(190 days)
RSI-MRI+ is indicated for use as automatic post-acquisition image processing software for analysis of diffusionweighted and anatomical magnetic resonance imaging data.
RSI-MRI+ is intended for automatic fusion of derived diffusion-weighted MRI data with anatomical T2-weighted MR images.
RSI-MRI+ is additionally intended to provide automatic prostate segmentation, and reporting of derived image metrics.
RSI-MRI+ is not intended for use in pediatric populations.
RSI-MRI+ is not intended to diagnose, treat, or prevent diseases or conditions.
RSI-MRI+ is intended to be used in a variety of settings such as hospitals, clinics, and medical offices.
RSI-MRI+ is standalone software that is used by radiologists, and other clinicians to assist with analysis and interpretation of medical images. RSI-MRI+ accepts DICOM images using supported protocols and performs automatic post-acquisition analysis of diffusion-weighted magnetic resonance imaging (DWI) data and optional automated fusion of derived image data with anatomical T2weighted MR images.
Some of the features of RSI-MRI+ include:
- Restricted Signal Map: The derived image data produced by RSI-MRI+ includes an enhanced DWI map (the Restricted Signal Map), which demonstrates improved conspicuity of restricted diffusion compared to standard DWI maps.
- Color Fusion Series: RSI-MRI+ can be configured to produce a color fusion series which overlays the Restricted Signal Map intensity onto the anatomical T2-weighted image series.
- Automated Prostate Segmentation: RSI-MRI+ uses artificial intelligence (Al) powered by a deep learning algorithm to automatically segment the prostate on anatomical T2-weighted images. The segmentation result is provided in the separate Prostate Seqmentation Series.
- Automated Segmentation Report: RSI-MRI+ generates a report of segmentation volume and images of the segmented prostate as a colored outline on the anatomical image.
- Export: RSI-MRI+ outputs are provided in standard DICOM format, which is compatible with most third-party commercial PACS workstation software.
The provided 510(k) summary for RSI-MRI+ (K191278) mentions performance testing related to "Accuracy of automated segmentation compared to manual radiologist segmentations" but does not explicitly state acceptance criteria or the full details of the study that proves the device meets those criteria in a structured format. Similarly, it mentions "Increased conspicuity of the RSI-MRI+ Restricted Signal Map to standard DWI maps," but without specific metrics or studies described.
Based on the available text, I can infer and assemble some of the requested information, but certain details regarding the study for acceptance criteria are not fully elaborated.
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 acceptance criteria with corresponding performance metrics. However, it states that "Accuracy of automated segmentation compared to manual radiologist segmentations" was tested. Without specific thresholds, it's difficult to formulate a complete table.
Inferred Acceptance Criteria (for prostate segmentation) and Reported Performance:
| Acceptance Criteria (Inferred) | Reported Device Performance |
|---|---|
| Prostate Segmentation Accuracy | Successfully compared to manual radiologist segmentations. (Specific metrics like Dice score, Hausdorff distance not provided in the summary) |
| BIFROST (Restricted Signal Map Conspicuity) | "Increased conspicuity of the RSI-MRI+ Restricted Signal Map to standard DWI maps in regions of restricted diffusion." (No quantitative metric provided) |
| Diffusion Signal Normalization | Successful normalization across acquisitions and scanners. (No quantitative metric provided) |
2. Sample Size Used for the Test Set and Data Provenance
The document states:
"Retrospective clinical data (including professionally labeled regions-of-interest, T2-weighted anatomical data, and raw diffusion image series) were used for verification and validation of the diffusion analysis (diffusion signal normalization and Restricted Signal Map conspicuity)."
- Sample size for the test set: Not explicitly stated.
- Data Provenance: Retrospective clinical data. The country of origin of the data is not specified.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The document states:
"...professionally labeled regions-of-interest..."
- Number of experts: Not explicitly stated (it mentions "radiologist segmentations" in plural, implying more than one, but no specific number is given).
- Qualifications of experts: The experts performed "manual radiologist segmentations" and provided "professionally labeled regions-of-interest." This implies they were radiologists, but specific experience levels (e.g., "10 years of experience") are not provided.
4. Adjudication Method for the Test Set
The document does not specify an adjudication method (e.g., 2+1, 3+1) for establishing the ground truth from the expert segmentations. It only mentions "manual radiologist segmentations."
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done
The document does not indicate that an MRMC comparative effectiveness study was done comparing human readers with AI assistance vs. without AI assistance. The performance testing focuses on the standalone algorithm's accuracy for segmentation and conspicuity of the Restricted Signal Map.
6. If a Standalone (i.e. Algorithm Only Without Human-in-the-Loop Performance) Was Done
Yes, a standalone performance evaluation was conducted. The assessment described for "Accuracy of automated segmentation compared to manual radiologist segmentations" and "Increased conspicuity of the RSI-MRI+ Restricted Signal Map" appears to be an algorithm-only evaluation against established ground truth.
7. The Type of Ground Truth Used
For prostate segmentation, the ground truth was established by expert consensus/manual radiologist segmentations.
For diffusion analysis and conspicuity, it was based on "professionally labeled regions-of-interest."
8. The Sample Size for the Training Set
The document does not explicitly state the sample size for the training set. It mentions "RSI-MRI+ uses artificial intelligence (AI) powered by a deep learning algorithm to automatically segment the prostate," implying a training phase, but the size of the training data is not provided in this summary.
9. How the Ground Truth for the Training Set Was Established
The method for establishing ground truth for the training set is not explicitly detailed in the provided text. However, given that "professionally labeled regions-of-interest" were used for validation, it is highly probable that similar expert labeling would have been used to generate ground truth for the training of the deep learning algorithm.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).