Search Results

Ask a specific question about this device

Ask a specific question about this device

Ask a specific question about this device

Ask a specific question about this device

Philips Magnetic Resonance (MR) systems are Medical Electrical Systems indicated for use as a diagnostic device.

This MR system enables trained physicians to obtain cross-sectional images, spectroscopic images and/or spectra of the internal structure of the head, body or extremities, in any orientation, representing the spatial distribution of protons or other nuclei with spin.

Image appearance is determined by many different physical properties of the tissue and the anatomy, the MR scan technique applied, and presence of contrast agents. The use of contrast agents for diagnostic imaging applications should be performed consistent with the approved labeling for the contrast agent.

The trained clinical user can adjust the MR scan parameters to customize image appearance, accelerate image acquisition, and synchronize with the patient's breathing or cardiac cycle.

The systems can use combinations of images to produce physical parameters, and related derived images. Images, spectra, and measurements of physical parameters, when interpreted by a trained physician, provide information that may assist diagnosis and therapy planning. The accuracy of determined physical parameters depends on system and scan parameters and must be controlled and validated by the clinical user.

In addition, the Philips MR systems provide imaging capabilities, such as MR fluoroscopy, to guide and evaluate interventional and minimally invasive procedures in the head, body and extremities. MR Interventional procedures, performed inside or adjacent to the Philips MR system, must be performed with MR Conditional or MR Safe instrumentation as selected and evaluated by the clinical user for use with the specific MR system configuration in the hospital. The appropriateness and use of information from a Philips MR system for a specific interventional procedure and specific MR system configuration must be validated by the clinical user.

The subject Ingenia, Ingenia CX, Ingenia Elition S/X, Ingenia Ambition S/X, BlueSeal SE/XE/QE, MR 5300 and MR 7700 MR Systems including Upgrades are 60 cm and 70 cm bore 1.5 and 3.0 Tesla (1.5T and 3.0T) Magnetic Resonance Diagnostic Devices.

The following software features are contained in software R13 for the Ingenia, Ingenia CX, Ingenia Elition S/X, Ingenia Ambition S/X, BlueSeal SE/XE/QE, MR 5300 and MR 7700 MR Systems including Upgrades:

• SmartHeart includes an automated cardiac MRI planning feature that uses a CNN-based deep learning model to analyze 3D survey images and suggest standard cardiac imaging planes. The technologist can fully edit these suggestions if needed. The model was trained on de-identified 3D survey images of adult patients from multiple geographies using expert-annotated reference planes. Testing on separate datasets showed performance is within the range expected from trained human operators. This feature supports workflow optimization only and does not perform any image reconstruction or diagnosis.

• CardiacQuant Perfusion enables assessment of the myocardial blood flow. It is a dual sequence method to allow for the simultaneous acquisition of both the Arterial Input Function and the perfusion data.

• Cardiac Motion Correction enables inline motion correction for the following 2D CMR sequences: Late Gadolinium enhancement (LGE), CardiacQuant Perfusion and T1 mapping. It makes use of Fast Elastic Image Registration between images across time points for the same slice to compensate for motion caused by breathing or differences in cardiac phase.

• CINE FreeBreating is a free-breathing sequence for Cine imaging using respiratory gating to reduce respiratory motion. It allows the user to perform a 2D CINE acquisition in a patient without breath hold commands.

• 4D MR-RT is a free-breathing scanning method to acquire 3D images for multiple respiratory phases. 4D MR-RT enables radiotherapy planning using MR for moving targets in the abdomen, and breath holds are not mandatory.

• IRIS is a Multi-shot SE-EPI diffusion imaging technique which enables higher resolution imaging of the (female) pelvis, prostate and breast.

The introduction of these software features required updates only to the MR system software.

The provided FDA 510(k) clearance letter and summary for K253648 focuses on the substantial equivalence of the Philips MR Systems (Ingenia family, MR 5300, MR 7700) to a predicate device, primarily through non-clinical performance testing and outlining new software features.

However, the document does not contain specific acceptance criteria or detailed study results for the performance of the new software features, particularly the SmartHeart AI-based cardiac MRI planning feature. It states general compliance with international and FDA consensus standards for safety and essential performance of medical electrical equipment.

