Search Results
Found 1 results
510(k) Data Aggregation
(52 days)
The Cardiac CT Function Software Application is indicated to be used with multi-phase, multi-slice cardiovascular CT angiography images to assist qualified medical professionals in assessing and evaluating cardiac function. CT Function includes manual and semi-automatic heart segmentation of 2 chambers (LV and RV) and calculation of cardiac function metrics including end diastolic volume, end systolic volume, stroke volume, ejection fraction, cardiac output, cardiac index, and LV myocardial mass.
Circle's Cardiac CT Function Software Application ("CT Function Module" or "CT Function", for short) is a software device that enables the analysis of cardiac images acquired using computed tomography (CT) scanners. It is designed to support physicians in the visualization, and analysis of heart function through the calculation of parameters such as volume and mass. The device is intended to be used as an aid to the existing standard of care and does not replace existing software applications that physicians use. The CT Function Module does not interface directly with any data collection equipment, and its functionality is dependent on the type of vendor acquisition equipment. The analysis results are available on-screen or can be saved for future review.
CT Function consists of multiplanar reconstruction (MPR) views and 3D rendering of the original CT data. The module displays three MPR views that the user can freely adjust to any position and orientation. Lines and regions of interest (ROIs) can be manually drawn on these MPR images for quantitative measurements.
The CT Function Module implements an Artificial Intelligence / Machine Learning (AI/ML) algorithm to detect and segment heart structures and post-processing methods to convert the heart segments to editable surfaces. All surfaces generated by the system are editable and users are advised to verify and correct any errors; if desired, users can also perform seqmentation of the cardiac structures manually using surface contour generation tools.
Here's a breakdown of the acceptance criteria and study details for the Cardiac CT Function Software Application, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Metric | Region | Acceptance Criteria (Implied from "less than X%" or "above Y%") | Reported Device Performance |
|---|---|---|---|
| Mean Absolute Error (MAE) for Volume Prediction | LV Cavity | < 10% | < 10% |
| Dice Coefficient | LV Cavity | > 86% | > 86% |
| 3D Hausdorff Distance (HD) | LV Cavity | < 9.5 mm | < 9.5 mm |
| Ejection Fraction (EF) Bias | LV Cavity | Not explicitly stated as acceptance criteria, but reported. | 1.3% (95% CI: [-12, 14]) |
| Mean Absolute Error (MAE) for Volume Prediction | RV Cavity | < 18% | < 18% |
| Dice Coefficient | RV Cavity | > 85% | > 85% |
| 3D Hausdorff Distance (HD) | RV Cavity | < 18 mm | < 18 mm |
| Ejection Fraction (EF) Bias | RV Cavity | Not explicitly stated as acceptance criteria, but reported. | -5.5% (95% CI: [-15, 4.4]) |
| Mean Absolute Error (MAE) for Volume Prediction | LV Myocardium | < 17% | < 17% |
| Dice Coefficient | LV Myocardium | > 82% | > 82% |
| 3D Hausdorff Distance (HD) | LV Myocardium | < 15 mm | < 15 mm |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size: Not explicitly stated. The document mentions "All data used for validation were not used during the development of the ML algorithms" and "Image information for all samples was anonymized... The validation data was sourced from 9 different sites." It does not provide a specific number of cases or patients in the test set.
- Data Provenance: Retrospective, sourced from 9 different sites, with 90% of the data sampled from US sources. The data included images from major CT imaging device vendors.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
- Number of Experts: Three expert readers.
- Qualifications of Experts: Not explicitly stated. The document refers to them simply as "expert readers."
4. Adjudication Method for the Test Set
- Adjudication Method: Not explicitly stated. The text mentions "Compared to a reference standard established from three expert readers," implying a consensus or individual reference for each expert, but the method of combining or adjudicating their readings to form the final ground truth is not described.
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
- No, an MRMC comparative effectiveness study involving human readers with and without AI assistance was not done. The study described is a standalone performance validation of the AI model against an expert-established ground truth.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
- Yes, a standalone performance study of the algorithm was done. The reported metrics (MAE, Dice, HD, EF bias) directly assess the ML model's performance in segmenting and calculating cardiac function metrics against a reference standard.
7. The Type of Ground Truth Used
- Expert Consensus (or reference standard): The ground truth was "a reference standard established from three expert readers."
8. The Sample Size for the Training Set
- Sample Size: Not explicitly stated. The document mentions the ML algorithms "have been trained and tested on images acquired from major vendors of CT imaging devices," but does not provide a specific number for the training set.
9. How the Ground Truth for the Training Set was Established
- Ground Truth Establishment: Not explicitly stated. While the document mentions the ML algorithms were trained, it does not describe the specific process or type of ground truth used during the training phase. It only clarifies that the validation data was not used during development.
Ask a specific question about this device
Page 1 of 1