(104 days)
The Swoop Point-of-Care Magnetic Resonance Imaging System is a bedside magnetic resonance imaging device for producing images that display the internal structure of the head where full diagnostic examination is not clinically practical. When interpreted by a trained physician, these images provide information that can be useful in determining a diagnosis.
The Swoop™ Point-of-Care MRI System is a portable MRI device that allows for patient bedside imaging. The system enables visualization of the internal structures of the head using standard magnetic resonance imaging contrasts. The main interface is a commercial off-the-shelf device that is used for operating the system, providing access to patient data, exam setup, exam execution, viewing MRI image data for quality control purposes, and cloud storage interactions. The system can generate MRI data sets with a broad range of contrasts. The Swoop™ Point-of-Care MRI System user interface includes touchscreen menus, controls, indicators, and navigation icons that allow the operator to control the system and to view imagery.
This subject device in this submission includes a change to the image reconstruction algorithm of the Swoop POC MRI device for the T1W, T2W, and FLAIR sequences. The image reconstruction change utilizes deep learning to provide improved image quality, specifically in terms of reductions in image noise and blurring. This change replaces the non-uniform FFT-gridding operation in the reconstruction pipeline with Advanced Gridding and adds an Advanced Denoising step in the image postprocessing stage. All other sections of the image reconstruction pipeline are unchanged with respect to those used in the previously cleared system (K201722/K211818).
The provided text describes the acceptance criteria for a new image reconstruction algorithm in the Swoop™ Point-of-Care Magnetic Resonance Imaging (POC MRI) System and the testing conducted to demonstrate substantial equivalence.
Here's a breakdown of the requested information:
1. A table of acceptance criteria and the reported device performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Advanced Reconstruction Verification | |
| Advanced reconstruction models do not alter image features or introduce artifacts. | Passed |
| Ability for expert-mode users to toggle between linear reconstruction and advanced reconstruction. | Passed |
| Image quality with advanced reconstruction is acceptable. | Passed (specifically, provides "improved image quality, specifically in terms of reductions in image noise and blurring.") |
| Basic software functionality is unchanged between releases. | Passed |
| NESSUS scan test to verify any vulnerabilities and serve as a security baseline. | Passed |
| Advanced Reconstruction Performance Analysis | |
| Robustness, stability, and generalizability of the advanced reconstruction models. | Passed |
| Image Performance | |
| Meets all image quality criteria (based on NEMA and ACR standards). | Passed |
| Advanced Reconstruction Validation | |
| Meets user needs and performs as intended. | Passed |
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
The document does not explicitly state the sample size for the test set or the data provenance (country of origin, retrospective/prospective). It mentions "testing to verify image quality with advanced reconstruction is acceptable" and "Validation studies to confirm that the device meets user needs and performs as intended" but lacks specific details about the patient data used for these tests.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)
The document does not specify the number of experts or their qualifications used to establish ground truth for the test set. It mentions "expert-mode users" and "trained physician" (in the Indications for Use), but no further details about their roles in evaluating the test set.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
The document does not describe any specific adjudication method (like 2+1 or 3+1) used for the 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
The document does not mention a multi-reader multi-case (MRMC) comparative effectiveness study or any effect size related to human reader improvement with AI assistance. The focus of the submission is on the image reconstruction algorithm itself, which utilizes deep learning to improve image quality, rather than focusing on a human-in-the-loop performance study.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
Yes, standalone performance was assessed for the algorithm. The testing described focuses on the "image reconstruction algorithm" and its "image quality" improvements, such as "reductions in image noise and blurring." This implies an evaluation of the algorithm's output (images) without necessarily involving human interpretation as the primary endpoint for all tests. The "Advanced Reconstruction Verification," "Advanced Reconstruction Performance Analysis," and "Image Performance" tests are all indicative of standalone algorithm evaluation.
7. The type of ground truth used (expert concensus, pathology, outcomes data, etc)
The document does not explicitly state the type of ground truth used for evaluating the image quality and performance of the advanced reconstruction algorithm. Given the nature of MRI image quality assessment, it is highly likely that expert visual assessment and potentially quantitative metrics (derived from NEMA and ACR standards, which are referenced) against established benchmarks or phantom data would have been used. However, "expert consensus," "pathology," or "outcomes data" are not directly cited as the ground truth.
8. The sample size for the training set
The document does not specify the sample size used for the training set of the deep learning image reconstruction algorithm.
9. How the ground truth for the training set was established
The document does not describe how the ground truth for the training set was established. It only states that the device "utilizes deep learning to provide improved image quality" but does not elaborate on the training process or ground truth generation.
§ 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.