(98 days)
The qER-Quant device is intended for automatic labeling, visualization of segmentable brain structures from a set of Non-Contrast head CT (NCCT) images. The software is intended to automate the current manual process of identifying, labeling and quantifying the volume of segmentable brain structures identified on NCCT images. qER-Quant provides volumes from NCCT images acquired at a single time point and provides a table with comparative analysis for two or more images that were acquired on the same image acquisition protocol for the same individual at multiple time points.
The qER-Quant software is indicated for use in the following structures: Intracranial Hyperdensities, Lateral Ventricles and Midline Shift.
qER-Quant is a standalone software device that processes non-contrast head CT scans to outline and quantify the structures described in the intended use. The qER-Quant software interacts with the user's picture archiving and communication system (PACS) to receive scans and returns the results to the same destination.
The analysis module of the qER-Quant software contains of a set of pre-trained convolutional neural networks (CNNs), that form the core processing component shown in Figure 1. This core processing component is coupled with a pre-processing module to prepare input digital imaging and communications in medicine (DICOMs) for processing by the CNNs and a post-processing module to convert the output into visual and tabular output for users.
Here's a breakdown of the acceptance criteria and study details for the qER-Quant device, based on the provided text:
qER-Quant Device Performance Study Details
1. Acceptance Criteria and Reported Device Performance
The acceptance criteria were defined based on the accuracy of the qER-Quant system when compared against manually labeled ground truth. The reported device performance met these pre-set criteria.
| Metric | Acceptance Criteria (Implied / Context) | Reported Device Performance (Mean ± SD / Mean (95% CI) / Median (10th-90th Percentile)) |
|---|---|---|
| Intracranial Hyperdensity | ||
| Absolute Error (Volume) | Exceeds preset acceptance criteria | 6.56 (7.33) ml (Mean ± SD) |
| 3.98 (0.52 - 17.35) ml (Median (10th - 90th percentile)) | ||
| Dice Score (Segmentation Accuracy) | Exceeds preset acceptance criteria | 0.75 (0.72 - 0.78) (Mean (95% CI)) |
| Midline Shift | ||
| Absolute Error (Shift) | Exceeds preset acceptance criteria | 1.37 (1.23) mm (Mean ± SD) |
| 1.15 (0.23 - 2.59) mm (Median (10th - 90th percentile)) | ||
| Dice Score (Segmentation Accuracy) | Not Applicable | Not applicable |
| Left Lateral Ventricle | ||
| Absolute Error (Volume) | Exceeds preset acceptance criteria | 2.09 (1.88) ml (Mean ± SD) |
| 1.60 (0.29 - 4.24) ml (Median (10th - 90th percentile)) | ||
| Dice Score (Segmentation Accuracy) | Exceeds preset acceptance criteria | 0.79 (0.78 - 0.81) (Mean (95% CI)) |
| Right Lateral Ventricle | ||
| Absolute Error (Volume) | Exceeds preset acceptance criteria | 2.18 (1.72) ml (Mean ± SD) |
| 1.88 (0.40 - 4.53) ml (Median (10th - 90th percentile)) | ||
| Dice Score (Segmentation Accuracy) | Exceeds preset acceptance criteria | 0.75 (0.73 - 0.77) (Mean (95% CI)) |
2. Sample Size and Data Provenance
- Test Set Sample Sizes:
- Intracranial Hyperdensity: 183 scans
- Midline Shift: 188 scans
- Left Lateral Ventricle: 210 scans
- Right Lateral Ventricle: 210 scans
- Reproducibility testing was done on 20% of these CT scans.
- Data Provenance: The document does not explicitly state the country of origin or whether the data was retrospective or prospective. It uses "a set of head CT scans."
3. Number of Experts and Qualifications for Ground Truth Establishment
- Number of Experts: The document states "experts" (plural) were used but does not specify the exact number.
- Qualifications of Experts: Not specified beyond being "experts" in the context of manually labeling CT scans.
4. Adjudication Method for the Test Set
The document does not explicitly state an adjudication method (e.g., 2+1, 3+1). It only mentions that the ground truth was established by "manually labeled by experts."
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- No, an MRMC comparative effectiveness study was not reported. The performance testing was a "standalone" evaluation of the device's accuracy against expert-generated ground truth.
6. Standalone Performance (Algorithm Only)
- Yes, a standalone performance study was conducted. The document states: "Qure.ai performed standalone consisted of a set of head CT scans with the outlines of the target structures manually labeled by experts." The results detailed in Table 2 are of this standalone performance.
7. Type of Ground Truth Used
- The ground truth used was expert consensus / manual labeling. The document clearly states: "manually labeled by experts."
8. Sample Size for the Training Set
- The document does not provide the sample size for the training set. It only describes the architecture of the analysis module as "a set of pre-trained convolutional neural networks (CNNs)."
9. How Ground Truth for the Training Set Was Established
- The document does not explicitly state how the ground truth for the training set was established. It describes the CNNs as "pre-trained," which implies a training phase using labeled data, but the method of ground truth establishment for that specific data is not detailed.
§ 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).