(189 days)
Yes.
The device description explicitly states, "MR Contour DL uses deep learning segmentation algorithms that have been designed and trained specifically for the task of generating organ at risk contours from MR images." Deep learning is a subset of AI.
No.
This device is an aid for radiation therapy planning by generating initial contours; it does not directly treat or diagnose a disease.
No
Explanation: This device is intended to aid in radiation therapy planning by generating initial contours of organs at risk. Its purpose is to accelerate workflow for planning, not to diagnose a condition. The output requires verification and correction by a medical professional, indicating it's a supportive tool rather than a diagnostic one.
Yes
The device is a software-only medical device because it is described as a "post processing application" that generates DICOM data. It operates on existing MR images and does not have its own user interface, indicating it integrates with other software/hardware platforms. There is no mention of the device including any physical components or requiring hardware for its core function of algorithmically generating contours. Its output is digital (DICOM RTSS).
No.
The device is a medical image processing software that analyzes MR images to generate contours of organs for radiation therapy planning. It does not perform in vitro examination of specimens derived from the human body.
No
The letter does not explicitly state that the FDA has reviewed and approved or cleared a PCCP for this specific device. The provided text only indicates "Not Found" for "Control Plan Authorized (PCCP) and relevant text".
Intended Use / Indications for Use
MR Contour DL generates a Radiotherapy Structure Set (RTSS) DICOM with segmented organs at risk which can be used by trained medical professionals. It is intended to aid in radiation therapy planning by generating initial contours to accelerate workflow for radiation therapy planning. It is the responsibility of the user to verify the processed output contours and user-defined labels for each organ at risk and correct the contours/labels as needed. MR Contour DL is intended to be used with images acquired on MR scanners, in adult patients.
Product codes
QKB
Device Description
MR Contour DL is a post processing application intended to assist a clinician by generating contours of organ at risk (OAR) from MR images in the form of a DICOM Radiotherapy Structure Set (RTSS) series. MR Contour DL is designed to automatically contour the organs in the head/neck, and in the pelvis for Radiation Therapy (RT) planning of adult cases. The output of the MR Contour DL is intended to be used by radiotherapy (RT) practitioners after review and editing, if necessary, and confirming the accuracy of the contours for use in radiation therapy planning.
MR Contour DL uses customizable input parameters that define RTSS description, RTSS labeling, organ naming and coloring. MR Contour DL does not have a user interface of its own and can be integrated with other software and hardware platforms. MR Contour DL has the capability to transfer the input and output series to the customer desired DICOM destination(s) for review.
MR Contour DL uses deep learning segmentation algorithms that have been designed and trained specifically for the task of generating organ at risk contours from MR images. MR Contour DL is designed to contour 37 different organs or structures using the deep learning algorithms in the application processing workflow.
The input of the application is MR DICOM images in adult patients acquired from compatible MR scanners. In the user-configured profile, the user has the flexibility to choose both the covered anatomy of input scan and the specific organs for segmentation. The proposed device has been tested on GE HealthCare MR data.
Mentions image processing
Yes
Mentions AI, DNN, or ML
Yes
Input Imaging Modality
MR images
Anatomical Site
head/neck, pelvis
Indicated Patient Age Range
adult patients
Intended User / Care Setting
trained medical professionals, Radiologists, Radiation oncologists, Dosimetrists, and Medical Physicists
Description of the training set, sample size, data source, and annotation protocol
Not Found, refers to testing data being separated from development data cohorts before models were trained.
Description of the test set, sample size, data source, and annotation protocol
Bench testing data: 105 retrospectively collected exams with a total of 1350 MR Contour DL generated contours for head/neck and 510 MR Contour DL generated contours for pelvis that were compared to ground truth contours. The test data included (23 head/neck, 32 pelvis) cases from data cohorts which were collected independently from the development data and cases (27 head/neck, 23 pelvis) which were separated from the development data cohorts before the models were trained. Therefore in total test data included 50 head/neck and 55 pelvis cases.
