(50 days)
Agfa's DX-D Imaging Package with XD Detectors is indicated for use in general projection radiographic applications to capture for display diagnostic quality radiographic images of the human anatomy. The DX-D Imaging Package with XD Detectors may be used wherever conventional screen-film systems may be used.
Agfa's DX-D Imaging Package with XD Detectors is not indicated for use in mammography.
The DX-D Imaging Package, previously cleared under K142184, is a solid state x-ray system, a direct radiography (DR) system (product code MQB) intended to capture general radiographic images of the human body. It is a combination of Agfa's NX workstation with MUSICA TM image processing and one or more flat-panel detectors of the scintillator-photodetector type (Cesium Iodide - CsI or Gadolinium Oxysulfide - GOS). It is capable of replacing other direct radiography, including computed radiography systems with conventional or phosphorous film cassettes.
This submission is to add the XD Detectors (XD 10/10+, XD 14/14+ and XD 17/17+) Flat Panel Detectors to Agfa's DX-D Imaging Package portfolio. Agfa's XD Detectors are currently marketed by Vieworks Co. Ltd. as FXRD-4343VAW/VAW Plus. FXRD-3643VAW/VAW Plus. and FXRD-2530VAW/VAW Plus which is the predicate for this submission (K200418).
The optional image processing allows users to conveniently select image processing settings for different patient sizes and examinations. The image processing algorithms in the new device are identical to those previously cleared in the DX-D Imaging Package - DX-D 40 (K142184reference) device and other devices in Agfa's radiography portfolio today. The addition of the offline workflow is identical to the Vivix-S VW (K200418) predicate device.
Principles of operation and technological characteristics of the new, predicate and reference devices are the same. There are no changes to the intended use/indications of the device. The new device is physically and electronically similar to the predicate device (K200418) which includes the addition of an offline workflow capable of storing up to 200 images on the flat-panel detector for later viewing. It uses the same NX workstation with MUSICA™ image processing as the reference device (K142184) and the same flat panel detectors of the scintillator-photodetector type (Cesium Iodide - Csl or Gadolinium Oxysulfide - GOS) to capture and digitize the images as the predicate device (K200418). Laboratory data and image quality evaluations conducted with internal specialists confirm that performance is equivalent to the predicate. Differences in devices do not alter the intended diagnostic effect nor do they impact the safety and effectiveness of the device.
The provided text describes a 510(k) premarket notification for Agfa’s DX-D Imaging Package with XD Detectors. The submission focuses on demonstrating substantial equivalence to a predicate device (Vieworks Vivix-S VW, K200418) and a reference device (Agfa DX-D Imaging Package -DX-D 40, K142184) rather than establishing new performance criteria or conducting extensive clinical trials of the AI component.
Key takeaway: This submission primarily focuses on the physical components of the imaging system (detectors, workstation, image processing) and their equivalence to existing cleared devices. It does not contain details about specific acceptance criteria or performance metrics for an AI algorithm in the way one might expect for a novel AI-driven diagnostic device. The "AI" mentioned is related to existing "MUSICA image processing" which is identical to previously cleared versions and is referred to as "image processing algorithms", rather than a new AI model with specific diagnostic performance targets.
Therefore, many of the requested details, especially those related to AI-specific performance criteria, ground truth establishment for a test set, and multi-reader studies, are not explicitly present in the provided document. The performance data presented refers to the physical detector characteristics (Spatial Resolution, DQE, MTF).
Here's an attempt to answer the questions based on the provided text, highlighting where information is absent:
1. A table of acceptance criteria and the reported device performance
The document does not define explicit "acceptance criteria" in terms of clinical performance metrics for an AI algorithm. Instead, it demonstrates performance equivalence of the new detector models to existing ones.