The information regarding acceptance criteria and a study proving those criteria are met, as requested in your prompt, is largely absent from this particular FDA document for the SmartHeart feature. The provided text indicates that the SmartHeart feature is for "workflow optimization only and does not perform any image reconstruction or diagnosis." This classification likely means less rigorous clinical performance studies are required compared to a diagnostic AI device.

Below, I've extracted all available relevant information and noted where information is missing based on your request.

1. Table of Acceptance Criteria and Reported Device Performance

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

• SmartHeart:
  • Test Set Size: Not specified in the document.
  • Data Provenance (Training Data): "de-identified 3D survey images of adult patients from multiple geographies." (The text refers to training data for provenance, not explicitly for the test set, but implies similar data characteristics for testing.)
  • Data Provenance (Test Set): "separate datasets" (No explicit geography or whether prospective/retrospective mentioned for the test set, only for training).

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

• SmartHeart:
  • Number of Experts: Not specified.
  • Qualifications of Experts: "expert-annotated reference planes" (No specific qualifications like "radiologist with 10 years of experience" are provided).

4. Adjudication Method for the Test Set

• SmartHeart: Not specified. The document only mentions "expert-annotated reference planes" for training, implying a consensus or single expert approach for ground truth creation. No specific adjudication method for the test set's ground truth is detailed.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size

• SmartHeart: No MRMC comparative effectiveness study was explicitly described in this document for the SmartHeart feature in the context of human reader improvement. The statement "performance is within the range expected from trained human operators" suggests a comparison, but not an MRMC study with specific effect sizes of AI assistance.
• The document states that the SmartHeart feature "supports workflow optimization only and does not perform any image reconstruction or diagnosis," which might explain the absence of an MRMC study focused on diagnostic improvement.

6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study Was Done

• SmartHeart: Yes, a standalone evaluation of the algorithm's performance was implied by the statement: "Testing on separate datasets showed performance is within the range expected from trained human operators." This indicates the algorithm's output was compared against expert annotations directly.

7. The Type of Ground Truth Used

• SmartHeart: "expert-annotated reference planes." This implies expert consensus or individual expert delineation of anatomical landmarks for cardiac imaging planes.

8. The Sample Size for the Training Set

• SmartHeart: Not specified in the document.

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

• SmartHeart: "expert-annotated reference planes."

Summary of Missing Information:

The provided FDA document is a clearance letter, which typically summarizes the information provided in the 510(k) submission, rather than containing exhaustive details of every study. Key missing details for a comprehensive understanding of the SmartHeart feature's validation include:

• Specific metrics for "performance within the range expected from trained human operators" (e.g., plane alignment accuracy, deviation from expert-defined landmarks).
• Exact sample sizes for both training and test sets.
• Specific countries of origin for the data, and whether the test set was prospective or retrospective.
• The number and precise qualifications of the experts establishing ground truth.
• The specific adjudication method used to determine ground truth.
• Details of any comparative studies with human performance.

Ask a specific question about this device

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, depending on optional local coils that have been configured with the system, 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 Flow.Ace & MAGNETOM Flow.Plus with software Syngo MR XB10 include new and modified hardware and software compared to the predicate devices, MAGNETOM Flow.Ace & MAGNETOM Flow.Plus with software syngo MR XA70A.

New compared to predicate:

• Spine support respiratory (Cushion) as a part of BM Spine Coil Set 1.5T (including new surface material)
• Gradient Configuration Upgrade

Modified same as predicate, but with new claim introduced:

• PETRA (new claim for the existing sequence)

In addition, the following hardware and software are transferred from the reference device MAGNETOM Flow.Neo with software Syngo MR XB10 (K252838), to the subject devices without any modifications:

Hardware (New compared to predicate, same as reference):

• BioMatrix Dockable Table with / without eDrive
• Comfort Sound: Cushion

Hardware (Modified compared to predicate, same as reference):

• Comfort Sound: BM Head/Neck Coil
• Relocatable Option

Software (New compared to predicate, same as reference):

• Open Workflow

Software (Modified compared to predicate, same as reference):

• BioMatrix Motion Sensor (SAMER)
• CS_VIBE
• SPAIR FatSat Improvements: SPAIR "Abdomen & Pelvis" mode and SPAIR Breast mode
• Deep Resolve Boost for FL3D_VIBE and SPACE
• Deep Resolve Sharp for FL3D_VIBE and SPACE
• Preview functionality for Deep Resolve Boost
• myExam Implant Suite
• GRE_PC
• Open Recon 2.0
• Deep Resolve Boost for TSE
• "MTC Mode" for SPACE