Head/neck evaluation: 50 head/neck cases (of 50 subjects). The test data was collected from USA (72%) and European (NL, 28%) clinical sites, acquired with GEHC (76%) and other MR scanners, using standard 2-dimensional T2 (FRFSE, TSE) imaging protocol. The test data was collected from male (64%) and female subjects (36%), with average age 58.9 years (SD 14.2) and weight 78.2kg (SD 16.3). The test data included healthy subjects (42%) and patients (58%) with different types of Squamous Cell Carcinoma (Oropharynx, Hypopharynx, Larynx, Salivary gland).
Pelvis evaluation: 55 pelvis cases of (55 subjects). The test data was collected from USA (58%) and UK (42%) clinical sites, acquired with GEHC (100%) MR scanners, using standard 3-dimensional T2 CUBE imaging protocol. The test data was collected from male (81%) and female (19%) subjects, with average age 64.4 years (SD 13.1) and weight 82.3kg (SD 20.0). The test data included healthy subject (31%) and patients (69%) with different types of pelvis cancer (prostate, rectal, anal).
Ground-truth was created following: 1) manual contour was delineated by GEHC operators trained using international guidelines (DAHANCA, RTOG), 2) manual contours were revised (corrected and approved) by three (2 USA, 1 EUR) board certified radiation oncologists, 3) all (3) independently validated ground-truth contours were incorporated in the performance evaluation. The test set and the ground-truth were stored separately form the development data.
Clinical Testing: A reader study evaluation was conducted using a database of sample clinical MR images. The 30 head/neck and 40 pelvis cases were randomly selected from the original set of 50 head/neck and 55 pelvis cases by ensuring that all clinical sites, scan types, and subject cohorts are represented in the test set. The clinical testing was performed as follows: 1) The auto-contour was generated with MR Contour DL for the selected test cases, 2) the auto-contours were reviewed and scored by three (2 USA, 1 Europe) certified radiation oncologists and, 3) all (3) independently provided Likert Scores were incorporated in the performance evaluation.
Summary of Performance Studies (study type, sample size, AUC, MRMC, standalone performance, key results)
Bench Testing:
Study Type: Algorithm bench testing assessed the performance of MR Contour DL using the DICE score and Hausdorff Distance 95th percentile (HD95) as primary metrics.
Sample Size: 105 retrospectively collected exams with 1350 head/neck contours and 510 pelvis contours. Test data included 50 head/neck and 55 pelvis cases.
Key Results:
- DSC acceptance criteria based on organ size (small 50%, medium 65%, large 80%).
- Average HD95 performance of 4.7 mm, which is smaller than the average corresponding HD95 values of the predicate device.
- MR Contour DL had an improved or equivalent HD95 value in 24/28 of the organs analyzed.
- Overall model performance was similar across sub-cohorts (region, subject, gender).
- Mean DSC accuracy (all 37 organs): 81.1% (all), 81.3% (male), 83.4% (female), 82.7% (USA), 80.4% (Europe), 81.6% (patient), and 80.5% (healthy cases).
Clinical Testing (Reader Study):
Study Type: Reader study evaluation using a database of sample clinical MR images.
Sample Size: 30 head/neck and 40 pelvis cases, a subset of the data used for non-clinical testing.
Key Results:
- Predefined acceptance criterion: mean Likert Score for each organ shall be greater than or equal to 3.0.
- All organs met the acceptance criterion for Likert score.
- The results showed that the MR Contour DL generated organ contours are adequate for use in radiotherapy planning.
Key Metrics (Sensitivity, Specificity, PPV, NPV, etc.)
DICE Similarity Coefficient (DSC), 95th percentile Hausdorff Distance (HD95), Likert Score.
Predicate Device(s)
Reference Device(s)
Predetermined Change Control Plan (PCCP) - All Relevant Information
Not Found
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).
FDA 510(k) Clearance Letter - MR Contour DL
Page 1
U.S. Food & Drug Administration
10903 New Hampshire Avenue
Silver Spring, MD 20993
www.fda.gov
Doc ID # 04017.07.05
April 1, 2025
GE Healthcare
Allena Holzworth
Regulatory Affairs Leader
500 W. Monroe Street
Chicago, Illinois 60661
Re: K242925
Trade/Device Name: MR Contour DL
Regulation Number: 21 CFR 892.2050
Regulation Name: Medical Image Management And Processing System
Regulatory Class: Class II
Product Code: QKB
Dated: February 27, 2025
Received: February 27, 2025
Dear Allena Holzworth:
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.
