(265 days)
In Vision Precision LVEF is used to process previously acquired trans thoracic cardiac ultrasound images, and manipulate and make measurements on images using an ultrasound device, personal computer, or a compatible DICOM compliant PACS system to provide an automated estimation of LVEF. This measurement can be used to assist the clinician in a cardiac evaluation. In Vision Precision is indicated for use in patients 22 years and older by sonographers and physicians evaluating cardiac ultrasound.
InVision Precision LVEF is a software as a medical device (SaMD), manufactured by InVision Medical Technology Corporation, intended as an aid in diagnostic review and analysis of echocardiographic data, including the evaluation of left ventricular ejection fraction (LVEF) in cardiovascular ultrasound images in DICOM format. The software interfaces with data files uploaded to a PACS by any ultrasound or data collection equipment. It selects a set of echocardiogram videos of the correct view and generates semi-automatic segmentations of the left ventricle using a machine learning algorithm to form the basis for the calculator of the LVEF output. The analysis results are visualized by the clinician's integrated image view application as adjustable annotations. The user has the option to modify the semi-automatic segmentations suggested by the software. The EF calculation is updated in real-time with the user's modification of the segmentation. A cardiologist can adjust the annotations and the downstream measurement of LVEF prior to finalization.
Here's a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criterion | Reported Device Performance |
|---|---|
| Root Mean Square Deviation (RMSD) of LVEF vs. reference ground truth EF | Biplane view: ~6.06 |
| A4C view: 6.17 | |
| A2C view: 7.12 | |
| Dice score for A4C segmentation | 0.89 |
| Dice score for A2C segmentation | 0.90 |
| Overall functional performance | Met all endpoints |
| Accuracy of algorithm | Met all endpoints |
| Image video clip selection function performance | Met all endpoints |
Note: The document states "Root Mean Square Deviation below a set threshold" and "Dice score above a set threshold," but the specific thresholds are not explicitly given. The reported performance values are provided instead.
2. Sample Size Used for the Test Set and Data Provenance
- Test Set Sample Size: Not explicitly stated. The document mentions "A retrospective, multicenter study" and "Images and cases used for verification and validation testing were separate and carefully segregated from training datasets," but does not give a specific number for the test set.
- Data Provenance: Retrospective, multicenter study. It included a variety of imaging equipment manufacturers (Philips, GE, Siemens), implying data from different sites. The country of origin is not specified.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Their Qualifications
- Number of Experts: Three cardiologists.
- Qualifications of Experts: Not explicitly stated, but they are identified as "cardiologists," implying medical doctors specializing in cardiology. No experience level (e.g., 10 years) is provided.
4. Adjudication Method for the Test Set
- Adjudication Method: Consensus annotation of three cardiologists. This suggests a consensus-based method, where agreement among the three experts formed the ground truth. It is not explicitly stated as 2+1 or 3+1, but rather a collective agreement by all three.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done
- The document does not mention a Multi-Reader Multi-Case (MRMC) comparative effectiveness study to evaluate human readers' improvement with AI vs. without AI assistance. The study focuses on the device's performance against a ground truth established by experts.
6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) Was Done
- Yes, a standalone performance evaluation was done. The study "evaluated the capability of the Precision machine learning model in calculating LVEF against ground truth." The reported RMSD and Dice scores are measures of the algorithm's performance.
- It's important to note that the device description states the user has the option to modify the semi-automatic segmentations, implying a human-in-the-loop aspect in the clinical use, but the reported performance data appears to be for the standalone algorithm's initial output before human modification.
7. The Type of Ground Truth Used
- Type of Ground Truth: Expert consensus. Specifically, "the consensus annotation of three cardiologists."
8. The Sample Size for the Training Set
- Training Set Sample Size: Not explicitly stated. The document only mentions that "Images and cases used for verification and validation testing were separate and carefully segregated from training datasets."
9. How the Ground Truth for the Training Set Was Established
- Ground Truth Establishment for Training Set: Not explicitly stated. The document only mentions "training datasets" but does not describe the method used to establish their ground truth. This is a common gap in publicly available summary documents.
§ 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).