Search Results

Vantage Galan 3T systems are indicated for use as a diagnostic imaging modality that produces crosssectional 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 K241496. This system is based upon the technology and materials of previously marketed Canon Medical Systems and is intended to acquire and display crosssectional transaxial, coronal, sagittal, and oblique images of anatomic structures of the head or body.

This document describes a 510(k) premarket notification for the Vantage Galan 3T, MRT-3020, V10.0 with AiCE Reconstruction Processing Unit for MR. This submission concerns a modification to an already cleared device, primarily involving the addition of a standard gradient system and the extension of the Precise IQ Engine (PIQE) to new scan families, weightings, and anatomical areas.

Here's a breakdown of the requested information based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly present a table of quantitative acceptance criteria for PIQE performance. Instead, it describes acceptance in qualitative terms based on expert review.

2. Sample Size Used for the Test Set and Data Provenance

• Test Set Sample Size: 106 unique subjects.
• Data Provenance: Two sites in the USA and one in Japan. This data is described as "separate from the training data sets." The document states that the multinational study population is expected to be representative of the intended US population for PIQE, as PIQE is an image post-processing algorithm not disease-specific or dependent on acquisition parameters that might be affected by population variation. Comparisons were internal (each subject as its own control).

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

• Number of Experts: 14 USA board-certified radiologists and cardiologists (3 reviewers per anatomy).
• Qualifications: "USA board-certified radiologists and cardiologists." Specific experience levels (e.g., years of experience) are not provided.

4. Adjudication Method for the Test Set

The document describes a scoring process by multiple reviewers but does not specify a formal adjudication method (e.g., 2+1, 3+1). It states: "scored by 3 reviewers per anatomy in various clinically-relevant categories... Reviewer scoring data was analyzed for reviewer agreement and differences between reconstruction techniques using Gwet's Agreement Coefficient and Generalized Estimating Equations, respectively." This suggests that the scores from the three reviewers were aggregated and analyzed statistically, rather than undergoing a consensus or tie-breaking adjudication process for each individual case.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, What was the Effect Size of How Much Human Readers Improve with AI vs. Without AI Assistance

• MRMC Study: Yes, a multi-site, randomized, blinded clinical image review study was conducted.
• Effect Size (AI-assisted vs. without AI assistance): This was not an AI-assisted reader study. The study compared images reconstructed with the conventional method (matrix expansion with Fine Reconstruction and typical clinical filter) against images reconstructed with PIQE. The purpose was to evaluate the image quality produced by PIQE, not to assess reader performance with or without AI assistance. Therefore, no effect size on human reader improvement with AI assistance is reported. The study aimed to demonstrate that PIQE-reconstructed images are clinically acceptable and offer benefits like sharpness and ringing reduction.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

Yes, a standalone performance evaluation of the PIQE algorithm was conducted through "bench testing." This involved evaluating metrics like Edge Slope Width, Ringing Variable Mean, Signal-to-Noise ratio, and Contrast Change Ratio on typical clinical images from various anatomical regions. This bench testing demonstrated that PIQE "generates images with sharper edges while mitigating the smoothing and ringing effects and maintaining similar or better contrast and SNR."

7. The Type of Ground Truth Used

• For Bench Testing: The "ground truth" implicitly referred to established quantitative image quality metrics (Edge Slope Width, Ringing Variable Mean, Signal-to-Noise ratio, and Contrast Change Ratio) and comparisons against conventional reconstruction methods.
• For Clinical Image Review Study: The "ground truth" was established by expert consensus/evaluation, where 14 board-certified radiologists and cardiologists scored images on various clinically-relevant categories (ringing, sharpness, SNR, overall IQ, and feature conspicuity) using a modified 5-point Likert scale.

8. The Sample Size for the Training Set

The document explicitly states that the "106 unique subjects... from two sites in USA and one in Japan... were scanned... to provide the test data sets (separate from the training data sets)." The sample size for the training set is not provided in the document.

9. How the Ground Truth for the Training Set Was Established

The document does not provide information on how the ground truth for the training set was established, as details about the training set itself are omitted.

Ask a specific question about this device

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 K230355. This system is based upon the technology and materials of previously marketed Canon Medical Systems and is intended to acquire and display crosssectional transaxial, coronal, sagittal, and oblique images of anatomic structures of the head or body.

The provided document describes a 510(k) premarket notification for a modified MRI system (Vantage Galan 3T, MRT-3020, V10.0 with AiCE Reconstruction Processing Unit for MR) by Canon Medical Systems Corporation. The primary purpose of this submission is to demonstrate substantial equivalence to a previously cleared predicate device (Vantage Galan 3T, MRT-3020, V9.0 with AiCE Reconstruction Processing Unit for MR, K230355) despite hardware and software changes.

