(196 days)
EchoGo Core is intended to be used for quantification and reporting of cardiovascular function to support physician diagnosis. EchoGo Core is indicated for use in adult populations.
EchoGo Core is a standalone software application that automatically measures standard cardiac parameters including Ejection Fraction (EF), Global Longitudinal Strain (GLS), and Left Ventricular (LV) volume. EchoGo Core provides a service to calculate those values and provide a report back to the clinician. EchoGo Core analyzes echocardiogram cardiac parameters. An echocardiogram is performed using standard echocardiograms protocols. The anonymized echocardiogram dataset is transferred to Ultromics and once received the images are processed through the application's workflow. Once the technical QC has been performed, the automated contour detection of the endocardium of the LV is reviewed and approved by trained operators. Identifying and outlining the LV is standard practice in echocardiography and present in many other cleared devices, including the predicate device TomTec Arena TTA2 (K150122). In the proposed device, an auto-contouring algorithm places points around the LV that sufficiently capture the LV shape. These contours are used in calculations for geometric parameters. As with the TomTec Arena TTA2, the following parameters are calculated for inclusion in the report: EF: calculated from Simpson's Biplane Method (SBM) LV Volumes. GLSAVG: standard calculations, averaged from 2 chamber and 4 chamber views. LV volume: SBM calculated left ventricular volumes. A worksheet is automatically generated from the calculated parameters which is returned to the interpreting clinician. This report is intended as an additional input to standard diagnostic pathways and is only to be used by qualified clinicians.
Here's a breakdown of the acceptance criteria and study details for the EchoGo Core device, as extracted from the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Parameter | Acceptance Criteria (RMSE Threshold relative to TomTec Arena TTA2) | Reported Device Performance (RMSE) |
|---|---|---|
| Ejection Fraction (EF) | Below a set threshold | 5.02% |
| Global Longitudinal Strain (GLS) | Below a set threshold | 2.89% |
| End-Diastolic Volume | Below a set threshold | 8.0 ml |
| End-Systolic Volume | Below a set threshold | 11.1 ml |
(Note: The exact numerical thresholds for "Below a set threshold" are not explicitly stated for all parameters, but the reported performance met these unquantified thresholds.)
2. Sample Size and Data Provenance for Test Set
- Sample Size: 378 previously acquired studies.
- Data Provenance: The text does not explicitly state the country of origin. It indicates the data was "previously acquired studies." The study was a "retrospective, non-interventional validation study."
3. Number of Experts and Qualifications for Ground Truth
The text states that the reference values, against which EchoGo Core was compared, were "generated using TomTec Arena TTA2 by a range of users (echocardiographers and cardiologists)." It also mentions that "EchoGo Core volume, ejection fraction and global longitudinal strain measurements were produced by a range of users (echocardiographers and cardiologists)."
- Number of Experts: "a range of users" (not a specific number provided).
- Qualifications of Experts: Echocardiographers and cardiologists.
- Note: The ground truth itself was not established by these experts directly, but rather by the predicate device (TomTec Arena TTA2) operated by these experts.
4. Adjudication Method for the Test Set
The text does not describe an explicit adjudication method for the test set. The primary comparison was between the EchoGo Core's output and the values generated by the predicate device (TomTec Arena TTA2). There's no mention of multiple experts independently evaluating and then adjudicating discrepancies for the ground truth establishment.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
There is no mention of a formal MRMC comparative effectiveness study that assesses how much human readers improve with AI vs. without AI assistance. The study described focuses on the standalone performance of the EchoGo Core device compared to a predicate device.
However, the document does make a comparison related to user interaction with contours:
- "Inter and intra-operator variability (0% bias and RMSE) was eliminated" for EchoGo Core.
- "inter and intra-operator variability remained for the predicate, TomTec Arena TTA2 (5.3-12.0% and 2.5-4.7% RMSE for EF and GLS, respectively)."
- "manual contour editing was required in up to 100% of cases processed through the predicate, while contour rejection was required in up to 29% of cases processed through EchoGo Core."
This indicates an implicit benefit in eliminating operator variability, but it's not quantified in terms of improved human reader performance with AI assistance in a diagnostic task.
6. Standalone Performance Study
Yes, a standalone study was performed. The entire performance testing section describes the EchoGo Core device calculating parameters independently and then comparing these calculated values to those produced by the predicate device (TomTec Arena TTA2). The performance metrics (RMSE) directly refer to the algorithm's output.
7. Type of Ground Truth Used
The ground truth for the performance study was the measurements derived from the predicate device, TomTec Arena TTA2, operated by echocardiographers and cardiologists. This can be considered a form of "expert consensus/device-derived reference," where the extensively validated predicate device, used by experts, served as the reference standard.
8. Sample Size for the Training Set
The document explicitly states: "Test datasets were strictly segregated from algorithm training datasets." However, it does not provide the sample size for the training set.
9. How the Ground Truth for the Training Set Was Established
The document states: "Test datasets were strictly segregated from algorithm training datasets." While it confirms segregation, it does not provide information on how the ground truth for the training set was established.
§ 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).