Search Results
Found 1 results
510(k) Data Aggregation
(27 days)
DIGITAL RADIOGRAPHY CXDI-Pro, D1
The DIGITAL RADIOGRAPHY CXDI-Pro / D1 provides digital image capture for conventional film/screen radiographic examinations. This device is intended to capture, for display, radiographic images of human anatomy, and to replace radiographic film/screen systems in all general purpose diagnostic procedures. This device is not intended for mammography applications.
The DIGITAL RADIOGRAPHY CXDI-Pro, also called the D1, (hereinafter referred to as CXDI-Pro) is a solid-state x-ray imager. The CXDI-Pro is a series of detectors, currently consisting of the CXDI-703C Wireless detector unit, also called the AR-D3543W detector. The detector intercepts x-ray photons, and the scintillator emits visible spectrum photons that illuminate an array of photodetectors that create electrical signals. After the electrical signals are generated, the signals are converted to digital values, and the images will be displayed on monitors. The digital value can be communicated to the operator console via a wired or wireless connection.
The subject of this Special 510(k) submission is a change to the DIGITAL RADIOGRAPHY CXDI-710C Wireless (hereinafter referred to as CXDI-710C) to make the CXDI-Pro. This change will remove the Docking Station, Multi Box, and Status Indicator as optional components and add the X-ray Interface Box and Power Box as optional components. The software has been updated from CXDI Control Software V2.16 to CXDI Control Software V3.10. The case material has been changed from fiberglass to magnesium alloy. Bluetooth function has been added, and Standalone mode has been removed as a photographing mode. Together, these changes make up the CXDI-Pro.
The provided text is a 510(k) Summary for the Canon DIGITAL RADIOGRAPHY CXDI-Pro D1. This document focuses on demonstrating substantial equivalence to a predicate device based on modifications, rather than establishing de novo performance criteria for an AI/ML device. Therefore, it does not explicitly detail acceptance criteria and a study proving a new device meets these criteria in the context of AI/ML performance metrics (e.g., sensitivity, specificity, AUC).
Instead, the document emphasizes that the fundamental scientific technology of the device has not been modified, and the changes are limited to:
- Software update (from V2.16 to V3.10)
- Changes to optional accessories
- Change in case material (fiberglass to magnesium alloy)
- Addition of Bluetooth
- Removal of Standalone mode as a photographing mode
The "performance" section states: "Evaluation of the changes to the CXDI-710C confirmed that the changes did not impact CXDI-Pro conformance with the U.S. Performance Standard for radiographic equipment and with relevant voluntary safety standards for Electrical safety and Electromagnetic Compatibility testing, specifically IEC standards 60601-1, 60601-1-2, 60601-2-54, 60601-1-6, and IEC 60529."
This indicates that the acceptance criteria are primarily related to safety and effectiveness standards for general radiographic equipment, as well as maintaining the existing performance characteristics (like spatial resolution) of the previous device. The study proving the device meets these criteria involved verification/validation activities demonstrating continued compliance with these standards despite the modifications.
Given this context, I will extract and infer information relevant to your request, acknowledging that it's not a study designed to prove AI performance in the way you might expect for a diagnostic algorithm.
Here's a breakdown based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criterion (Type) | Reported Device Performance |
|---|---|
| Safety and Electrical Standards | Conformance with U.S. Performance Standard for radiographic equipment and relevant voluntary safety standards for Electrical safety and Electromagnetic Compatibility testing (specifically IEC standards 60601-1, 60601-1-2, 60601-2-54, 60601-1-6, and IEC 60529). |
| Functional Equivalence | The changes (software update, accessories, case material, Bluetooth, removal of Standalone mode) are demonstrated to not negatively impact the fundamental scientific technology or performance characteristics compared to the predicate device. |
| Spatial Resolution | Proposed Device: 35% [MTF@2lp/mm] |
| Predicate Device: 35% [MTF@2lp/mm] | |
| This indicates that the spatial resolution performance metric was maintained as identical to the predicate, serving as an implicit acceptance criterion for this characteristic. | |
| Indications for Use (Effectiveness) | The "Indication for Use statement is identical to the predicate device." The device is intended to "capture, for display, radiographic images of human anatomy, and to replace radiographic film/screen systems in all general purpose diagnostic procedures." The modifications did "not change as a result of the modification(s)." This implies the continued ability to effectively perform its stated intended use. |
| Ingress Protection (IP) Level | Proposed Device: IP55 |
| Predicate Device: IPX7 | |
| Reference Devices: IP54 | |
| While modified from the predicate, the new IP55 rating would have been an acceptance criterion that the modified device met, indicating a specific level of protection against solids and liquids. | |
| Pixel Pitch | Proposed Device: 140μm |
| Predicate Device: 125μm | |
| Reference Devices: 125μm | |
| While modified from the predicate, the device demonstrably has a pixel pitch of 140μm, implying this new specification was acceptable and likely evaluated not to degrade image quality below an acceptable threshold for its indicated use. |
2. Sample Size Used for the Test Set and Data Provenance
The document does not detail a "test set" in the context of clinical images for diagnostic AI/ML performance evaluation. The "study" here is a series of "verification/validation activities" focused on engineering, software, and hardware changes. Therefore, there is no mention of a specific sample size of patient data or its provenance (country of origin, retrospective/prospective). The evidence presented focuses on compliance testing with engineering and safety standards.
3. Number of Experts Used to Establish Ground Truth and Qualifications
Not applicable for this type of submission. Ground truth, in the context of medical image interpretation by experts, is not relevant to this submission, which focuses on hardware and software modifications and compliance with engineering standards.
4. Adjudication Method for the Test Set
Not applicable. There is no mention of a test set requiring adjudication in the context of diagnostic interpretation.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No. This submission is for a digital radiography system, not an AI-powered diagnostic assistant. Therefore, no MRMC study comparing human readers with and without AI assistance was performed or described.
6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study
No. While the document mentions "Standalone mode" was removed as a photographing mode for the device itself, this does not refer to an AI algorithm's standalone diagnostic performance. The device is a digital X-ray imager, which functions as a component in the diagnostic workflow, not an autonomous diagnostic algorithm.
7. Type of Ground Truth Used
The "ground truth" here is the adherence to established engineering, safety, and performance standards (e.g., IEC 60601 series, MTF measurements, IP ratings). It's not clinical ground truth derived from expert consensus, pathology, or outcomes data.
8. Sample Size for the Training Set
Not applicable. This document describes hardware and software changes to an established medical device, not the training of a new AI/ML algorithm.
9. How the Ground Truth for the Training Set Was Established
Not applicable, as no AI/ML training set is discussed. The "ground truth" for the device's functional integrity is presumably established through rigorous engineering testing, quality control, and adherence to manufacturing specifications, rather than through labeled datasets for machine learning.
Ask a specific question about this device
Page 1 of 1