The document primarily focuses on verifying that the changes do not adversely affect the device's safety and effectiveness and that the modified device maintains performance comparable to the predicate. It does not describe a study proving the device meets specific acceptance criteria in the context of diagnostic accuracy, particularly for an AI-assisted diagnostic device, as the "AiCE Reconstruction Processing Unit" is for image reconstruction, not for AI-based diagnosis.

Therefore, many of the requested fields related to diagnostic performance studies (like multi-reader multi-case studies, expert consensus ground truth, effect size of AI assistance for human readers, or standalone AI performance) are not applicable or not provided in this regulatory submission, as this is a modification of an imaging device itself, not a new AI diagnostic algorithm.

Based on the provided text, here's a breakdown of the requested information:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not present a formal table of "acceptance criteria" for diagnostic accuracy or clinical utility that an AI diagnostic algorithm would typically have, nor does it report performance metrics against such criteria. Instead, the testing focuses on ensuring the new features and hardware maintain image quality, safety, and functionality comparable to the predicate device.

However, the document does list testing performed for new features. We can infer the "acceptance criteria" for these were successful confirmation of functionality and image quality.

2. Sample Size Used for the Test Set and Data Provenance

• 4D Flow: "a phantom with known flow values" and "volunteers." Specific numbers are not provided.
• Zoom DWI: "phantom images" and "representative volunteer images." Specific numbers are not provided.
• 3D-QALAS: "multiple volunteers." Specific numbers are not provided.
• Data Provenance: Not explicitly stated, but given Canon Medical Systems Corporation is based in Japan (manufacturer) and the U.S. (agent), it's likely a mix or either. The studies are described as "bench testing" and using "volunteers," implying prospective data collection for these specific tests.

3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications

• 3D-QALAS: "three experienced reviewers" compared images. Their specific qualifications (e.g., "radiologist with 10 years of experience") are not detailed, but their role as "reviewers" suggests they are professionals qualified to assess image quality.
• Other features (4D Flow, Zoom DWI): The ground truth appears to be established by comparison to known phantom values or visual confirmation of expected image quality improvements (e.g., artifact suppression for Zoom DWI). No external "experts" beyond the testing team are mentioned for establishing ground truth in these cases, which is typical for image quality and functional assessments.

4. Adjudication Method for the Test Set

• For 3D-QALAS, comparison was made by "three experienced reviewers." The document does not specify an adjudication method (e.g., 2+1, 3+1 consensus). It simply states they "compared" the images.
• For other features, adjudication methods are not applicable as the "ground truth" relies on phantom measurements or visual confirmation against expected technical performance.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done

No, a MRMC study comparing human readers with and without AI assistance was not reported. This submission concerns hardware and image reconstruction software changes for an MRI system, not an AI diagnostic algorithm intended for human reader assistance in interpretation. The "AiCE Reconstruction Processing Unit" processes raw MR data into images, it does not interpret those images for diagnostic findings. Therefore, the effect size of human readers improving with AI vs without AI assistance is not relevant or measured here.

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

This refers to the performance of the image reconstruction itself. The testing described (e.g., for 4D Flow, Zoom DWI, 3D-QALAS) demonstrates the standalone technical performance of these new imaging capabilities and the AiCE reconstruction unit in producing images with desired characteristics (e.g., flow visualization, artifact suppression, specific contrast weighting). The "performance" is that the images are generated accurately according to the algorithms' design and meet technical quality metrics.

7. The Type of Ground Truth Used

• 4D Flow: Phantom with "known flow values" (objective physical ground truth) and visual assessment from "volunteer images."
• Zoom DWI: Phantom images and visual assessment from "volunteer images" (technical image quality and accuracy of ADC values).
• 3D-QALAS: Comparison against "reference images published in the literature" (literature-based reference) and assessment by "three experienced reviewers" on image quality metrics (expert qualitative assessment against a standard).
• General System Performance: Compliance with recognized consensus standards (e.g., IEC, NEMA) and comparison to the characteristics of the predicate device (regulatory/technical ground truth).

8. The Sample Size for the Training Set

The document does not describe a "training set" in the context of supervised machine learning for diagnostic tasks. The AiCE (Artificial intelligence Clear Engine) is mentioned as a "Reconstruction Processing Unit," suggesting it's an AI reconstruction algorithm, not an AI diagnostic algorithm. Image reconstruction algorithms may use learned models, but the source document does not provide details on their training data.

9. How the Ground Truth for the Training Set was Established

Not applicable, as a "training set" in the context of a diagnostic AI algorithm is not described. If the AiCE reconstructor uses a deep learning approach, its "training" would likely involve large datasets of raw MR data and corresponding high-quality reference images (e.g., from conventional reconstruction or higher-resolution scans) to learn the mapping from raw data to reconstructed images; however, this level of detail is not provided in a 510(k) summary focused on substantial equivalence of an entire MRI system.

Ask a specific question about this device