(22 days)
The KODAK DirectView CR 800 System and the KODAK Direct View CR 900 System are compact laser scanners capable of reading the latent image formed on a storage phosphor imaging plate and producing a digital image for projection radiography applications.
The KODAK DirectView CR 800 is a single cassette computed radiography (CR) system while the KODAK DirectView CR 900 is an autoloading CR system. Each CR system contains a storage phosphor reader, CR cassette, HeNe laser, QA workstation with image processing software, and a DICOM network connection. These CR systems are capable of reading the latent image formed on a storage phosphor imaging plate in a CR cassette and producing a digital image. The system scans the phosphor plate, communicates the image data to a host (e.g. soft copy review workstation or laser printer) and erases the residual image on the plate. An internal imaging plate eraser removes the residual image. The plate is then returned to the cassette, ready for the next exposure.
The provided text describes a software modification to the KODAK DirectView CR 800 and CR 900 Systems to reduce grid artifacts in radiographic images. The study conducted aimed to demonstrate that this software modification, specifically the grid detection and suppression algorithm, does not negatively impact diagnostic image quality.
Here's a breakdown of the acceptance criteria and study details based on the provided information:
Acceptance Criteria and Reported Device Performance
| Acceptance Criteria Category | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Diagnostic Image Quality | Radiologists rate images as "acceptable for interpretation" or better (defined as 7 or higher on a 1-9 scale). | Average scores ranged from 7.2 (with no grid suppression) to 7.7 (with maximum grid suppression applied). The differences were deemed "not clinically significant." |
| Grid Artifact Reduction | The algorithm should be effective at reducing grid-induced artifacts. | The algorithm was effective at reducing grid induced artifacts for a wide variety of grids. |
Study Details
-
Sample size used for the test set and the data provenance:
- Sample Size: Not explicitly stated how many images/cases were included in the evaluation. It mentions "a variety of radiographic exams (e.g. chest, abdomen, spine, shoulder, pelvis)"
- Data Provenance: The study used "hard copy CR images." The country of origin and whether the data was retrospective or prospective are not specified.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Number of Experts: Ten radiologists.
- Qualifications: "Radiologists" is the only qualification given; no specific experience level is mentioned.
-
Adjudication method for the test set:
- The radiologists rated the exams individually, and average scores were presented. There is no mention of a formal adjudication method (e.g., 2+1, 3+1).
-
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:
- This was not an MRMC comparative effectiveness study in the typical sense of measuring human reader improvement with AI assistance. Instead, it was an evaluation of the impact of grid suppression on diagnostic image quality. It compared images with and without the grid suppression algorithm applied, and the outcome measured was the radiologists' acceptability rating of the images.
- Effect Size: The average scores shifted from 7.2 (no suppression) to 7.7 (maximum suppression). This suggests a slight improvement in perceived quality with suppression, but the text states these differences were "not clinically significant" in relation to maintaining diagnostic acceptability.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Yes, an initial analytical evaluation of the algorithm's performance was done: "The analytical characteristics of the grid detection and suppression algorithm were evaluated for a wide variety of grids commonly in use in the medical community. The algorithm was effective at reducing grid induced artifacts." This indicates a standalone assessment of the algorithm's technical function prior to human evaluation.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- Expert Opinion/Rating: The ground truth for image quality was established by the subjective ratings of ten radiologists regarding the "acceptability for interpretation" of the images. There is no mention of a true diagnosis based on pathology or outcomes data.
-
The sample size for the training set:
- Not provided. The text only mentions the evaluation of the algorithm, not its training data.
-
How the ground truth for the training set was established:
- Not provided. As training set details are absent, so is the method for establishing its ground truth.
§ 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.