(488 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 enhanced contrast with respect to the surrounding soft tissue. It is to be used in the pelvic region, which includes the boney anatomy of the sacrum, hip bones and femoral heads. Warning: BoneMRI images are not intended to replace CT images and are not to be used for diagnosis or monitoring of (primary or metastatic) tumors.
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, 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 orthopaedic surgeons, typically in a physician's office. The BoneMRI application is a server application running in the clinic or hospital networks. It returns the reconstructed BoneMRI images as DICOM images.
Here's an analysis of the acceptance criteria and the study that proves the device meets them, based on the provided text:
Acceptance Criteria and Device Performance
| Acceptance Criteria (Quantitative Endpoints) | Reported Device Performance |
|---|---|
| 3D bone morphology with a mean absolute cortical delineation error below 1.0 mm | The data provided demonstrate that BoneMRI application v1.2 can accurately reconstruct the 3D bone morphology with a mean absolute cortical delineation error below 1.0 mm on average. |
| Tissue radiodensity with a mean deviation below 10 HU | The data provided demonstrate that BoneMRI application v1.2 can accurately reconstruct the tissue radiodensity with a mean deviation below 10 HU on average. |
| Bone radiodensity with a mean deviation below 55 HU | The data provided demonstrate that BoneMRI application v1.2 can accurately reconstruct the tissue radiodensity with a mean deviation below 55 HU for bone specifically. |
| Tissue radiodensity contrast with a mean HU correlation coefficient above 0.80 | The data provided demonstrate that BoneMRI application v1.2 can accurately reconstruct the tissue radiodensity contrast with a mean HU correlation coefficient above 0.80 on average. |
| Bone radiodensity contrast with a mean HU correlation coefficient above 0.75 | The data provided demonstrate that BoneMRI application v1.2 can accurately reconstruct the tissue radiodensity contrast with a mean HU correlation coefficient above 0.75 for bone specifically. |
Study Details
-
Sample Size used for the test set and data provenance:
- Sample Size: 61 patients.
- Data Provenance: The text states, "imaging data from 61 patients, consisting of the BoneMRI and CT of the same patient, acquired during the previously conducted clinical investigations." This implies the data is retrospective as it was "previously conducted." The country of origin is not specified.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- The provided text does not specify the number of experts used or their qualifications. The ground truth was established by comparing BoneMRI outputs directly to co-registered CT scans.
-
Adjudication method for the test set:
- The text describes a "voxel-by-voxel analysis" using an "in-house developed algorithm validation pipeline, the core validation framework." This suggests an automated, quantitative comparison against a reference standard (CT), rather than an expert adjudication method like 2+1 or 3+1. Therefore, the adjudication method was none in the traditional sense of human consensus.
-
If a multi-reader multi-case (MRMC) comparative effectiveness study was done, if so, what was the effect size of how much human readers improve with AI vs without AI assistance:
- No, an MRMC comparative effectiveness study was not done. The performance data section focuses on quantitative validation against CT scans, not on reader performance with or without AI assistance.
-
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- Yes, a standalone performance study was done. The "Voxel-by-Voxel analysis" directly compares the output of the BoneMRI algorithm to CT scans, without involving human readers in the quantitative performance metrics.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- The ground truth used was co-registered CT scans. The study directly validated BoneMRI outputs (3D bone morphology, radiodensity, and radiodensity contrast) against these CT scans.
-
The sample size for the training set:
- The document does not explicitly state the sample size for the training set. It mentions that "The parameters of the model were obtained through an algorithm development pipeline," but does not give specific numbers for training data.
-
How the ground truth for the training set was established:
- The document does not explicitly describe how the ground truth for the training set was established. It only mentions that the "parameters of the model were obtained through an algorithm development pipeline," which implies data was used for training, but the process for establishing ground truth for that data is not detailed.
§ 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).