A_Subclavian_R (Update) | Small | 0.50 | 40 | 0.87 | 0.06 | 0.76 | 4.5 | | BileDuct_Common (Update) | Small | 0.50 | 25 | 0.66 | 0.12 | 0.76 | 4.5 | | Bone_Mandible (Update) | Medium | 0.65 | 22 | 0.90 | 0.21 | 0.56 | 4.7 | | Bone_Teeth (Update) | Medium | 0.65 | 9 | 0.87 | 0.02 | 0.84 | 4.7 | | Bowel_Bag (Update) | Large | 0.80 | 46 | 0.90 | 0.06 | 0.81 | 4.35 | | Bowel_Bag_F (Update) | Large | 0.80 | 20 | 0.89 | 0.05 | 0.81 | 4.1 | | Breast_Prone_RTOG | Large | 0.80 | 21 | 0.94 | 0.07 | 0.82 | 4 | | BuccalMucosa (Update) | Medium | 0.65 | 42 | 0.77 | 0.05 | 0.69 | 4.4 | | Chestwall_L (Update) | Large | 0.80 | 20 | 0.88 | 0.09 | 0.74 | 4.1 | | Chestwall_OAR (Update) | Large | 0.80 | 20 | 0.94 | 0.01 | 0.92 | 4 | | Chestwall_R (Update) | Large | 0.80 | 20 | 0.89 | 0.06 | 0.78 | 4 | | Chestwall_RC_L (Update) | Large | 0.80 | 20 | 0.91 | 0.05 | 0.83 | 4 | | Chestwall_RC_R (Update) | Large | 0.80 | 20 | 0.90 | 0.04 | 0.84 | 4.2 | | Dental_Artifact (Update) | Medium | 0.65 | 7 | 0.86 | 0.08 | 0.73 | 4.6 | | Gallbladder (Update) | Medium | 0.65 | 21 | 0.86 | 0.05 | 0.78 | 5 | | GreatVessels (Update) | Large | 0.80 | 20 | 0.91 | 0.01 | 0.89 | 4.2 | | HDR_Bowel (Update) | Large | 0.80 | 21 | 0.89 | 0.07 | 0.78 | 4.4 | | HDR_Canal_Anal (Update) | Medium | 0.65 | 20 | 0.78 | 0.09 | 0.63 | 4.4 | | HDR_Colon_Sigmoid (Update) | Medium | 0.65 | 20 | 0.73 | 0.13 | 0.52 | 4.4 | | HDR_Rectum (Update) | Medium | 0.65 | 20 | 0.85 | 0.07 | 0.74 | 4.4 | | HDR_UteroCervix | Medium | 0.65 | 13 | 0.88 | 0.05 | 0.81 | 4.3 | | Heart_Prone (Update) | Large | 0.80 | 21 | 0.95 | 0.02 | 0.91 | 4.8 | | Liver (Update) | Large | 0.80 | 25 | 0.97 | 0.01 | 0.95 | 4.8 | | Lips (Update) | Medium | 0.65 | 22 | 0.81 | 0.06 | 0.71 | 5 | AutoContour Model RADAC V6 Special 510(k) 23 {27} | LN_Mediast_10 | Small | 0.50 | 43 | 0.67 | 0.07 | 0.56 | 3.8 | | --- | --- | --- | --- | --- | --- | --- | --- | | LN_Mediast_11 | Medium | 0.65 | 43 | 0.71 | 0.09 | 0.58 | 4.4 | | LN_Mediast_1L | Medium | 0.65 | 43 | 0.84 | 0.06 | 0.74 | 4.4 | | LN_Mediast_1R | Medium | 0.65 | 43 | 0.84 | 0.06 | 0.74 | 4 | | LN_Mediast_2L | Medium | 0.65 | 43 | 0.80 | 0.09 | 0.66 | 4.4 | | LN_Mediast_2R | Medium | 0.65 | 43 | 0.81 | 0.09 | 0.67 | 4 | | LN_Mediast_3A | Medium | 0.65 | 43 | 0.78 | 0.14 | 0.56 | 3.9 | | LN_Mediast_3P | Medium | 0.65 | 43 | 0.69 | 0.14 | 0.45 | 4.1 | | LN_Mediast_4L | Small | 0.50 | 43 | 0.72 | 0.06 | 0.63 | 4.5 | | LN_Mediast_4R | Medium | 0.65 | 43 | 0.82 | 0.07 | 0.70 | 4.6 | | LN_Mediast_5 | Small | 0.50 | 42 | 0.69 | 0.16 | 0.43 | 4.3 | | LN_Mediast_6 | Medium | 0.65 | 43 | 0.67 | 0.16 | 0.41 | 4.4 | | LN_Mediast_7 | Small | 0.50 | 43 | 0.71 | 0.08 | 0.58 | 4.4 | | LN_Mediast_8 | Medium | 0.65 | 43 | 0.71 | 0.06 | 0.61 | 3.9 | | LN_Neck_CerVII_2d_L (Update) | Large | 0.80 | 22 | 0.91 | 0.04 | 0.84 | 4 | | LN_Neck_CerVII_2d_R (Update) | Large | 0.80 | 22 | 0.90 | 0.04 | 0.84 | 4 | | LN_Neck_II-IV_2d_L (Update) | Large | 0.80 | 22 | 0.91 | 0.04 | 0.84 | 4 | | LN_Neck_II-IV_2d_R (Update) | Large | 0.80 | 22 | 0.90 | 0.04 | 0.83 | 4 | | LN_Neck_IV_2d_L (Update) | Medium | 0.65 | 22 | 0.89 | 0.04 | 0.83 | 4 | | LN_Neck_IV_2d_R (Update) | Medium | 0.65 | 22 | 0.86 | 0.05 | 0.78 | 4.1 | | LN_Neck_IVA_2d_L | Medium | 0.65 | 22 | 0.86 | 0.05 | 0.77 | 4.4 | | LN_Neck_IVA_2d_R | Medium | 0.65 | 22 | 0.82 | 0.07 | 0.70 | 4.5 | | LN_Neck_IVB_2d_L | Medium | 0.65 | 22 | 0.81 | 0.05 | 0.73 | 4 | | LN_Neck_IVB_2d_R | Medium | 0.65 | 22 | 0.76 | 0.05 | 0.68 | 4 | | LN_Neck_VC_2d_L | Medium | 0.65 | 22 | 0.67 | 0.21 | 0.32 | 4.4 | | LN_Neck_VC_2d_R | Medium | 0.65 | 22 | 0.65 | 0.27 | 0.20 | 4.4 | | LN_Rectum_EIN_L | Medium | 0.65 | 23 | 0.89 | 0.05 | 0.81 | 4.5 | | LN_Rectum_EIN_R | Medium | 0.65 | 41 | 0.70 | 0.08 | 0.58 | 4.45 | | LN_Paraaortic (Update) | Large | 0.80 | 41 | 0.70 | 0.12 | 0.51 | 4.45 | | Lung_L (Update) | Large | 0.80 | 25 | 0.98 | 0.01 | 0.98 | 4.9 | | Lung_R (Update) | Large | 0.80 | 25 | 0.99 | 0.01 | 0.98 | 4.9 | | Lungs (Update) | Large | 0.80 | 20 | 0.98 | 0.01 | 0.97 | 4.9 | | Pericardium (Update) | Large | 0.80 | 20 | 0.98 | 0.01 | 0.97 | 4 | AutoContour Model RADAC V6 Special 510(k) 24 {28} | V_Brachioceph_L (Update) | Medium | 0.65 | 40 | 0.87 | 0.10 | 0.71 | 4.5 | | --- | --- | --- | --- | --- | --- | --- | --- | | V_Brachioceph_R (Update) | Small | 0.50 | 40 | 0.89 | 0.06 | 0.80 | 4.4 | AutoContour Model RADAC V6 Special 510(k) 25 {29} | **MR Models** | **Pass Criteria** | **HD95** | | --- | --- | --- | | A_Aorta (Update) | 33.40 | 2.93 | | A_Aorta_Asc (Update) | 20.30 | 3.08 | | A_Aorta_Dsc (Update) | 34.68 | 2.96 | | A_Brachiocephls (Update) | 10.38 | 4.52 | | A_LAD (Update) | 15.88 | 7.21 | | A_Subclavian_L (Update) | 16.71 | 9.01 | | A_Subclavian_R (Update) | 16.22 | 5.32 | | BileDuct_Common (Update) | 14.051 | 7.78 | | Bone_Mandible (Update) | 12.03 | 1.74 | | Bone_Teeth (Update) | 10.03 | 1.91 | | Bowel_Bag (Update) | 65.41 | 19.38 | | Bowel_Bag_F (Update) | 37.16 | 21.59 | | Breast_Prone_RTOG | 42.75 | 13.67 | | BuccalMucosa (Update) | 8.87 | 3.38 | | Chestwall_L (Update) | 25.53 | 17.67 | | Chestwall_OAR (Update) | 32.63 | 8.04 | | Chestwall_R (Update) | 21.92 | 20.23 | | Chestwall_RC_L (Update) | 25.14 | 15.94 | | Chestwall_RC_R (Update) | 21.26 | 16.93 | | Dental_Artifact (Update) | 15.238 | 3.59 | | Gallbladder (Update) | 17.98 | 5.99 | | GreatVessels (Update) | 19.68 | 3.87 | | HDR_Bowel (Update) | 31.182 | 5.17 | | HDR_Canal_Anal (Update) | 4.49 | 3.89 | | HDR_Colon_Sigmoid (Update) | 35.13 | 16.65 | | HDR_Rectum (Update) | 17.13 | 8.19 | | HDR_UteroCervix | 15.72 | 7.39 | | Heart_Prone (Update) | 10.70 | 5.33 | | Liver (Update) | 31.00 | 4.33 | | Lips (Update) | 8.17 | 3.82 | AutoContour Model RADAC V6 Special 510(k) 26 {30} | LN_Mediast_10 | 10.083 | 4.8 | | --- | --- | --- | | LN_Mediast_11 | 20.83 | 10.99 | | LN_Mediast_1L | 10.43 | 4.16 | | LN_Mediast_1R | 10.24 | 4.48 | | LN_Mediast_2L | 9.228 | 3.98 | | LN_Mediast_2R | 10.87 | 4.5 | | LN_Mediast_3A | 16.71 | 6.3 | | LN_Mediast_3P | 19.19 | 8.11 | | LN_Mediast_4L | 10.46 | 4.34 | | LN_Mediast_4R | 11.51 | 4.06 | | LN_Mediast_5 | 8.92 | 5.46 | | LN_Mediast_6 | 12.09 | 5.88 | | LN_Mediast_7 | 11.87 | 5.59 | | LN_Mediast_8 | 16.63 | 8.67 | | LN_Neck_CerVII_2d_L (Update) | 11.46 | 4.25 | | LN_Neck_CerVII_2d_R (Update) | 11.98 | 4.77 | | LN_Neck_II-IV_2d_L (Update) | 11.67 | 4.72 | | LN_Neck_II-IV_2d_R (Update) | 12.13 | 4.89 | | LN_Neck_IV_2d_L (Update) | 8.83 | 4.25 | | LN_Neck_IV_2d_R (Update) | 9.07 | 5.46 | | LN_Neck_IVA_2d_L | 9.50 | 5.88 | | LN_Neck_IVA_2d_R | 8.42 | 6.62 | | LN_Neck_IVB_2d_L | 8.62 | 5.12 | | LN_Neck_IVB_2d_R | 7.87 | 5.9 | | LN_Neck_VC_2d_L | 8.58 | 10.5 | | LN_Neck_VC_2d_R | 9.36 | 10.96 | | LN_Paraaortic (Update) | 14.41 | 9.56 | | LN_Rectum_EIN_L | 12.35 | 12.34 | | LN_Rectum_EIN_R | 12.43 | 14.96 | | Lung_L (Update) | 31.04 | 2.63 | | Lung_R (Update) | 33.54 | 2.47 | | Lungs (Update) | 31.86 | 2.78 | | Pericardium (Update) | 19.96 | 4.4 | AutoContour Model RADAC V6 Special 510(k) 27 {31} | V_Brachioceph_L (Update) | 12.18 | 2.52 | | --- | --- | --- | | V_Brachioceph_R (Update) | 8.33 | 3.33 | The MR training data set used for initial testing of the Brain models (CaudateNucleus_L/R, Fornix_L/R, Glnd_Lacrimal_L/R, Periventricular_Space, and Pineal) had an average of 362 training image sets and 69 testing image sets and were acquired from several different institutions in several countries. The MR training data used for initial testing of the MR Abdomen models (Bowel, Duodenum, Esophagus, Gallbladder, Kidney_L/R, Liver, Pancreas, Spleen, Stomach, and ThecalSac) had an average of 714 training image sets and 99 testing image sets and were taken from several institutions within several countries, such as the United States, the Netherlands, and France. Datasets used for testing were removed from the training dataset pool before model training began and used exclusively for testing. Ground truthing of each test dataset was generated manually using consensus (NRG/RTOG) guidelines as appropriate by six clinically experienced experts consisting of 2 radiation therapy physicists, 1 radiation dosimetrist, and 3 radiation therapists with specialized training in radiation therapy contouring. For MR Structure models, a mean training DSC of 0.90+/-0.13 was found for large models, 0.86+/-0.15 for medium models, and 0.75+/- 0.12 for small models. All structures passed the minimum training DSC criteria and one structure passed all other tests aside from DSC, including with a qualitative clinical external review with a score of 3.92, so is deemed clinically acceptable. | Table 11: MR Training Data Results for AutoContour Model RADAC V6 | | | | | | | | | --- | --- | --- | --- | --- | --- | --- | --- | | MR Models | Size | Pass Criteria | # of Training Sets | # of Testing Sets | DSC (Avg) | DSC Std Dev (Avg) | Lower Bound 95% Confidence Interval | | Bowel | Large | 0.80 | 769 | 84 | 0.77 | 0.10 | 0.61 | | CaudateNucleus_L | Small | 0.50 | 370 | 42 | 0.82 | 0.25 | 0.40 | | CaudateNucleus_R | Small | 0.50 | 370 | 42 | 0.82 | 0.25 | 0.40 | | Duodenum | Medium | 0.65 | 502 | 120 | 0.82 | 0.20 | 0.50 | | Esophagus | Small | 0.50 | 787 | 89 | 0.82 | 0.11 | 0.64 | | Fornix_L | Small | 0.50 | 415 | 80 | 0.60 | 0.12 | 0.40 | | Fornix_R | Small | 0.50 | 415 | 80 | 0.61 | 0.11 | 0.43 | | Gallbladder | Medium | 0.65 | 502 | 120 | 0.93 | 0.11 | 0.74 | | Glnd_Lacrimal_L | Small | 0.50 | 318 | 79 | 0.73 | 0.07 | 0.62 | | Glnd_Lacrimal_R | Small | 0.50 | 318 | 79 | 0.73 | 0.07 | 0.62 | | Kidney_L | Large | 0.80 | 850 | 97 | 0.92 | 0.01 | 0.91 | | Kidney_R | Large | 0.80 | 850 | 97 | 0.92 | 0.01 | 0.91 | AutoContour Model RADAC V6 Special 510(k) 28 {32} | Liver | Large | 0.80 | 898 | 51 | 0.91 | 0.17 | 0.63 | | --- | --- | --- | --- | --- | --- | --- | --- | | Pancreas | Medium | 0.65 | 502 | 120 | 0.84 | 0.16 | 0.57 | | Periventricular_Space* | Medium | 0.65 | N/A | N/A | N/A | N/A | N/A | | Pineal | Small | 0.50 | 331 | 81 | 0.67 | 0.17 | 0.39 | | Spleen | Large | 0.80 | 842 | 97 | 0.95 | 0.14 | 0.71 | | Stomach | Large | 0.80 | 502 | 120 | 0.89 | 0.15 | 0.65 | | ThecalSac | Medium | 0.65 | 851 | 96 | 0.84 | 0.16 | 0.57 | *These structures were generated from post-processing operations from previously released or re-tested models, Brain_Ventricle, rather than from a CNN model. Qualitative and Quantitative analysis of these structures is still performed. Additional external clinical testing was performed in order to validate the accuracy of the models on image sets acquired that were unique to the training datasets. | Table 12: MR External Clinical Dataset References | | | | --- | --- | --- | | Model Group | Data Source ID | Data Citation | | MR Brain | MR - Renown | N/A | | MR Pelvis | Gold Atlas Pelvis | Nyholm, Tufve, Stina Svensson, Sebastian Andersson, Joakim Jonsson, Maja Sohlin, Christian Gustafsson, Elisabeth Kjellén, et al. 2018. “MR and CT Data with Multi Observer Delineations of Organs in the Pelvic Area - Part of the Gold Atlas Project.” Medical Physics 12 (10): 3218–21. doi:10.1002/mp.12748. | | MR Pelvis_2 | SynthRad | Thummerer A, van der Bijl E, Galapon Jr A, Verhoeff JJ, Langendijk JA, Both S, van den Berg CAT, Maspero M. 2023. SynthRAD2023 Grand Challenge dataset: Generating synthetic CT for radiotherapy. Medical Physics, 50(7), 4664-4674. https://doi.org/10.1002/mp.16529 | | MRLinac Pelvis | MR Linac | N/A- Testing data was shared by 2 institutions utilizing MR Linacs for image acquisitions. | | MR Female HDR Brachy | Female HDR MR Pelvis | N/A- Testing data was shared by 1 institution in Canada | | MR Pelvis Barrigel | Barrigel | N/A- Testing data was shared by several institutions in Australia. | | MR Abdomen | MR_Abdomen | N/A- Testing data was shared by 1 institution in the United States | For the Brain models, datasets acquired via data-use agreement from a clinical partner were acquired containing 20 MR T1 Ax post (BRAVO) image scans acquired with a GE MR750w scanner. Images had an average slice thickness of 1.6mm, In-plane resolution between 0.94 mm, and acquisition parameters of TR=5.98ms, TE=96.8s. Data