Page 2
K242925 - Allena Holzworth 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 System (QS) regulation (21 CFR Part 820), which includes, but is not limited to, 21 CFR 820.30, Design controls; 21 CFR 820.90, Nonconforming product; and 21 CFR 820.100, Corrective and preventive action. Please note that regardless of whether a change requires premarket review, the QS regulation requires device manufacturers to review and approve changes to device design and production (21 CFR 820.30 and 21 CFR 820.70) and document changes and approvals in the device master record (21 CFR 820.181).
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 systems (QS) regulation (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
Page 3
K242925 - Allena Holzworth Page 3
the DICE website (https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-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
Page 4
DEPARTMENT OF HEALTH AND HUMAN SERVICES
Food and Drug Administration
Indications for Use
Form Approved: OMB No. 0910-0120
Expiration Date: 07/31/2026
See PRA Statement below.
| Submission Number (if known) | Device Name |
|---|---|
| K242925 | MR Contour DL |
Indications for Use (Describe)
MR Contour DL generates a Radiotherapy Structure Set (RTSS) DICOM with segmented organs at risk which can be used by trained medical professionals. It is intended to aid in radiation therapy planning by generating initial contours to accelerate workflow for radiation therapy planning. It is the responsibility of the user to verify the processed output contours and user-defined labels for each organ at risk and correct the contours/labels as needed. MR Contour DL is intended to be used with images acquired on MR scanners, in adult patients.
Type of Use (Select one or both, as applicable)
☒ Prescription Use (Part 21 CFR 801 Subpart D)
☐ Over-The-Counter Use (21 CFR 801 Subpart C)
CONTINUE ON A SEPARATE PAGE IF NEEDED.
This section applies only to requirements of the Paperwork Reduction Act of 1995.
DO NOT SEND YOUR COMPLETED FORM TO THE PRA STAFF EMAIL ADDRESS BELOW.
The burden time for this collection of information is estimated to average 79 hours per response, including the time to review instructions, search existing data sources, gather and maintain the data needed and complete and review the collection of information. Send comments regarding this burden estimate or any other aspect of this information collection, including suggestions for reducing this burden, to:
Department of Health and Human Services
Food and Drug Administration
Office of Chief Information Officer
Paperwork Reduction Act (PRA) Staff
PRAStaff@fda.hhs.gov
"An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB number."
Page 5
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 1 of 10
510(k) SUMMARY
This 510(k) summary of Safety and Effectiveness information is submitted in accordance with the requirement of 21 CFR Part 807.92:
Date: February 19, 2025
Submitter: GE HealthCare
500 W. Monroe Street
Chicago, IL 60661
Primary Contact: Allena Holzworth
Regulatory Affairs Leader
GE HealthCare
Phone: +1 (414) 208-5162
Email: allena.holzworth@gehealthcare.com
Secondary Contacts: Michelle Huettner
Director Regulatory Affairs
GE HealthCare
Phone: +1 (901) 558-8035
Email: michelle.huettner@gehealthcare.com
Subject Device Name: MR Contour DL
Device Classification: Class II
Regulation Number: 21 CFR 892.2050 Medical image management and processing system
Product Code: QKB
Predicate Device Information
Device Name: Auto Segmentation
Manufacturer: GE Medical Systems, LLC
510(k) Number: K230082
Regulation Number: 21 CFR 892.2050 Medical image management and processing system
Product Code: QKB
Reference Device 1 Information
Device Name: AutoContour Model RADAC V2
Manufacturer: Radformation, Inc.
510(k) Number: K220598
Regulation Number: 21 CFR 892.2050 Medical image management and processing system
Product Code: QKB
Page 6
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 2 of 10
Reference Device 2 Information
Device Name: Contour ProtégéAI
Manufacturer: MIM Software Inc.