Other Modifications and / or Minor Changes (New compared to predicate, same as reference):

• Eco Power Mode Pro

Other Modifications and / or Minor Changes (Modified compared to predicate, same as reference):

• Off-Center Planning Support
• Flip Angle Optimization (Lock TR and FA)
• ID Gain (re-naming)
• Marketing bundle "myExam Companion"

Acceptance Criteria and Study Details for Siemens MAGNETOM Flow.Ace and Flow.Plus

Based on the provided 510(k) clearance letter, the acceptance criteria and study details are as follows. It's important to note that this document primarily focuses on demonstrating substantial equivalence to a predicate device rather than presenting a detailed de novo clinical trial for device efficacy. Therefore, specific metrics like sensitivity, specificity, or AUC for diagnostic performance are not explicitly stated as acceptance criteria in the typical sense for a new diagnostic claim.

The acceptance criteria are generally focused on demonstrating that the new and modified features of the MAGNETOM Flow.Ace and Flow.Plus systems maintain equivalent safety and performance to the predicate device.

1. Table of Acceptance Criteria and Reported Device Performance

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

The document does not explicitly state a specific sample size for a test set in the context of clinical images or patient data.

• Data Provenance: "Sample clinical images" were provided as "claim evidence." The origin (e.g., country) and whether this data was retrospective or prospective are not specified in the provided text. The clinical publications referenced are peer-reviewed articles, which would typically involve patient data.

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

The document does not provide information regarding the number of experts, their qualifications, or how ground truth was established for any "test set" of images. The phrase "interpreted by a trained physician" is used in the Indications for Use, which is a general statement about MR diagnostic devices.

4. Adjudication Method for the Test Set

The document does not specify any adjudication method (e.g., 2+1, 3+1, none) for a test set.

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

No, an MRMC comparative effectiveness study was not explicitly done or reported in this document. The submission relies on "sample clinical images" as "claim evidence" and references existing clinical publications for certain features to demonstrate substantial equivalence, rather than a direct comparative study showing improvement with AI assistance.

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

The document discusses "software verification and validation" and "nonclinical tests" which would imply standalone performance testing of the algorithms and features. However, it does not provide specific metrics or results for standalone algorithm performance (e.g., sensitivity, specificity, or AUC if applicable to specific features). The focus is on the performance of the integrated system.

7. The Type of Ground Truth Used

The document does not explicitly state the type of ground truth used for any specific evaluation. The "Indications for Use" mention that images "when interpreted by a trained physician yield information that may assist in diagnosis." For the referenced clinical publications, the ground truth would depend on the methodology of those studies (e.g., pathology, clinical follow-up, expert consensus in a research setting), but this information is external to the 510(k) summary itself.

8. The Sample Size for the Training Set

The document does not provide any information regarding the sample size for a training set. This is a 510(k) submission for an updated MR system, not a de novo AI/ML device where training data details are typically prominent. While features like "Deep Resolve Boost" and "Deep Resolve Sharp" likely leverage AI/ML, the details of their development and training are not disclosed in this summary.

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

The document does not provide any information on how ground truth was established for a training set.

Ask a specific question about this device

The SIGNA™ Bolt system is a whole body magnetic resonance scanner designed to support high resolution, high signal-to-noise ratio, and short scan times. It is indicated for use as a diagnostic imaging device to produce axial, sagittal, coronal, and oblique images, spectroscopic images, parametric maps, and/or spectra, dynamic images of the structures and/or functions of the entire body, including, but not limited to, head, neck, TMJ, spine, breast, heart, abdomen, pelvis, joints, prostate, blood vessels, and musculoskeletal regions of the body. Depending on the region of interest being imaged, contrast agents may be used.

The images produced by the SIGNA™ Bolt system reflect the spatial distribution or molecular environment of nuclei exhibiting magnetic resonance. These images and/or spectra when interpreted by a trained physician yield information that may assist in diagnosis.

SIGNA™ Bolt is a whole body magnetic resonance scanner designed to support high resolution, high signal-to-noise ratio, and short scan times, and is designed for improved patient comfort and workflow. The system features a 3.0T superconducting magnet with a 70 cm bore size and can image in the sagittal, coronal, axial, oblique, and double oblique planes, using various pulse sequences, imaging techniques and reconstruction algorithms. SIGNA™ Bolt is designed to conform to NEMA DICOM standards.

