(158 days)
BoneMRI is an image processing software that can be used for image enhancement in MRI images. It can be used to visualize the bone structures in MRI images with increased contrast with respect to the surrounding soft tissue. It is to be used in the pelvic region, which includes the bony anatomy of the sacrum, hip bones and the spine, which includes the bony anatomy of the cervical, thoracic, lumbar, and S1 vertebrae. BoneMRI is indicated for use in patients 12 years and older.
BoneMRI is not to be used for diagnosis or monitoring of (primary or metastatic) tumors. BoneMRI images are not intended to replace CT images in general but can be used to visualize 3D bone morphology, tissue radiodensity and tissue radiodensity contrast.
The BoneMRI application is a standalone image processing software application that analyses 3D gradient echo MRI scans acquired with a dedicated MRI scan protocol. From the analysis of the gradient echo MRI scan, 3D tomographic radiodensity contrast images, called BoneMRI images, are constructed.
The BoneMRI images can be used to visualize the bone structures in MR images with enhanced contrast with respect to the surrounding soft tissue. The application is designed to be used by imaging experts, such as radiologists or orthopedic surgeons, typically in a physician's office.
The BoneMRI application is a server application running on the clinic or hospital networks. It is available as fully on-premise software with specific GPU hardware requirements, or partly running as a managed cloud service, for which the environment in which the managed modules run is controlled by MRIguidance. The on-premise software is fully controlled by the clinic or hospital, and as such, no protected health information (PHI) will leave the clinic or hospital network. All data sent to the managed cloud server will be de-identified before it leaves the clinic or hospital network, and as such, the managed cloud service will not receive PHI.
Within the hospital network, the application communicates with a DICOM compatible imaging archive (e.g., a PACS) to receive input MRI and to return BoneMRI images. Reading of the resulting BoneMRI images is performed using regular DICOM compatible medical image viewing software.
The BoneMRI application uses an algorithm to detect bone images from MRIs obtained using a specific gradient echo acquisition sequence. The algorithm training sets included images from multiple clinical sites, multiple anatomies, and multiple scanners to ensure that the trained algorithm was robust with respect to the approved indications for use. None of the data used in the training dataset was used subsequently in the validation dataset.
Here's a summary of the acceptance criteria and the study proving the device meets them, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance:
| Performance Metric | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Cortical Delineation Error (3D bone morphology) | 0.75 (specifically for bone) | > 0.75 (specifically for bone for all subgroups) |
2. Sample Size Used for the Test Set and Data Provenance:
- Pelvic Region: 76 patients
- Spine Region: 117 patients
- Data Provenance: Retrospective clinical data from various medical sites in the US and EU. The test data was acquired at different medical sites/departments/clinical studies than the training data and was unseen.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications:
The document does not explicitly state the number of experts or their qualifications for establishing the ground truth. It mentions that "the objective was to validate the quantitative accuracy of BoneMRI using rigorous, objective, and unbiased statistical tests comparing bone morphology, radiodensity, and radiodensity contrast in BoneMRI and CT images." This implies a comparison against existing CT scans as the reference.
4. Adjudication Method for the Test Set:
The document does not explicitly describe an adjudication method involving multiple experts. The validation was "voxel-by-voxel" comparing BoneMRI images to co-registered CT scans.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:
A multi-reader multi-case (MRMC) comparative effectiveness study was not explicitly mentioned or described in the provided text. The performance validation focused on the algorithm's standalone quantitative accuracy against CT images.
6. Standalone (Algorithm Only) Performance:
Yes, a standalone performance study was conducted. The "Performance Validation" section describes a "quantitative voxel-by-voxel validation of BoneMRI" where the algorithm's output (BoneMRI images) was compared directly against CT images, without human interpretation in the validation process itself.
7. Type of Ground Truth Used:
The ground truth used was co-registered Computed Tomography (CT) scans. The study aimed to compare the bone morphology, radiodensity, and radiodensity contrast generated by BoneMRI to those in CT images.
8. Sample Size for the Training Set:
The document states, "The algorithm training sets included images from multiple clinical sites, multiple anatomies, and multiple scanners." However, it does not specify the exact sample size of the training set.
9. How the Ground Truth for the Training Set Was Established:
The document states that "The algorithm training sets included images from multiple clinical sites, multiple anatomies, and multiple scanners to ensure that the trained algorithm was robust with respect to the approved indications for use." It also mentions "Hounsfield Unit (HU) value" assigned to volume elements. This implies that the training data likely consisted of MRI-CT pairs where the CT scans provided the ground truth for bone morphology and radiodensity, and these were used to train the convolutional neural network to assign HU values based on MRI intensity and contextual information. The specific process of establishing ground truth for individual training cases is not detailed beyond this.
§ 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).