510(k) Number: K213976
Regulation Number: 21 CFR 892.2050 Medical image management and processing system
Product Code: QKB
Device Description
MR Contour DL is a post processing application intended to assist a clinician by generating contours of organ at risk (OAR) from MR images in the form of a DICOM Radiotherapy Structure Set (RTSS) series. MR Contour DL is designed to automatically contour the organs in the head/neck, and in the pelvis for Radiation Therapy (RT) planning of adult cases. The output of the MR Contour DL is intended to be used by radiotherapy (RT) practitioners after review and editing, if necessary, and confirming the accuracy of the contours for use in radiation therapy planning.
MR Contour DL uses customizable input parameters that define RTSS description, RTSS labeling, organ naming and coloring. MR Contour DL does not have a user interface of its own and can be integrated with other software and hardware platforms. MR Contour DL has the capability to transfer the input and output series to the customer desired DICOM destination(s) for review.
MR Contour DL uses deep learning segmentation algorithms that have been designed and trained specifically for the task of generating organ at risk contours from MR images. MR Contour DL is designed to contour 37 different organs or structures using the deep learning algorithms in the application processing workflow.
The input of the application is MR DICOM images in adult patients acquired from compatible MR scanners. In the user-configured profile, the user has the flexibility to choose both the covered anatomy of input scan and the specific organs for segmentation. The proposed device has been tested on GE HealthCare MR data.
Intended Use
MR Contour DL is intended to be used as a workflow tool for initial anatomy segmentation of organs at risk on MR images as an aid in radiation therapy planning after user confirmation.
Indications for Use
MR Contour DL generates a Radiotherapy Structure Set (RTSS) DICOM with segmented organs at risk which can be used by trained medical professionals. It is intended to aid in radiation therapy planning by generating initial contours to accelerate workflow for radiation therapy planning. It is the responsibility of the user to verify the processed output contours and user-defined labels for each organ at risk and correct the contours/labels as needed. MR Contour DL is intended to be used with images acquired on MR scanners, in adult patients.
Technology
The proposed device, MR Contour DL, employs similar fundamental scientific technology as its predicate device. Both the proposed device and the predicate device use deep learning algorithms to segment organs at risk.
Page 7
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 3 of 10
Comparisons
MR Contour DL software is substantially equivalent to the predicate device, Auto Segmentation (K230082). The proposed device is based on the same fundamental technology as the predicate device using deep learning algorithms for organ at risk segmentation. The comparisons of the intended use, indications for use, and devices technical characteristics demonstrate that the proposed device, MR Contour DL, is as safe and effective and substantially equivalent to that of the predicate device. MR Contour DL did not introduce any new questions of safety and effectiveness.
The tables below summarize the substantive feature/technological similarities and differences between the proposed device, predicate device and reference devices.
Table 1: Comparison of Intended Use between Proposed Device and Predicate Device
| Items | Proposed MR Contour DL | Predicate: Auto Segmentation (K230082) | Comparison |
|---|---|---|---|
| Intended Use | MR Contour DL is intended to be used as a workflow tool for initial anatomy segmentation of organs at risk on MR images as an aid in radiation therapy planning after user confirmation. | Auto Segmentation is intended to be used as a workflow tool for initial anatomy segmentation of organs at risk on CT images as an aid in radiation therapy planning after user confirmation. | Both the proposed device and the predicate device are intended to be used as a workflow tool for initial anatomy segmentation of organs at risk as an aid in radiation therapy planning after user confirmation. The proposed device is intended for post-processing MR images, whereas the predicate is intended for post-processing of CT images. All features and performance of the proposed device have been verified and validated per GE HealthCare's quality system. No safety and effectiveness issues were raised. In addition, the performance testing conducted on MR Contour DL successfully demonstrate the devices performance, particularly the anatomy segmentation of organs at risk on MR images. The intended use of the proposed device falls within the general intended use of the predicate device and do not create a new Intended Use. |
Table 2: Comparison of Indications for Use between Proposed Device and Predicate Device
| Items | Proposed MR Contour DL | Predicate: Auto Segmentation (K230082) | Comparison |
|---|---|---|---|
