(131 days)
SmartGrid feature is a software option that provides, upon request by user, a diagnostic radiograph image with a reduction in visible x-ray scatter similar to the effect of an anti-scatter grid.
The SmartGrid software is designed to improve contrast and reduce the appearance of scatter in radiographic images that have been acquired without a physical grid. SmartGrid encapsulates an algorithm for estimating and removing scatter from radiographic images. The SmartGrid feature is accessible through DirectView DR Product application software. Users will be able to select SmartGrid processing before an image is acquired, or to change whether SmartGrid processing is applied to a previously acquired image.
The provided text describes the SmartGrid software, which aims to reduce visible x-ray scatter in diagnostic radiographic images. The acceptance criteria and the study proving the device meets these criteria can be extracted from the "Discussion of Testing" section.
Here's the breakdown of the information requested:
1. A table of acceptance criteria and the reported device performance
The text doesn't explicitly list acceptance criteria in a quantitative table format. Instead, it describes the findings of the study related to image quality. We can infer the acceptance criteria from these reported outcomes.
| Acceptance Criterion (Inferred from Study Results) | Reported Device Performance (SmartGrid) |
|---|---|
| Production of diagnostic quality images. | "SmartGrid processing software produces diagnostic quality images." |
| Image quality compared to non-grid reference images (at all exposure levels). | "At all exposure levels, SmartGrid processing produced images rated as good as or better than the non-grid reference images." |
| Diagnostic quality compared to grid reference acquisitions at lower exposures. | "SmartGrid processing software produces images with statistically equivalent diagnostic quality at lower exposures than the grid reference acquisitions." |
| Image quality after Scatter Factor Estimation and Scatter Correction. | Both SmartGrid and the predicate device "depend on the proper estimation of the scatter-to-primary ratio (SPR) to calculate scatter distribution and perform image enhancement." SmartGrid's performance was found to be substantially equivalent. |
| Effectiveness of noise suppression. | Both SmartGrid and the predicate device "suppress noise." SmartGrid's performance was found to be substantially equivalent. |
2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
- Sample Size for Test Set: The text states that "images of cadaveric specimens and phantoms were acquired and used in the study." It does not specify the exact number of images or cases in the test set.
- Data Provenance:
- Country of Origin: Not specified.
- Retrospective or Prospective: Not explicitly stated, but the acquisition of "images of cadaveric specimens and phantoms" for the study suggests a prospective acquisition for the purpose of the study, rather than leveraging pre-existing clinical images.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)
- Number of Experts: The study was conducted as a "radiologist reader study," implying multiple radiologists. The exact number is not specified.
- Qualifications of Experts: The individuals are referred to as "Radiologists." Specific experience levels or board certifications are not mentioned.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
The text states that "Radiologists reviewed and rated investigational and reference images (both grid and non-grid), for diagnostic quality, using a Radlex subjective diagnostic rating scale." It does not describe any specific adjudication method (e.g., majority vote, consensus after discussion, or a senior radiologist as tie-breaker) if there were disagreements among readers.
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
- MRMC Study: Yes, a "radiologist reader study" was performed, comparing SmartGrid processed images against grid and non-grid reference images. This inherently involves multiple readers (radiologists) reviewing multiple cases (images).
- Effect Size of Human Reader Improvement: The study focused on the diagnostic quality of the images produced by SmartGrid, rather than the improvement of human readers with AI assistance.
- It reports that SmartGrid images were rated "as good as or better than the non-grid reference images."
- And "statistically equivalent diagnostic quality at lower exposures than the grid reference acquisitions."
The study does not quantify how much human readers "improved" in their diagnostic performance when assisted by SmartGrid compared to not using SmartGrid. Instead, it validates the image quality produced by the software.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
No, the primary study described is a "radiologist reader study," which by definition involves human readers evaluating the images processed by the algorithm. There is no mention of a standalone algorithm performance evaluation without human input in the document.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The ground truth was established by expert subjective ratings using a "Radlex subjective diagnostic rating scale." This is a form of expert consensus on image quality rather than an objective clinical truth like pathology or patient outcomes. The text explicitly states that the images were rated for "diagnostic quality."
8. The sample size for the training set
The provided text does not mention the sample size for the training set used to develop the SmartGrid algorithm. The "Discussion of Testing" solely focuses on the performance evaluation study.
9. How the ground truth for the training set was established
The provided text does not mention how the ground truth for the training set was established, as it does not discuss the training phase of the algorithm development.
§ 892.1680 Stationary x-ray system.
(a)
Identification. A stationary x-ray system is a permanently installed diagnostic system intended to generate and control x-rays for examination of various anatomical regions. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II (special controls). A radiographic contrast tray or radiology diagnostic kit intended for use with a stationary x-ray system only is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9.