(131 days)
ClearRead CT CAC is image processing software designed to aid physicians in assessing coronary artery calcification on non-gated, non-contrast, standard, or low-dose chest CT scans of adult patients 30 years of age or older. The primary output of ClearRead CT CAC is the Agatston score. ClearRead CT CAC locates calcified coronary lesions and assigns them to one of several coronary arteries: the right coronary artery (RCA), the left main artery (LM), the left anterior descending artery (LAD), and the left circumflex artery (LCX). The software calculates the Agatston score as the sum for each artery while also providing the total across all coronary arteries. The total Agatston score is calibrated to 3.0mm slice thickness and is used to derive a CAC category and the arterial age. The provided segmentations are for illustrative purposes only and are not intended for diagnostic use.
ClearRead CT CAC provides adjunctive information and is not intended to be used without clinical expert review.
ClearRead CT CAC is a computer aided quantification system intended to aid in the assessment of coronary artery calcification. The system receives chest Computed Tomography (CT) studies as input, in DICOM® format, and generates output in DICOM (or other) format.
Here's a detailed breakdown of the acceptance criteria and the study proving the device meets them, based on the provided document:
1. Table of Acceptance Criteria and Reported Device Performance
| Metric / Endpoint | Acceptance Criteria (Internal Investigation/Literature) | Reported Device Performance (Clinical Study) |
|---|---|---|
| Agatston Score Category Agreement | Weighted Kappa > 0.85 (considered excellent) | 0.959 (95% CI: 0.943-0.975) |
| CAC Negative Agreement (NPA) | Not explicitly stated as acceptance criteria, but reported. | 99.1% (95% CI: 98.0%, 100.0%) for Category 0 vs. Not Category 0 |
| CAC Positive Agreement (PPA) | Not explicitly stated as acceptance criteria, but reported. | PPA for Category 0: 88.1% (95% CI: 82.5%, 93.2%) |
| PPA for Category 1-99: 92.8% (95% CI: 87.7%, 96.9%) | ||
| PPA for Category 100-399: 88.1% (95% CI: 80.5%, 94.4%) | ||
| PPA for Category ≥ 400: 98.6% | ||
| Overall Continuous Agatston Score Agreement | Bland-Altman analysis showing strong agreement (qualitative, based on Figure 2 from non-clinical). No specific numerical acceptance criteria provided for LoA. | Limits of Agreement (LoA) of -12.20 ± 437.84 |
| Coronary Artery Calcification Segmentation Accuracy | Not explicitly stated as a separate acceptance criterion with a specific numerical threshold. | Average Dice score: 0.892 (0.871, 0.911) (Device vs. GT) |
2. Sample Size for Test Set and Data Provenance
- Non-clinical testing: Approximately 400 cases. Data provenance is not specified, but it's referred to as "internal investigation."
- Clinical testing: 491 cases (143 cases with no CAC, 348 with CAC). The data included samples from four CT manufacturers across five clinical sites. The specific country of origin is not mentioned, nor is it explicitly stated whether the data was retrospective or prospective, though the mention of "cases collected and held out from development" suggests retrospective.
3. Number of Experts and Qualifications for Ground Truth
- Clinical testing: Three radiologists established the consensus ground truth review. Their specific qualifications (e.g., years of experience, subspecialty) are not detailed in the provided text.
4. Adjudication Method for the Test Set
- Clinical testing: A "consensus ground truth review by three radiologists" was used. This implies that the three radiologists collectively determined the ground truth for each case. The specific mechanism (e.g., majority vote, discussion until agreement) is not specified, but it suggests a form of 3-reader consensus.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- No MRMC comparative effectiveness study was done. The study performed was a standalone model performance assessment comparing the algorithm's output to ground truth measurements. The document does not describe human readers improving with AI vs. without AI assistance.
6. Standalone Performance Assessment
- Yes, a standalone (algorithm only without human-in-the-loop performance) study was done. The document explicitly states: "This study was a standalone model performance assessment of ClearRead CT CAC."
7. Type of Ground Truth Used
- Clinical testing: "Consensus ground truth review by three radiologists." This is a form of expert consensus.
8. Sample Size for Training Set
- The sample size for the training set is not provided in the given document. The document mentions that approximately 400 cases were "held out from development of any kind to estimate the test performance," implying these were for the test set, not training.
9. How the Ground Truth for the Training Set was Established
- The document does not provide information on how the ground truth for the training set was established, as the details focus on the "held out" test set for performance estimation. For the test set, "Regions associated with coronary artery calcification were manually outlined, including the particular artery, from which the ground truth Agatston score was obtained."
§ 892.1750 Computed tomography x-ray system.
(a)
Identification. A computed tomography x-ray system is a diagnostic x-ray system intended to produce cross-sectional images of the body by computer reconstruction of x-ray transmission data from the same axial plane taken at different angles. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II.