Search Results
Found 1 results
510(k) Data Aggregation
(97 days)
VBrain-OAR
VBrain-OAR is a software device intended to assist trained radiotherapy personnel including, but not limited to, radiologists, radiation oncologists, neurosurgeons, radiation therapists, and medical physicists, during their clinical workflows of brain tumor radiation therapy treatment planning, by providing initial object contours of organs at risk in the brain (i.e., the region of interest, ROI) on axial T1 contrast-enhanced brain MRI images. VBrain-OAR is intended to be used on adult patients only.
VBrain-OAR uses an artificial intelligence algorithm (i.e., deep learning neural networks) to contour (segment) organs at risk (brain stem, eyes, optic nerves, optic chiasm) in the brain on MRI images for trained radiotherapy personnel's attention, which is meant for informational purposes only and not intended for replacing their current standard practice of manual contouring process. VBrain-OAR does not alter the original MRI image, nor does it intend to detect tumors for diagnosis. VBrain-OAR is intended only for contouring and generating contours of organs at risk in the brain; it is not intended to be used with images of other body parts.
VBrain-OAR also contains the automatic image register volumetric medical image data. (e.g., MR, CT). It allows rigid image registration to adjust the spatial position and orientation of two images. Radiation therapy treatment personnel must finalize (confirm or modify) the contours generated by VBrain-OAR, as necessary, using an external platform available at the facility that supports DICOM-RT viewinglediting functions, such as image visualization software and treatment planning system.
VBrain-OAR is a software application system indicated for use in the contouring (segmentation) of brain MRI images for the organs at risk (OAR) in the brain during radiation treatment planning and in the registration of multi-modality images. The device consists of 2 algorithm modules, which are contouring algorithm module and registration algorithm module, and a workflow management module. The modules can work independently, and yet can be integrated with each other.
The contouring (segmentation) algorithm module consists of image preprocessing, deep learning neural networks, and postprocessing components, and is intended to contour organs at risk in the brain on the axial T1 contrast-enhanced MR images. It utilizes deep learning neural networks to generate contours for the organs at risk in the brain and export the results as DICOM-RT objects (using the RT Structure Set ROI Contour attribute, RTSTRUCT).
The registration algorithm module registers volumetric medical image data (e.g., MR, CT). It allows rigid image registration to adjust the spatial position and orientation of two images.
The workflow management module is configured to work on a PACS network. Upon user's request, it will pull patient scans or users can send corresponding DICOM images, and it will trigger a predefined workflow, in which different algorithm modules are executed to generate the DICOM output. The DICOM output of a workflow are sent back to the PACS.
This document, a 510(k) premarket notification for Vysioneer Inc.'s VBrain-OAR, focuses on the device's technical characteristics and claims of substantial equivalence to predicate devices, but does not provide the specific acceptance criteria and detailed study results (including performance metrics like Dice Similarity Coefficient, Hausdorff Distance, or expert review scores) that would typically be required to fully describe how the device met these criteria.
The document states that "performance testing was conducted to evaluate the contouring (segmentation) performance and registration performance of VBrain-OAR" and that "the auto-segmentation algorithm of the VBrain-OAR algorithm module provides clinically acceptable contours for organs at risk in the brain structures on an image of a patient." However, it does not explicitly define what constitutes "clinically acceptable" or provide the quantitative results from these tests.
Therefore, many of the requested details cannot be extracted directly from the provided text. I will provide information based on what is available and indicate where details are missing.
Acceptance Criteria and Device Performance Study (Based on Provided Text)
The document generally states that the device's performance was evaluated, and it met "clinically acceptable" standards. However, the specific quantitative acceptance criteria (e.g., minimum Dice Similarity Coefficient, maximum Hausdorff Distance) are not detailed in the provided text. Similarly, the reported device performance (quantitative results) against these criteria is also not included in this summary.
In the absence of specific acceptance criteria and performance results, the table below is illustrative of what would typically be included in such a section, but the "Acceptance Criteria" and "Reported Device Performance" columns cannot be filled with concrete numbers from the provided document.
1. Table of Acceptance Criteria and Reported Device Performance
| Feature/Metric | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Contouring (Segmentation) Performance | Not explicitly stated in document (e.g., Min. Dice Similarity Coefficient, Max. Hausdorff Distance, Expert Review Score) | Not explicitly stated in document (e.g., Achieved Dice scores, Hausdorff distances, Qualitative assessment) |
| Brain Stem Contouring | Clinically acceptable* | Clinically acceptable* |
| Eyes Contouring | Clinically acceptable* | Clinically acceptable* |
| Optic Nerves Contouring | Clinically acceptable* | Clinically acceptable* |
| Optic Chiasm Contouring | Clinically acceptable* | Clinically acceptable* |
| Registration Performance | Not explicitly stated in document (e.g., Registration accuracy in mm) | Not explicitly stated in document (e.g., Achieved registration accuracy) |
| Rigid image registration | Substantially equivalent to predicate device | Substantially equivalent to predicate device |
Note: The term "clinically acceptable" is used in the document but is not defined with quantitative metrics.
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Test Set: The document states "VBrain-OAR was tested on datasets from multiple institutions" for standalone performance testing. However, the exact number of cases or scans in the test set is not specified.
- Data Provenance: The document mentions "datasets from multiple institutions" and "data across patient sex, multiple imaging hardware and protocols" was used for testing. However, the country of origin of the data is not specified. The document also does not explicitly state whether the data was retrospective or prospective, though typical 510(k) submissions for AI devices often rely on retrospective datasets for performance testing.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications
The provided document does not specify the number of experts used to establish the ground truth for the test set, nor does it detail their qualifications (e.g., radiologist with X years of experience). It's implied that "trained radiotherapy personnel" are involved in the standard practice of manual contouring which the device aims to assist, but this does not directly describe the ground truth establishment process for the test data.
4. Adjudication Method for the Test Set
The document does not describe any specific adjudication method (e.g., 2+1, 3+1, none) used for establishing the ground truth for the test set.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done
The provided summary does not indicate that an MRMC comparative effectiveness study was performed to assess how human readers improve with AI vs. without AI assistance. The device is described as an "assistance" tool, but an MRMC study demonstrating improvement in human performance is not mentioned. The focus is on the standalone performance of the AI algorithm.
6. If a Standalone (Algorithm Only) Performance Study Was Done
Yes, a standalone performance study was done. The document explicitly states under "5.7 Non-Clinical Test (Standalone Performance Data)":
"Standalone performance testing was conducted to evaluate the contouring (segmentation) performance and registration performance of VBrain-OAR."
7. The Type of Ground Truth Used
The type of ground truth used is implied to be expert consensus or expert-derived manual contours. The device aims to "provide initial object contours... on axial T1 contrast-enhanced brain MRI images" and states it's "not intended for replacing their current standard practice of manual contouring process." This suggests that human expert manual contours would serve as the ground truth against which the AI's generated contours are compared. However, the exact methodology for establishing this ground truth (e.g., single expert, multi-expert consensus) for the test set is not detailed.
8. The Sample Size for the Training Set
The document does not specify the sample size used for the training set. It mentions the use of "deep learning neural networks" implying a training phase, but provides no details on the data used.
9. How the Ground Truth for the Training Set Was Established
The document does not specify how the ground truth for the training set was established. While it is implied that expert manual contours would be used given the device's function, the details of this process for the training data are not provided.
Ask a specific question about this device
Page 1 of 1