Search Results

The MAGNETOM system is indicated for use as a magnetic resonance diagnostic device (MRDD) that produces transverse, sagittal, coronal and oblique cross sectional images, spectroscopic images and/or spectra, and that displays the internal structure and/or function of the head, body, or extremities. Other physical parameters derived from the images and/or spectra may also be produced. Depending on the region of interest, contrast agents may be used. These images and/or spectra and the physical parameters derived from the images and/or spectra when interpreted by a trained physician yield information that may assist in diagnosis.

The MAGNETOM system may also be used for imaging during interventional procedures when performed with MR compatible devices such as in-room displays and MR Safe biopsy needles.

MAGNETOM Vida with software syngo MR XA50A includes new and modified software compared to the predicate device, MAGNETOM Vida with software syngo MR XA31A. A high-level summary of the new and modified hardware and software is provided below:

Software
New Features and Applications

• Deep Resolve Swift Brain is a protocol for fast routine brain imaging primarily based on echo planar imaging (EPI) pulse sequences. Its main enablers are multi-shot (ms) EPI pulse sequence types and a deep learning-based image reconstruction.
• Deep Resolve Boost is a novel deep learning-based image reconstruction algorithm for 2D TSE data, which reconstructs images from k-space raw-data.
• BLADE diffusion is a multi-shot imaging method based on TSE or TGSE (when EPI factor > 1) readout and a BLADE trajectory with diffusion preparation to enable diffusion weighted imaging with reduced sensitivity to B0 inhomogeneity and reduced T2 decay caused image blurring.
• HASTE diffusion (HASTE DIFF) is a single-shot imaging method based on TSE readout with diffusion preparation to enable diffusion weighted imaging with reduced sensitivity to B0 inhomogeneity.

Modified Features and Applications

• Fast GRE RefScan: A speed-optimized reference scan for GRAPPA and SMS kernel calibration for echo planar imaging pulse sequence types.
• Static Field Correction is a reconstruction option reducing susceptibilityinduced distortions and intensity variations.
• Deep Resolve Gain is a reconstruction option which improves the SNR of the scanned images. The functionality has been extended to pulse sequence types SE and TSE DIXON.
• Deep Resolve Sharp is an interpolation algorithm which increases the perceived sharpness of the interpolated images. The functionality has been extended to pulse sequence types SE and TSE DIXON.
• The myExam Angio Advanced Assist provides an assisted and quided workflow for peripheral angiography examination using care bolus. The main advantage of this new workflow is a simplified and improved planning procedure of multi-station peripherical angiography measurements.

Other Modifications and / or Minor Changes

• TSE MoCo is an image-based motion correction in the average-dimension for the TSE pulse sequence type.
• MR Breast Biopsy is improved with an automatic fiducial detection.

This document describes regulatory approval for an MRI system and its software updates, rather than a device with specific performance criteria evaluated against a ground truth in the context of AI or advanced image analysis. Therefore, much of the requested information cannot be extracted from the provided text.

Here's why and what can be extracted:

Why most of the requested information cannot be provided:

• No specific "device" for performance claims: The submission is for a Magnetic Resonance Diagnostic Device (MRDD) system (MAGNETOM Vida) and its software updates (syngo MR XA50A). It's essentially an upgrade to an existing MRI machine, not a new AI/CADx device making specific diagnostic claims that would require detailed performance metrics like sensitivity, specificity, AUC, etc.
• Focus on Substantial Equivalence: The primary objective of this 510(k) summary is to demonstrate that the upgraded MRI system is "substantially equivalent" to a legally marketed predicate device (MAGNETOM Vida with syngo MR XA31A). This demonstration typically involves showing that the new features do not raise new questions of safety or effectiveness and perform as intended, rather than proving a superior or specific diagnostic accuracy against a clinical ground truth.
• "Acceptance Criteria" are not clinical performance metrics: The "acceptance criteria" discussed in the document are related to the successful completion of performance tests (e.g., image quality assessments, software verification and validation) to ensure the device performs as intended and conforms to relevant standards, not clinical performance acceptance thresholds for a diagnostic task.
• No detailed clinical study for performance claims: The document explicitly states: "No additional clinical tests were conducted to support substantial equivalence for the subject devices; however, as stated above, sample clinical images were provided." This confirms that a dedicated clinical study to prove diagnostic performance metrics (e.g., of an AI algorithm) was not performed or presented here. The "clinical publications" referenced are primarily previous research papers related to the underlying technologies (e.g., deep learning for reconstruction, new pulse sequences), not a study directly validating the clinical performance of this specific device's new features against a ground truth.

Information that can be extracted:

Here's what can be provided based on the text:

1. A table of (apparent) acceptance criteria and the reported device performance

Based on the nature of this 510(k) for an MRI system upgrade, the "acceptance criteria" are related to the successful completion of engineering and non-clinical performance evaluations, demonstrating the device performs as intended and is safe and effective when compared to the predicate. No specific numerical performance metrics (e.g., sensitivity, specificity for a diagnostic task) are provided as acceptance criteria or reported performance for a "device" in the context of AI.

2. Sample sized used for the test set and the data provenance

• Sample Size: Not specified for any clinical performance evaluation, as no dedicated clinical study was performed. The document mentions "sample clinical images" were provided, but the quantity or characteristics of these images are not detailed for a statistical test set.
• Data Provenance: Not specified, as no formal clinical test set with defined provenance was used to demonstrate performance against acceptance criteria in the manner requested.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

• Not applicable. No formal expert-adjudicated ground truth for a test set was established for a clinical performance study of this upgraded MRI system. The interpretation of images is generally described as being by a "trained physician" as part of the Indications for Use.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

• Not applicable. No formal adjudication methods were used for a clinical performance test set.

5. 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. A multi-reader multi-case (MRMC) comparative effectiveness study comparing human readers with and without AI assistance was not conducted or reported in this 510(k) summary. The new features like "Deep Resolve Swift Brain" and "Deep Resolve Boost" are described as deep learning-based reconstruction algorithms or methods for faster imaging, not diagnostic AI assistants for human readers.

6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done

• No. This submission is for an MRI system and its software, which includes deep learning for image reconstruction. It is not a standalone diagnostic algorithm that operates without human-in-the-loop for interpretation and performance evaluation. The "nonclinical tests" included "Image quality assessments by sample clinical images" and "Software verification and validation," which are typical for imaging device changes.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

• Not applicable for performance claims of an AI algorithm against a diagnostic ground truth. The "ground truth" in the context of MRI system performance typically refers to physical phantom measurements, simulated data, or established image quality metrics, not clinical diagnostic outcomes adjudicated by experts or pathology for AI algorithm evaluation.

8. The sample size for the training set

• Not specified. While the document mentions "deep learning-based image reconstruction," it does not provide details about the training set size for these deep learning components. The referenced clinical publications (e.g., [2], [3], [4], [5], [6], [8], [9]) discuss deep learning for reconstruction and accelerated MRI but are not specific to the training data used for this particular device's integrated deep learning features.

9. How the ground truth for the training set was established

• Not specified. For deep learning-based image reconstruction, the "ground truth" typically involves high-quality, fully sampled MRI data used to train models to reconstruct images from undersampled or noisy data. However, the exact methodology for establishing this ground truth for the incorporated deep learning models is not detailed in this regulatory summary.

Ask a specific question about this device