| Performance Characteristic | DX-D 40 Flat-Panel Detector (K142184) (Reference) | XD 10/10+ Wireless Detector (New Device) | XD 14/14+ Wireless Detector (New Device) | XD 17/17+ Wireless Detector (New Device) |
|---|---|---|---|---|
| Spatial Resolution | 3.5 lp/mm (for 6007/100 & 6007/200) | 4.0 lp/mm | 3.5 lp/mm | 3.5 lp/mm |
| DQE @ 1 lp/mm | CsI: 0.494, GOS: 0.259 | XD: 0.500, XD+: 0.587 | XD: 0.425, XD+: 0.587 | XD: 0.412, XD+: 0.587 |
| DQE @ 2 lp/mm | CsI: 0.379, GOS: 0.157 | XD: 0.401, XD+: 0.445 | XD: 0.321, XD+: 0.399 | XD: 0.345, XD+: 0.407 |
| DQE @ 3 lp/mm | CsI: 0.215, GOS: 0.061 | XD: 0.288, XD+: 0.316 | XD: 0.206, XD+: 0.257 | XD: 0.220, XD+: 0.280 |
| MTF @ 1 lp/mm | CsI: 0.685, GOS: 0.589 | XD: 0.729, XD+: 0.650 | XD: 0.751, XD+: 0.635 | XD: 0.726, XD+: 0.656 |
| MTF @ 2 lp/mm | CsI: 0.386, GOS: 0.266 | XD: 0.424, XD+: 0.315 | XD: 0.446, XD+: 0.302 | XD: 0.428, XD+: 0.311 |
| MTF @ 3 lp/mm | CsI: 0.209, GOS: 0.115 | XD: 0.236, XD+: 0.157 | XD: 0.247, XD+: 0.152 | XD: 0.231, XD+: 0.161 |
The "acceptance criteria" appear to be met by demonstrating that the new detectors (XD series) have comparable or superior technical performance characteristics (Spatial Resolution, DQE, MTF) to the existing reference devices, and that the image processing ("MUSICA™ image processing") is identical to previously cleared devices. The document states: "The results of these tests fell within the acceptance criteria for the DX-D Imaging Package with XD Detectors." However, the quantitative thresholds for these "acceptance criteria" are not specified beyond the presented performance values.
2. Sample size used for the test set and the data provenance:
- Test Set (for image quality evaluation): "anthropomorphic adult and pediatric phantoms".
- Data Provenance: Not explicitly stated (likely internal laboratory data, given "internal specialists"). This was bench testing, not clinical data from patients.
- Software Test Iterations (NX 23): Two software iterations were tested.
- Performance Functionality Evaluation: Not a sample size of data, but related to the number of experts (see below).
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Image Quality Evaluations: "qualified internal experts" (radiographers) evaluated the overall image quality using phantoms. The exact number of radiographers is not specified.
- Performance Functionality Evaluations: "four qualified experts". Their specific qualifications (e.g., radiologist, years of experience) are not detailed beyond "qualified experts".
- Ground Truth: For the phantom studies, the ground truth is inherently defined by the known properties of the phantoms (e.g., specific structures, resolution targets).
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set:
Not explicitly stated. The evaluations seem to be internal assessments for comparison rather than a formal human reader study with adjudication for a clinical diagnostic task.
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 MRMC study was performed. The document explicitly states: "No clinical trials were performed in the device. No animal or clinical studies were performed in the development of the new device. No patient treatment was provided or withheld."
- The "image processing algorithms in the new device are identical to those previously cleared." This suggests that the "AI" (MUSICA image processing) is not a new or modified component requiring a new MRMC study to demonstrate clinical impact or improvement for human readers.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Performance data (Spatial Resolution, DQE, MTF) are presented for the detectors themselves, which can be considered "standalone" technical performance metrics of the hardware.
- For the MUSICA image processing software, its "standalone" performance is implied to be equivalent to its previously cleared versions since it is "identical." No new quantitative standalone performance metrics for the algorithm itself are provided in this submission beyond this statement of identity.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- For the technical performance data (Spatial Resolution, DQE, MTF) related to the detectors: The ground truth is based on standard metrology and engineering principles for measuring detector performance using physical test objects (e.g., bar patterns, uniform fields) and laboratory equipment according to established standards.
- For the image quality evaluations: The ground truth was based on "anthropomorphic adult and pediatric phantoms," meaning the content and structures within the phantoms served as the reference.
- For the software testing: Ground truth for software verification and validation is against pre-defined requirements and design specifications, with "deviations or variances...documented in a defect database and addressed."
8. The sample size for the training set:
The document mentions "MUSICA™ image processing" which contains algorithms. However, this submission states these algorithms are "identical to those previously cleared" and does not describe the development or training of any new AI models. Therefore, information about a training set for a novel AI algorithm is not applicable to this 510(k) submission, as it is leveraging previously cleared technology.
9. How the ground truth for the training set was established:
As no new AI model training is described in this submission, information on how a training set's ground truth was established is not applicable.
§ 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.