The SIGNA™ Bolt system will be offered as two commercial configurations with the following features:

• Magnet: 3.0T superconducting magnet with a wide (70 cm) bore size and active shielding
• Maximum Gradient Strength: 80 mT/m (SuperXG Gradient), 65 mT/m (SuperXF Gradient)
• Maximum Slew Rate: 200 T/m/s (SuperXG Gradient and SuperXF Gradient)
• RF Transmit: A liquid cooled In-Scan-Room RF transmit architecture with a peak power capability of 36 kW and 3.0T Platform Body Coil
• RF Receive Chain: 162 Ch available (SuperXG Gradient), 130 Ch available (SuperXF Gradient)
• Patient Table: Detachable SIGNA One Patient Table with embedded 3.0T AIR PA XL coil and up to four 32-channel high density auto-coil sensing connection ports
• Power Rating: 113 kVA (SuperXG Gradient), 90 kVA (SuperXF Gradient)
• Software: Software platform featuring various productivity enhancement features, designed to improve workflow and reduce scan time
  • AIRx (previously cleared in K183231) – AI-based automated slice prescription tool now extended with new deep learning models for spine and prostate imaging
  • SIGNA One Camera – Real-time AI-enabled image guidance that assists with automated patient positioning
• Gating Options: Wired, wireless, and contactless physiological gating options

This document outlines the acceptance criteria and supporting studies for the SIGNA™ Bolt device, based on the provided FDA 510(k) clearance letter.

Key Features and AI/ML Components of SIGNA™ Bolt:

The SIGNA™ Bolt system includes several AI/Machine Learning components:

1. AIRx: An AI-based automated slice prescription tool, previously cleared for brain and knee imaging (K183231), now extended with new deep learning models for spine and prostate imaging.
2. SIGNA One Camera: Real-time AI-enabled image guidance that assists with automated patient positioning.
3. Contactless Gating: This feature leverages underlying physiological signal detection that might involve advanced signal processing or AI techniques, though the document primarily describes its functional outcome.

Acceptance Criteria and Reported Device Performance

The following table summarizes the acceptance criteria and reported performance for the AI/ML components of the SIGNA™ Bolt device:

Study Details for AI/ML Components:

1. SIGNA One Camera

• Sample Size for Test Set: Data collected from 80 volunteers.
• Data Provenance: US and China (to ensure diverse datasets).
• Number of Experts & Qualifications for Ground Truth: Not explicitly stated for this component. Ground truth is described as "MR system coordinates of the camera-predicted anatomical landmarks against ground truth annotations," suggesting a technical or measurement-based ground truth rather than expert reads.
• Adjudication Method: Not specified.
• MRMC Comparative Effectiveness Study: Yes, a "time on task study" was conducted with 11 MR Scan Operators comparing the AI-powered workflow to conventional laser landmarking.
  • Effect Size: The camera workflow "consistently enabled faster setup times for landmarking." Specific quantitative improvement (e.g., % reduction in time) is not provided in text.
• Standalone Performance: Yes, "Accuracy was evaluated by comparing the MR system coordinates of the camera-predicted anatomical landmarks against ground truth annotations." This indicates an algorithm-only evaluation.
• Type of Ground Truth: MR system coordinates.
• Sample Size for Training Set: Not explicitly stated, but the test dataset was "entirely separate from the training and validation datasets."
• Ground Truth for Training Set: Not specified, but likely established in a similar manner to the test set (MR system coordinates or similar technical measurements).

2. AIRx Spine

• Sample Size for Test Set: 376 subjects.
• Data Provenance: Multiple clinical sites and internal GE HealthCare sites.
• Number of Experts & Qualifications for Ground Truth: Not explicitly stated. Ground truth is implied to be established for "accurate multi-slice, multi-angle prescriptions."
• Adjudication Method: Not specified.
• MRMC Comparative Effectiveness Study: Yes, "Comparative studies demonstrated that AIRx Spine reduced scan prescription times compared to manual workflows and minimized inter-operator variability."
  • Effect Size: "Reduced scan prescription times" and "minimized inter-operator variability" (confirmed by Structural Similarity Index (SSIM) analysis and visual comparisons). Specific quantitative improvement is not provided.
• Standalone Performance: Yes, "Performance testing was conducted on the AIRx Spine deep learning models," indicating an algorithm-only evaluation.
• Type of Ground Truth: Not explicitly stated but implied to be based on accurate anatomical prescriptions suitable for diagnostic imaging. SSIM analysis and visual comparisons suggest a comparison against an ideal or expert-defined prescription.
• Sample Size for Training Set: Not explicitly stated, but the test dataset was "held separate from training and validation data."
• Ground Truth for Training Set: Not specified, but likely established to enable the model to learn "patient-specific spinal curvatures" and "accurate multi-slice, multi-angle prescriptions."