| Indications for Use | MR Contour DL generates a Radiotherapy Structure Set (RTSS) DICOM with segmented organs at risk which can be used by trained medical professionals. It is intended to aid in radiation therapy | Auto Segmentation generates a Radiotherapy Structure Set (RTSS) DICOM with segmented organs at risk which can be used by dosimetrists, medical physicists, and radiation oncologists as | Substantially Equivalent – Both the proposed device and the predicate device are indicated to generate a Radiotherapy Structure Set (RTSS) DICOM with segmented organs at risk. Both the proposed device and the predicate device are: • Used by trained medical |
Page 8
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 4 of 10
| Items | Proposed MR Contour DL | Predicate: Auto Segmentation (K230082) | Comparison |
|---|---|---|---|
| planning by generating initial contours to accelerate workflow for radiation therapy planning. It is the responsibility of the user to verify the processed output contours and user-defined labels for each organ at risk and correct the contours/labels as needed. MR Contour DL is intended to be used with images acquired on MR scanners, in adult patients. | initial contours to accelerate workflow for radiation therapy planning. It is the responsibility of the user to verify the processed output contours and user-defined labels for each organ at risk and correct the contours/labels as needed. Auto Segmentation may be used with images acquired on CT scanners, in adult patients. | professionals, including the professions listed by the predicate device. • Intended to aid in radiation therapy planning by generating initial contours to accelerate workflow for radiation therapy planning. • Declare that it is the responsibility of the user to verify the processed output contours and user-defined labels for each organ at risk and to correct the contours/labels as needed. MR Contour DL is intended to be used with images acquired on MR scanners in adult patients while the predicate device is intended to be used on images acquired on CT scanners in adult patients. All features and performance of the proposed device have been verified and validated per GEHC's quality system. No safety and effectiveness issues were raised. Additionally, the testing performed on MR Contour DL demonstrates the devices performance, particularly the anatomy segmentation of organs at risk on MR images. The indications for use for the proposed device is substantially equivalent to that of the predicate. |
Table 3: Comparison of technical characteristics between proposed device, predicate device, and reference devices
| Items | Proposed MR Contour DL | Predicate: Auto Segmentation (K230082) | Reference 1: AutoContour Model RADAC V2 (K220598) | Reference 2: Contour ProtégéAI (K213976) | Comparison |
|---|---|---|---|---|---|
| Product Code | QKB | QKB | QKB | QKB | Identical |
| Patient Population | Adult only | Adult only | Adult only | Adult only | Identical |
Page 9
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 5 of 10
| Items | Proposed MR Contour DL | Predicate: Auto Segmentation (K230082) | Reference 1: AutoContour Model RADAC V2 (K220598) | Reference 2: Contour ProtégéAI (K213976) | Comparison |
|---|---|---|---|---|---|
| Intended Users | Radiologists, Radiation oncologists, Dosimetrists, and Medical Physicists | Radiation Oncologists, Dosimetrists, and Medical Physicists | Medical Professionals who do radiation therapy treatment planning | Trained Medical Professionals | Substantially Equivalent – both the proposed device and the predicate device are intended for trained medical professionals in the radiation oncology field |
| Algorithm | Deep Learning | Deep Learning | Machine Learning/Deep Learning | Machine Learning | Substantially Equivalent – both the proposed device and the predicate device contain deep learning auto segmentation algorithms |
| Compatible Modality | MR images | CT images | CT images and MR images | CT images and MR images | Substantially Equivalent – While the predicate device is only intended to segment CT Images, both reference devices are intended to segment both CT and MR images. Reference 1 uses DICOM-compliant image data (CT or MR) to automatically contour various structures of interest for RT planning. Reference device 2 is intended to assist in the automated processing of digital medical images of modalities CT and MR to create contours using machine learning algorithms. The proposed device aligns with the performance of commercially available products on the market, specifically Reference Devices 1 and 2. |
| OAR Segmentation Anatomic Regions | Head/neck Pelvis | Head/neck Thorax Abdomen Pelvis | Head/neck Thorax Abdomen pelvis | Head/neck Prostate Thorax Abdomen Lungs and Liver | Substantially Equivalent – Both the predicate and proposed device cover the Head/neck and Pelvis regions. |
| Workflow | Automated | Automated | Manual and Automated | Automated | Identical |
