(172 days)
Vantage Galan 3T systems are indicated for use as a diagnostic imaging modality that produces cross-sectional transaxial, coronal, sagittal, and oblique images that display anatomic structures of the head or body. Additionally, this system is capable of non-contrast enhanced imaging, such as MRA.
MRI (magnetic resonance imaging) images correspond to the spatial distribution of protons (hydrogen nuclei) that exhibit nuclear magnetic resonance (NMR). The NMR properties of body tissues and fluids are:
·Proton density (PD) (also called hydrogen density)
·Spin-lattice relaxation time (T1)
·Spin-spin relaxation time (T2)
·Flow dynamics
·Chemical Shift
Depending on the region of interest, contrast agents may be used. When interpreted by a trained physician, these images yield information that can be useful in diagnosis.
The Vantage Galan (Model MRT-3020) is a 3 Tesla Magnetic Resonance Imaging (MRI) System, previously cleared under K181593. This system is based upon the technology and materials of previously marketed Canon Medical Systems and is intended to acquire and display cross-sectional transaxial, coronal, sagittal, and oblique images of anatomic structures of the head or body.
AiCE is an optional noise reduction algorithm that improves image quality and reduces thermal noise by employing Deep Convolutional Neural Network methods.
AiCE is designed to remove Gaussian distributed noise in MR images for reducing contributions of thermal noise. In order to train a DCNN that can learn a model that represents thermal noise, the training datasets are created by adding Gaussian noise of different amplitudes to high-SNR images acquired with large number of averages.
The device is targeted for Brain and knee regions.
This software and its associated hardware are used on Canon MRI systems that are designed to communicate with the AiCE Reconstruction Processing Unit for MR.
Here's a breakdown of the acceptance criteria and study information for the Canon Medical Systems Corporation's Vantage Galan 3T, MRT-3020, V6.0 with AiCE Reconstruction Processing Unit for MR, based on the provided document:
Acceptance Criteria and Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Maintained or improved Low Contrast Detectability (LCD) | Performance (bench testing) using a Model Observer study determined that image Low Contrast Detectability was maintained or improved. |
| Statistical preference for AiCE images compared to other preferred filters | A Human Observer study demonstrated a statistical preference of AiCE when compared to other preferred filters. |
| Overall safety and effectiveness for intended use | Concluded that the subject device is safe and effective for its intended use based on bench testing, phantom imaging, volunteer clinical imaging, successful completion of software validation, and application of risk management and design controls. |
Study Details
-
Sample size used for the test set and the data provenance:
- Human Observer Study: 160 images.
- Data Provenance: Not explicitly stated (e.g., country of origin, retrospective/prospective).
- Model Observer Study: Not specified, but generally involves simulated data or specific phantom images.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Human Observer Study: 6 physicians.
- Qualifications: Not specified (e.g., years of experience, subspecialty).
-
Adjudication method (e.g., 2+1, 3+1, none) for the test set:
- Not explicitly stated. The document mentions a "statistical preference," which implies a comparative assessment by the readers, but the method of resolving discrepancies or forming a consensus for ground truth is not detailed.
-
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:
- Yes, a Human Observer study was conducted with 6 physicians and 160 images.
- Effect Size: The document states that the study "demonstrated a statistical preference of AiCE when compared to other preferred filters." However, it does not quantify the effect size in terms of human reader improvement with AI vs. without AI assistance (e.g., AUC improvement, sensitivity/specificity increase). The study focused on preference rather than diagnostic accuracy improvement.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Yes, a "Model Observer study" was conducted to determine if image Low Contrast Detectability was maintained or improved. This is a form of standalone evaluation focusing on image quality metrics.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- Human Observer Study: The "statistical preference" suggests that the physicians were evaluating image quality, and their collective preference served as a form of expert consensus on the perceived image quality rather than a ground truth for a specific pathology.
- Model Observer Study: Ground truth would typically be established based on known physical properties of the phantoms used or simulated data, where the true low-contrast objects are precisely defined.
-
The sample size for the training set:
- The document states: "In order to train a DCNN that can learn a model that represents thermal noise, the training datasets are created by adding Gaussian noise of different amplitudes to high-SNR images acquired with large number of averages."
- The specific number of images or cases in the training set is not provided.
-
How the ground truth for the training set was established:
- The training datasets were created by "adding Gaussian noise of different amplitudes to high-SNR images acquired with large number of averages."
- This implies that the "ground truth" for training the deep convolutional neural network (DCNN) was essentially the original high Signal-to-Noise Ratio (SNR) images before the artificial noise was added. The DCNN was trained to learn how to transform noisy images back to these high-SNR "ground truth" representations.
§ 892.1000 Magnetic resonance diagnostic device.
(a)
Identification. A magnetic resonance diagnostic device is intended for general diagnostic use to present images which reflect the spatial distribution and/or magnetic resonance spectra which reflect frequency and distribution of nuclei exhibiting nuclear magnetic resonance. Other physical parameters derived from the images and/or spectra may also be produced. The device includes hydrogen-1 (proton) imaging, sodium-23 imaging, hydrogen-1 spectroscopy, phosphorus-31 spectroscopy, and chemical shift imaging (preserving simultaneous frequency and spatial information).(b)
Classification. Class II (special controls). A magnetic resonance imaging disposable kit intended for use with a magnetic resonance diagnostic device only is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9.