(105 days)
3D ASL is a software option intended for use on GE 1.5T and 3.0T MR systems. It is indicated for magnetic resonance imaging of the brain.
3D ASL allows for generation of maps representing blood flow without the use of an exogenous contrast agent. 3D ASL utilizes water in arterial blood as an endogenous contrast media, to visualize tissue perfusion and evaluate cerebral blood flow (CBF).
When interpreted by a trained physician, images generated by 3D ASL provide information that can be useful in determining a diagnosis.
Arterial Spin Labeling (ASL) is an MR technique using the water in arterial blood as an endogenous tracer to evaluate perfusion non-invasively. It provides a non-contrast way to visualize brain perfusion and functional physiology by allowing quantitative cerebral blood flow (CBF) measurements.
GE 3D ASL is an integration of a novel pulsed-continuous labeling technique and a 3D fast spin echo (FSE) acquisition.
The application involves performing inversion of the spins using multiple RF pulses (based on the theory of adiabatic inversion) to label the artery blood spins for the first labeling image. The second image acquisition (also known as control) is performed without inversion of the blood spins. The subtraction of the labeling and control images gives perfusion-weighted images. Additional acquisition of proton density weighted images is then used in combination with the difference images to compute quantitative cerebral blood flow.
The pulsed continuous labeling allows for high labeling efficiency leading to high SNR perfusion images while the 3D FSE readout allows for whole brain coverage and robustness to susceptibility artifacts. The high labeling efficiency is in part due to significantly reduced Magnetization Transfer (MT) effects. The short RF pulses also lead to a significant decrease in the duty cycle when compared to continuous labeling. In addition background suppression is used to reduce the sensitivity of 3D ASL to motion artifacts.
The provided document is a 510(k) summary for GE Healthcare's 3D ASL device. It describes the device, its intended use, and the studies conducted to establish substantial equivalence to predicate devices. However, explicit acceptance criteria values and a detailed breakdown of the study results that directly prove the device meets those criteria are not provided in a concise, easily extractable format. The document focuses on regulatory compliance and general performance validation rather than specific quantitative acceptance criteria for clinical performance metrics.
Based on the provided text, here's what can be extracted and inferred regarding acceptance criteria and the study that proves the device meets them:
1. Table of Acceptance Criteria and Reported Device Performance:
The document lists performance parameters that were measured, but does not provide specific quantitative acceptance criteria or the numerical results of these measurements.
| Acceptance Criteria Category | Specific Acceptance Criteria (Not explicitly stated with numerical values) | Reported Device Performance (Not explicitly stated with numerical values) |
|---|---|---|
| Performance Parameters | The device should demonstrate acceptable levels of: | "The following performance parameters have been measured: Reproducibility, Repeatability, Signal-to-noise ratio (SNR)" |
| Reproducibility | (Implied: High reproducibility) | (Measurement conducted, but specific metrics not provided) |
| Repeatability | (Implied: High repeatability) | (Measurement conducted, but specific metrics not provided) |
| Signal-to-Noise Ratio (SNR) | (Implied: Adequate SNR for clinical interpretation) | (Measurement conducted, but specific metrics not provided) |
| Safety Parameters | The device should demonstrate acceptable levels of: | "The following safety parameters have been measured: Acoustic noise, dB/dt, SAR" |
| Acoustic Noise | (Implied: Within acceptable limits) | (Measurement conducted, but specific metrics not provided) |
| dB/dt | (Implied: Within acceptable limits) | (Measurement conducted, but specific metrics not provided) |
| SAR (Specific Absorption Rate) | (Implied: Within acceptable limits) | (Measurement conducted, but specific metrics not provided) |
| Clinical Evaluation | (Implied: Images generated by 3D ASL are useful for determining a diagnosis when interpreted by a trained physician, and the technique is comparable to predicate devices.) | "The following clinical testing has been performed to validate the 3D ASL technique: Clinical evaluation, Volunteer imaging" |
Study that proves the device meets acceptance criteria:
The document states: "The following clinical testing has been performed to validate the 3D ASL technique: Clinical evaluation, Volunteer imaging." It concludes that "GE Healthcare considers the 3D ASL to be as safe, as effective, and performance is substantially equivalent to the predicate device(s)." This implies that the clinical evaluation and volunteer imaging demonstrated performance comparable to the predicate devices and met the implicit (but not explicitly detailed) acceptance criteria for safety and effectiveness.
2. Sample size used for the test set and the data provenance:
- Sample size for the test set: Not explicitly stated in the provided document. The mention of "Volunteer imaging" suggests healthy volunteers were part of the study, and "Clinical evaluation" would involve patient data, but specific numbers are absent.
- Data provenance: Not explicitly stated. The document does not mention the country of origin of the data or whether it was retrospective or prospective.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
This information is not provided in the document. While it states that images are "interpreted by a trained physician," it does not specify how many physicians were involved in establishing ground truth for the study, nor their specific qualifications.
4. Adjudication method for the test set:
This information is not provided in the document.
5. If a multi-reader, multi-case (MRMC) comparative effectiveness study was done, and the effect size of how much human readers improve with AI vs without AI assistance:
- MRMC study: The document does not describe a multi-reader, multi-case (MRMC) comparative effectiveness study comparing human readers with and without AI assistance. The 3D ASL is described as a software option for generating perfusion maps, which are then interpreted by a trained physician, suggesting it is an imaging technique rather than an AI-assisted diagnostic tool that directly impacts reader performance in the manner of an MRMC study setup.
- Effect size: Not applicable, as an MRMC study of this nature is not described.
6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
The device (3D ASL) itself is essentially a standalone algorithm for generating specific image types (blood flow maps). It operates without human-in-the-loop performance during the image generation process. The output (the maps) then requires interpretation by a human. The "Performance parameters" listed (Reproducibility, Repeatability, SNR) would have been measured on the output of this standalone algorithm.
7. The type of ground truth used:
The exact "type of ground truth" (e.g., pathology, outcomes data) for the clinical evaluation is not specified. However, for a perfusion imaging technique, ground truth might involve comparisons to other established perfusion measurement techniques or clinical outcomes related to cerebral blood flow. Given the context, general "clinical evaluation" and "volunteer imaging" likely referred to the ability of the 3D ASL images to consistently and accurately reflect expected physiological blood flow in volunteers and potentially known clinical conditions in patients.
8. The sample size for the training set:
The document does not mention a "training set" or its sample size. This suggests the 3D ASL is likely a physics-based image reconstruction and processing algorithm rather than a machine learning model that requires explicit training data in the modern sense. Its development would be based on underlying physical principles and signal processing, validated with various datasets during development and verification.
9. How the ground truth for the training set was established:
Not applicable, as a training set for a machine learning model is not mentioned.
§ 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.