| User Interface | Automated execution of the software with no user interaction, other than configuration settings. Generated | Automated execution of the software with no user interaction, other than configuration settings. Generated | Contains both an automated processing component, Data Visualization, and Graphical User Interface | Designed for use in the processing of medical images and operates on Windows, Mac, and Linux | Identical |
Page 10
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 6 of 10
| Items | Proposed MR Contour DL | Predicate: Auto Segmentation (K230082) | Reference 1: AutoContour Model RADAC V2 (K220598) | Reference 2: Contour ProtégéAI (K213976) | Comparison |
|---|---|---|---|---|---|
| contours are automatically transmitted to review workstation(s) supporting RTSS objects for review and editing, as needed. | contours are automatically transmitted to review workstation(s) supporting RTSS objects for review and editing, as needed. | computer systems. Deployed on a remote server using the MIMcloud service for data management and transfer; or locally on the workstation or server running MIM software. | |||
| Compatible Scanner Models | Compatible on MR Scanners, DICOM compliance required. | No limitation on scanner model, DICOM compliance required. | No limitation on scanner model, DICOM compliance required | No limitation on scanner model, DICOM compliance required | Substantially Equivalent – The proposed device is compatible with MR scanner models and DICOM compliance is required. The proposed device has only been tested on GEHC data. If a non-GEHC MR image is sent to MR Contour DL, a notice message will display after contouring. The predicate device has no limitation on scanner models and DICOM compliance is also required. This difference does not raise any new safety or effectiveness concerns. |
| Deployment Platform | Server-based deployment | Server-based deployment | Cloud and server-based deployment | Cloud-based deployment and locally deployed (or installed) | Identical – Both the proposed device and the predicate device are deployed on Edison HealthLink (EHL), GE Healthcare's computational platform providing hosting infrastructure and services. |
Page 11
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 7 of 10
Determination of Substantial Equivalence
Summary of Non-Clinical Testing
MR Contour DL has successfully completed the design control testing per GE HealthCare's quality system. It was designed under the Quality System Regulations of 21CFR 820 and ISO 13485. No new questions of safety and effectiveness and no unexpected test results were observed.
The following quality assurance measures have been applied to the development of the system:
- Requirement Definition
- Risk Analysis and Control
- Technical Design Reviews
- Formal Design Reviews
- Software Development Lifecycle
- Safety Testing (Verification)
- Performance Testing (Verification, Validation)
- Software Release
MR Contour DL has been successfully verified. The testing and results did not raise any new concerns of safety and effectiveness. Software documentation provided is for Enhanced Documentation.
The bench testing assessed the performance of MR Contour DL using the DICE score as a primary metric by comparing the segmentation accuracy of the proposed device to that of the predicate and reference devices for supported organs at risk. The MR Contour DL algorithm's capability was validated using a database of 105 retrospectively collected exams with a total of 1350 MR Contour DL generated contours for head/neck and 510 MR Contour DL generated contours for pelvis that were compared to ground truth contours generated by qualified radiotherapy practitioners. The result of the algorithm bench testing showed that MR Contour DL performs as expected.
The test data included (23 head/neck, 32 pelvis) cases from data cohorts which were collected independently from the development data and cases (27 head/neck, 23 pelvis) which were separated from the development data cohorts before the models were trained. Therefore in total test data included 50 head/neck and 55 pelvis cases.
Table 4 demonstrates the performance metrics (DSC, HD95) measured with ground-truth (see column ground-truth). The acceptance criterion was defined with DSC for each individual organ, based on the size of the organ (small 50%, medium 65%, large 80%). The HD95 metric was compared to the predicate device by evaluating the Lower Bound (LBCI95) and Upper Bound (UBCI95) of the 95% Confidence Interval against the mean HD95 of the predicate device for each organ. MR Contour DL's performance was considered improved from the predicate if UBCI95 was less than the HD95 mean of the predicate device, and performance was considered equivalent to the predicate if the HD95 mean of the predicate device was between LBCI95 and UBCI95. In summary, MR Contour DL had an improved or equivalent HD95 value in 24/28 of the organs analyzed and an average HD95 performance of 4.7 mm, which is smaller than the average corresponding HD95 values of the predicate device (for the Hausdorff Distance metric a smaller HD95 value signifies better segmentation). Therefore, the performance of the proposed device is substantially equivalent to that of the cleared predicate device.