3. AIRx Prostate

• Sample Size for Test Set: 785 exams.
• Data Provenance: Clinical sites in the US and Europe.
• Number of Experts & Qualifications for Ground Truth: Not explicitly stated.
• Adjudication Method: Not specified.
• MRMC Comparative Effectiveness Study: Not explicitly mentioned for this specific feature in the provided text.
• Standalone Performance: Yes, "Performance testing was conducted on the six deep learning models that comprise the AIRx Prostate feature," evaluating automated prostate scan plane prescription, indicating an algorithm-only evaluation.
• Type of Ground Truth: Not explicitly stated but implied to be based on accurate anatomical prescriptions for the prostate, using SSFSE localizer images.
• Sample Size for Training Set: Not explicitly stated, but the test dataset was "kept separate from the training and validation data."
• Ground Truth for Training Set: Not specified, but likely established to enable the model to learn "automated prostate scan plane prescription."

4. Contactless Gating

• Sample Size for Test Set: Not explicitly stated for this particular feature's performance validation.
• Data Provenance: Not specified.
• Number of Experts & Qualifications for Ground Truth: Not specified.
• Adjudication Method: Not specified.
• MRMC Comparative Effectiveness Study: Not mentioned.
• Standalone Performance: Yes, "Verification and validation testing confirmed that the contactless gating feature meets its performance specifications by accurately detecting and displaying respiratory and peripheral cardiac waveforms," indicating a system performance evaluation.
• Type of Ground Truth: Underlying physiological waveforms (respiratory and cardiac).
• Sample Size for Training Set: Not specified.
• Ground Truth for Training Set: Not specified, but likely established from physiological signal data.

Overall Conclusion from Performance Testing:

GE HealthCare concludes that the SIGNA™ Bolt is as safe and effective, with performance substantially equivalent to the predicate device, based on the nonclinical testing, including extensive software verification and validation, as well as specific performance evaluations for its new AI-enabled features. No clinical studies were required to support substantial equivalence.

Ask a specific question about this device

The SIGNA™ Sprint Select is a whole body magnetic resonance scanner designed to support high resolution, high signal-to-noise ratio, and short scan times. It is indicated for use as a diagnostic imaging device to produce axial, sagittal, coronal, and oblique images, spectroscopic images, parametric maps, and or spectra, dynamic images of the structures and/or functions of the entire body, including, but not limited to, head, neck, TMJ, spine, breast, heart, abdomen, pelvis, joints, prostate, blood vessels, and musculoskeletal regions of the body. Depending on the region of interest being imaged, contrast agents may be used.

The images produced by SIGNA™ Sprint Select reflect the spatial distribution or molecular environment of nuclei exhibiting magnetic resonance. These images and/or spectra when interpreted by a trained physician yield information that may assist in diagnosis.

SIGNA™ Sprint Select is a whole-body magnetic resonance scanner designed to support high resolution, high signal-to-noise ratio, and short scan time. The system uses a combination of time-varying magnet fields (Gradients) and RF transmissions to obtain information regarding the density and position of elements exhibiting magnetic resonance. The system can image in the sagittal, coronal, axial, oblique, and double oblique planes, using various pulse sequences, imaging techniques and reconstruction algorithms. The system features a 1.5T superconducting magnet with 70cm bore size. The system is designed to conform to NEMA DICOM standards (Digital Imaging and Communications in Medicine).

N/A

Ask a specific question about this device

The Ascent3T Neonatal Magnetic Resonance Imaging System (Ascent3T) is a whole-body magnetic resonance scanner designed for neonates and infants. The system can produce cross-sectional images of the internal structure of the head, body or extremities in any orientation.

Images produced by the Ascent3T show the spatial distribution of protons exhibiting magnetic resonance. Images produced by the Ascent3T, when interpreted by a trained physician, may provide information useful in diagnosis.