for testing of the MR Pelvis structure models were acquired from 2 publicly available datasets, which contained images of patients with prostate or rectal cancer, as well as 1 dataset shared from 2 institutions utilizing an MR Linac. Various scanner models and acquisition settings were used. Data for testing of the MR Pelvis HDR structure models were acquired from 1 institution using two different slice thicknesses, 1mm and 4mm, and two different in-plane resolutions, 1mm and .72mm. Data for AutoContour Model RADAC V6 Special 510(k) 29 {33} testing of the MR Abdomen structure models was acquired from 1 institution in the United States using a slice thickness of 3mm, and an in-plane resolution of 1.5mm. DSC values were calculated between ground truth contour data and AutoContour structures and rated on the same DSC passing criteria as was used for the training DSC validation. All structures passed the minimum DSC criteria for small, medium, and large structures with a mean DSC of 0.67+/-0.13, 0.76+/-0.08, and 0.92+/-0.03, respectively. In addition to the DSC value calculations, the 95th Percentile Hausdorff Distance was also calculated on the ground truth contour data, results in Table 9. The qualitative clinical appropriateness of AutoContour structures generated on these scans was graded by clinical experts. AutoContour structures were graded on a scale from 1 to 5, where 5 refers to a contour requiring no additional edits, and 1 refers to a score in which full manual re-contour of the structure would be required. Four structures passed the HD95 validation. Esophagus, Glnd_Lacrimal_L, Glnd_Lacrimal_R, and Pineal were passed according to predefined qualitative clinical review criteria, with scores of 3.4, 5, 5, and 4.9, respectively. An average score >= 3 was used to determine whether a structure model would ultimately be beneficial clinically. An average rating of 4.4 was found across all MR structure models, demonstrating that only minor edits would be required in order to make the structure models acceptable for clinical use. Table 13: MR External Reviewer Results for AutoContour Model RADAC V6 | MR Models | Size | Pass Criteria | # External Test Data Sets | Average DSC | Average DSC Std. Dev | Lower Bound 95% Confidence Interval | External Reviewer Average Rating (1-5) | | --- | --- | --- | --- | --- | --- | --- | --- | | Bowel | Large | 0.80 | 18 | 0.87 | 0.03 | 0.82 | 3.92 | | CaudateNucleus_L | Small | 0.50 | 20 | 0.78 | 0.09 | 0.63 | 3.98 | | CaudateNucleus_R | Small | 0.50 | 20 | 0.77 | 0.12 | 0.58 | 3.88 | | Duodenum | Medium | 0.65 | 18 | 0.70 | 0.10 | 0.54 | 3.78 | | Esophagus | Small | 0.50 | 18 | 0.59 | 0.09 | 0.44 | 3.80 | | Fornix_L | Small | 0.50 | 20 | 0.61 | 0.09 | 0.46 | 3.50 | | Fornix_R | Small | 0.50 | 20 | 0.63 | 0.07 | 0.51 | 3.77 | | Gallbladder | Medium | 0.65 | 11 | 0.79 | 0.11 | 0.61 | 3.50 | | Glnd_Lacrimal_L | Small | 0.50 | 20 | 0.70 | 0.14 | 0.46 | 4.10 | | Glnd_Lacrimal_R | Small | 0.50 | 20 | 0.66 | 0.23 | 0.27 | 3.90 | | Kidney_L | Large | 0.80 | 18 | 0.94 | 0.01 | 0.93 | 4.39 | | Kidney_R | Large | 0.80 | 18 | 0.93 | 0.02 | 0.90 | 4.11 | | Liver | Large | 0.80 | 18 | 0.94 | 0.02 | 0.91 | 3.94 | | Pancreas | Medium | 0.65 | 18 | 0.67 | 0.09 | 0.51 | 3.75 | | Periventricular_Space | Medium | 0.65 | 20 | 0.81 | 0.05 | 0.73 | 4.20 | | Pineal | Small | 0.50 | 20 | 0.60 | 0.21 | 0.26 | 3.78 | AutoContour Model RADAC V6 Special 510(k) 30 {34} | Spleen | Large | 0.80 | 18 | 0.93 | 0.02 | 0.90 | 3.72 | | --- | --- | --- | --- | --- | --- | --- | --- | | Stomach | Large | 0.80 | 18 | 0.88 | 0.07 | 0.78 | 3.75 | | ThecalSac | Medium | 0.65 | 18 | 0.83 | 0.04 | 0.76 | 4.08 | **Table 14: MR HD95 Results for AutoContour Model RADAC V6** | MR Models | Pass Criteria | HD95 | | --- | --- | --- | | Bowel | 59.20 | 5.8 | | CaudateNucleus_L | 7.34 | 4.13 | | CaudateNucleus_R | 7.17 | 4.16 | | Duodenum | 23.46 | 15.48 | | Esophagus | 20.35 | 25.22 | | Fornix_L | 5.68 | 3.27 | | Fornix_R | 5.52 | 3.05 | | Gallbladder | 17.81 | 7.3 | | Glnd_Lacrimal_L | 2.57 | 2.63 | | Glnd_Lacrimal_R | 2.55 | 3.52 | | Kidney_L | 13.30 | 3.86 | | Kidney_R | 11.24 | 3.87 | | Liver | 24.99 | 7.78 | | Pancreas | 25.17 | 13.38 | | Periventricular_Space | 11.28 | 2.21 | | Pineal | 1.94 | 2.83 | | Spleen | 23.35 | 3.93 | | Stomach | 28.56 | 10.32 | | ThecalSac | 24.33 | 12.81 | ### Validation Conclusion The outcomes of performance verification establish that AutoContour operates reliably across the diversity of anatomical presentations and clinical conditions characteristic of the intended patient demographic and standard radiotherapy practice. Evaluation of algorithm performance, as assessed by qualified radiation therapy professionals, indicates that the automated segmentation functionality meaningfully decreases the baseline time and effort otherwise dedicated to manual contour delineation while providing clinically acceptable contours. AutoContour Model RADAC V6 Special 510(k) 31 {35} Additionally, the validation evidence supports that the device consistently produces structure sets of sufficient clinical quality. ## **8. Conclusion** AutoContour RADAC V6 is deemed substantially equivalent to the primary Predicate Device AutoContour RADAC V5 (K260509). Verification tests were performed to ensure that the software works as intended, and pass/fail criteria were used to verify requirements. Validation testing was performed to ensure that the software was behaving as intended, and output results from AutoContour were validated against accepted results for known planning parameters from clinically-utilized treatment planning systems. All tests passed regression testing. Verification and validation testing, and the risk documentation demonstrate that AutoContour is as safe and effective as the Predicate Device. The minor technological differences between AutoContour Model RADAC V6 and the Predicate Device do not raise any significant questions on the safety and effectiveness of the Subject Device. AutoContour Model RADAC V6 Special 510(k) 32