Page 12
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 8 of 10
The evaluation for ground-truth involved 50 head/neck cases (of 50 subjects). The test data was collected from USA (72%) and European (NL, 28%) clinical sites, acquired with GEHC (76%) and other MR scanners, using standard 2-dimensional T2 (FRFSE, TSE) imaging protocol. The test data was collected from male (64%) and female subjects (36%), with average age 58.9 years (SD 14.2) and weight 78.2kg (SD 16.3). The test data included healthy subjects (42%) and patients (58%) with different types of Squamous Cell Carcinoma (Oropharynx, Hypopharynx, Larynx, Salivary gland).
The evaluation with ground-truth involved 55 pelvis cases of (55 subjects). The test data was collected from USA (58%) and UK (42%) clinical sites, acquired with GEHC (100%) MR scanners, using standard 3-dimensional T2 CUBE imaging protocol. The test data was collected from male (81%) and female (19%) subjects, with average age 64.4 years (SD 13.1) and weight 82.3kg (SD 20.0). The test data included healthy subject (31%) and patients (69%) with different types of pelvis cancer (prostate, rectal, anal).
Ground-truth was created following: 1) manual contour was delineated by GEHC operators trained using international guidelines (DAHANCA, RTOG), 2) manual contours were revised (corrected and approved) by three (2 USA, 1 EUR) board certified radiation oncologists, 3) all (3) independently validated ground-truth contours were incorporated in the performance evaluation. The test set and the ground-truth were stored separately form the development data.
The performance was evaluated on the whole testset, as well as on the following 6 sub-cohorts: 1) region (USA, Europe), 2) subject (patient, healthy), and 3) gender (male, female). The mean DSC accuracy (incorporating all 37 organs) was 81.1% for all, 81.3% for male, 83.4% for female, 82.7% for USA, 80.4% for Europe, 81.6% for patient, and 80.5% for healthy cases, which demonstrate the overall model performance was similar in all sub-cohorts.
Summary of Clinical Testing
A reader study evaluation was conducted using a database of sample clinical MR images to demonstrate that the contours generated by the MR Contour DL application are adequate for radiotherapy planning use. Each contour used in the evaluation was generated using the MR Contour DL application, and reviewed by three qualified radiotherapy practitioners, who provided an assessment of the adequacy of the subject device generated contours. The readers completed their assessments independently and were blinded to the results of the other readers' assessments. The results of the algorithm clinical testing shows that the MR Contour DL generated organ contours are adequate for use in radiotherapy planning.
Table 4 demonstrates the performance metric (Likert score) measured in the clinical testing (see column Reader-study). The predefined acceptance criterion was the mean Likert Score for each organ shall be greater than or equal to 3.0, with the following interpretation of the Likert Score: 1 – unacceptable (recontouring is needed), 2 – poor (significant correction is needed), 3 – good (some correction is needed), 4 – very good (minor correction is needed), 5 – excellent (no correction is needed).
The clinical testing was performed on a subset of the data used for non-clinical (bench) testing. The 30 head/neck and 40 pelvis cases were randomly selected from the original set of 50 head/neck and 55 pelvis cases by ensuring that all clinical sites, scan types, and subject cohorts are represented in the test set.
The clinical testing was performed as follows: 1) The auto-contour was generated with MR Contour DL for the selected test cases, 2) the auto-contours were reviewed and scored by three (2 USA, 1 Europe) certified radiation oncologists and, 3) all (3) independently provided Likert Scores were incorporated in the performance evaluation.
Page 13
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 9 of 10
Table 4: Performance metrics for all organs
(L – left, R – right, G – glottic, SG – supraglottic, PCM-Pharyngeal Constrictor Muscle, DSC – DICE Similarity Coefficient, HD95 – 95th percentile Hausdorff Distance).