The Ascent3T Neonatal Magnetic Resonance Imaging System is suitable for neonates and infants weighing up to 9kg (19.8 lbs).

The Ascent3T Neonatal Magnetic Resonance Imaging System (Ascent3T) is a high-field magnetic resonance imaging system, appropriately sized and optimized for the neonate and infant population, with a format that allows siting near point of care. The Ascent3T presents a solution for the technical limitations associated with using an adult-size MRI system and provides clinicians with an improved ability to visualize and diagnose disease in the neonatal patient population.

The Ascent3T is equipped with a small format superconducting magnet designed for neonate applications. The system is designed to operate at 3.0 Tesla and achieves a high level of homogeneity over a 24cm diameter spherical volume using passive shims. The magnet requires a minimal amount of helium and no quench pipe. These features, in combination with the size and weight of the magnet, support near-patient siting.

The Ascent3T patient table is detachable and can serve as a patient transport device. The patient table includes a tabletop cradle with features for securing the patient during scanning. The patient table is mobile, providing flexibility in workflow based on institutional needs and preferences.

The Ascent3T contains a menu of pulse sequences intended to provide the user with a variety of sequences useful for producing images for diagnostic purposes.

Key Features of the Ascent3T:

• 3T superconducting magnet with 25cm patient bore.
• Minimal helium capacity with no quench pipe required.
• Gradient system: 80 mT/m maximum amplitude per axis, 300 mT/m/ms slew rate per axis.
• Real-time SAR monitoring and alerts with Normal and First-Level Controlled Operating Modes.
• Capable of producing images in axial, sagittal, coronal, and oblique orientations.
• Accommodates neonates and infants weighing up to 9 kg (19.8 lbs).
• Detachable, mobile patient table with built-in safety features.

N/A

Ask a specific question about this device

AIR Recon DL is a deep learning based reconstruction technique that is available for use on GE HealthCare 1.5T, 3.0T, and 7.0T MR systems. AIR Recon DL reduces noise and ringing (truncation artifacts) in MR images, which can be used to reduce scan time and improve image quality. AIR Recon DL is intended for use with all anatomies, and for patients of all ages. Depending on the anatomy of interest being imaged, contrast agents may be used.

AIR Recon DL is a software feature intended for use with GE HealthCare MR systems. It is a deep learning-based reconstruction technique that removes noise and ringing (truncation) artifacts from MR images. AIR Recon DL is an optional feature that is integrated into the MR system software and activated through purchasable software option keys. AIR Recon DL has been previously cleared for use with 2D Cartesian, 3D Cartesian, and PROPELLER imaging sequences.

The proposed device is a modified version of AIR Recon DL that includes a new deep-learning phase correction algorithm for applications that create multiple intermediate images and combine them, such as Diffusion Weighted Imaging where multiple NEX images are collected and combined. This enhancement is an optional feature that is integrated into the MR system software and activated through an additional purchasable software option key (separate from the software option keys of the predicate device).

This document describes the acceptance criteria and the studies conducted to prove the performance of the AIR Recon DL device, as presented in the FDA 510(k) clearance letter.

1. Table of Acceptance Criteria and Reported Device Performance

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

• Nonclinical Testing:
  • Phantom testing was conducted for the DLPC and PC-ARDL models. No specific sample size (number of phantom scans) is provided, but it implies a sufficient number for evaluation.
• In-Vivo Performance Testing:
  • ADC Accuracy: Diffusion-weighted brain images were acquired at 1.5T with b-values = 50, 400, 800, 1200 s/mm². The number of subjects is not explicitly stated, but it's referred to as "diffusion images" and "diffusion-weighted brain images."
  • Low-Contrast Detectability: Raw data from 4 diffusion-weighted brain scans were used.
• Quantitative Post Processing (Repeatability Study):
  • 6 volunteers were recruited. 2 volunteers scanned on a 1.5T scanner, 4 on a 3T scanner.
  • Scanned anatomical regions included brain, spine, abdomen, pelvis, and breast.
  • Each sequence was repeated 4 times.
  • Data Provenance: The document states "in-vivo data" and "volunteer scanning was performed simulating routine clinical workflows." This suggests prospective scanning of human subjects, likely in a controlled environment. The country of origin is not specified, but given the FDA submission, it's likely U.S. or international data meeting U.S. standards. The statement "previously acquired de-identified cases" for the Clinical Image Quality Study refers to retrospective data for that specific study, but the volunteer scanning for repeatability appears prospective.
• Clinical Image Quality Study:
  • 34 datasets of previously acquired de-identified cases.
  • Data Provenance: "previously acquired de-identified cases" indicates retrospective data. The country of origin is not specified.