| Organ | Anatomy region | Ground-truth | Reader-study | |||
|---|---|---|---|---|---|---|
| Number of cases | DSC Acceptance criteria | DSC MEAN | HD95 MEAN | HD95 value compared to predicate device | ||
| bladder | pelvis | 53 | 80% | 92.4% | 4.7 | Improved |
| bowel-bag | pelvis | 40 | 80% | 90.3% | 13.3 | N/A |
| brainstem | head/neck | 50 | 65% | 94.3% | 2.1 | Improved |
| chiasm | head/neck | 50 | 50% | 72.7% | 2.5 | Improved |
| eye-L | head/neck | 50 | 65% | 95.5% | 1.4 | Improved |
| eye-R | head/neck | 50 | 65% | 95.4% | 1.4 | Improved |
| femoral-head-L | pelvis | 53 | 80% | 93.7% | 4.5 | Not-Improved |
| femoral-head-R | pelvis | 51 | 80% | 93.5% | 5.3 | Not-Improved |
| head-body | head/neck | 50 | 80% | 99.3% | 1.6 | Improved |
| inner-ear-L | head/neck | 50 | 50% | 88.4% | 1.4 | N/A |
| inner-ear-R | head/neck | 50 | 50% | 88.3% | 1.3 | N/A |
| lacrimal-L | head/neck | 50 | 50% | 67.4% | 4.2 | Equivalent |
| lacrimal-R | head/neck | 50 | 50% | 65.6% | 4.4 | Equivalent |
| larynx-G | head/neck | 49 | 50% | 67.1% | 3.9 | N/A |
| larynx-SG | head/neck | 50 | 65% | 85.3% | 4.9 | N/A |
| lens-L | head/neck | 50 | 50% | 86.7% | 1.2 | Improved |
| lens-R | head/neck | 50 | 50% | 86.1% | 1.3 | Improved |
| mandible | head/neck | 50 | 65% | 89.8% | 2.8 | Equivalent |
| optic-nerve-L | head/neck | 50 | 50% | 73.4% | 2.9 | Equivalent |
| optic-nerve-R | head/neck | 50 | 50% | 72.3% | 2.6 | Improved |
| oral-cavity | head/neck | 50 | 65% | 92.5% | 3.7 | Improved |
| parotid-L | head/neck | 50 | 65% | 85.9% | 4.9 | Improved |
| parotid-R | head/neck | 50 | 65% | 84.6% | 6.0 | Improved |
| PCM-inf | head/neck | 44 | 50% | 53.6% | 7.0 | Not-Improved |
| PCM-mid | head/neck | 50 | 50% | 60.1% | 6.1 | Equivalent |
| PCM-sup | head/neck | 50 | 50% | 57.6% | 6.8 | Improved |
| pelvis-body | pelvis | 50 | 80% | 97.5% | 10.0 | Not-Improved |
| penile-bulb | pelvis | 39 | 50% | 67.5% | 7.3 | N/A |
| pituitary | head/neck | 50 | 50% | 73.7% | 2.5 | Equivalent |
| prostate | pelvis | 43 | 65% | 83.0% | 5.6 | Equivalent |
| rectum | pelvis | 53 | 65% | 79.6% | 19.8 | N/A |
| seminal-vesicles | pelvis | 43 | 65% | 69.2% | 7.3 | N/A |
| spinal-cord | head/neck | 50 | 65% | 90.3% | 2.2 | Improved |
| submandibular-L | head/neck | 49 | 65% | 86.4% | 3.4 | Equivalent |
| submandibular-R | head/neck | 49 | 65% | 85.4% | 3.3 | Equivalent |
| urethra | pelvis | 43 | 50% | 35.8% | 9.5 | N/A |
| whole-brain | head/neck | 50 | 80% | 98.8% | 1.8 | Improved |
Page 14
GE HealthCare 510(k) Premarket Notification Submission – MR Contour DL
Page 10 of 10
Substantial Equivalence Conclusion
MR Contour DL and the predicate device have substantially equivalent indications for use, and represent equivalent technological characteristics, including the use of deep learning algorithms.
MR Contour DL was developed under GE HealthCare's quality system. Design verification and validation, along with bench testing and the clinical reader study provided in this submission demonstrate that the MR Contour DL software is substantially equivalent and, hence, as safe and effective as the legally marketed predicate device. GE HealthCare's quality system design, verification, and risk management processes did not identify any unexpected results or new questions of safety and effectiveness.
GE HealthCare believes that MR Contour DL is substantially equivalent to the predicate device and, hence, is safe and effective for its intended use.