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

• Nonclinical Testing: Ground truth established through phantom measurements and expected physical properties (e.g., signal bias, noise floor). No human experts involved in establishing ground truth here.
• In-Vivo Performance Testing:
  • ADC Accuracy: "Average ADC values were measured from regions of interest in the lateral ventricles." This implies expert selection of ROIs, but the number of experts is not specified. The ground truth for ADC is the expected isotropic Gaussian diffusion in these regions.
  • Low-Contrast Detectability: "The contrast ratio and contrast-to-noise ratio for each of the inserts were measured." This is a quantitative measure, not explicitly relying on expert consensus for ground truth on detectability, but rather on the known properties of the inserted synthetic objects.
• Quantitative Post Processing:
  • ADC Repeatability: Ground truth for repeatability is based on quantitative measurements and statistical analysis (coefficient of variability). ROI placement would typically be done by an expert, but the number is not specified.
  • Phase Correction Effectiveness: Ground truth is based on the theoretical expectation of signal distribution in real/imaginary channels after ideal phase correction.
• Clinical Image Quality Study:
  • One (1) U.S. Board Certified Radiologist was used.
  • Qualifications: "U.S. Board Certified Radiologist." No explicit number of years of experience is stated, but Board Certification indicates a high level of expertise.

4. Adjudication Method for the Test Set

• Nonclinical/Phantom Testing: No explicit adjudication method described beyond passing defined acceptance criteria for quantitative metrics.
• In-Vivo Performance Testing: Quantitative measurements (ADC values, contrast ratios, CNR) were used. Paired t-tests were conducted, which is a statistical comparison method, not an adjudication process as typically defined for expert readings.
• Quantitative Post Processing: Quantitative measurements and statistical analysis (coefficient of variability, comparison of real/imaginary channels).
• Clinical Image Quality Study: A single U.S. Board Certified Radiologist made the assessment. There is no stated adjudication method described, implying a single-reader assessment for clinical image quality.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• An MRMC comparative effectiveness study was not explicitly described as a formal study design in the provided text.
• The "Clinical Image Quality Study" involved only one radiologist, so it does not qualify as an MRMC study.
• There is no reported effect size of how much human readers improve with AI vs. without AI assistance. The study rather focused on the AI-reconstructed images' standalone diagnostic quality.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study

• Yes, performance was evaluated in a standalone manner.
  • Nonclinical Testing: Phantom studies directly evaluate the algorithm's output against known physical properties and defined metrics.
  • In-Vivo Performance Testing: ADC accuracy and low-contrast detectability were measured directly from the reconstructed images, which is a standalone evaluation of the algorithm's quantitative output.
  • Quantitative Post Processing: Repeatability and effectiveness of phase correction in real/imaginary channels are algorithm-centric evaluations.
  • Even the clinical image quality study, while involving a human reader, assessed the standalone output of the algorithm (AIR Recon DL with Phase Correction) for diagnostic quality.

7. Type of Ground Truth Used

• Expert Consensus: Not explicitly stated as the primary ground truth for quantitative metrics, but one radiologist's assessment served as the primary clinical ground truth.
• Pathology: Not used as ground truth in the provided study descriptions. While some datasets "included pathological features such as prostate cancer... hepatocellular carcinoma," the assessment by the radiologist was on "diagnostic quality" of the images, not a comparison against pathology reports for definitive disease identification.
• Outcomes Data: Not used as ground truth.
• Other:
  • Physical Properties/Known Standards: For phantom testing (e.g., signal bias, noise floor, SNR, sharpness), and for theoretical expectations of ADC values in specific regions (lateral ventricles).
  • Known Synthetic Inserts: For low-contrast detectability.
  • Theoretical Expectations: For phase correction effectiveness (signal in real, noise in imaginary).

8. Sample Size for the Training Set

• The document does not provide any specific sample size for the training set used for the deep learning models (DLPC and PC-ARDL). It only states that the models are "deep learning-based."

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

• The document does not provide any information on how the ground truth for the training set was established. It only describes the testing of the final, trained models.

Ask a specific question about this device