(75 days)
The Raven is a free standing device used to print diagnostic images for viewing on a standard view box. It may be used in any situation in which a hard copy of an image generated by a medical imaging device is required or desirable.
The device accepts electrical image signals and produces hard copy images. The image signal source may be digital formatted image data from image readers or unformatted image data from other imaging modalities (e.g. CT, MRI). The image signal source may be analog or digital. The Helios Laser Imager uses the information in the image signals digitally record diagnostic images and patient data on a proprietary product specific medical imaging media. The Helios does not use conventional light-sensitive silver halide photographic media, requires no dark room, film processor, processing chemicals, water, drainage, or dryer ventilation. It produces no chemical waste, and requires no space for chemical storage.
The provided text is a 510(k) premarket notification for a medical device called the "Helios Laser Imager." This document focuses on demonstrating substantial equivalence to a previously cleared device, not on proving device performance against specific acceptance criteria for an AI algorithm. Therefore, much of the requested information regarding AI device performance, sample sizes, ground truth establishment, expert adjudication, and MRMC studies is not available in this document.
Here's an analysis based on the information provided, highlighting the differences in context:
1. A table of acceptance criteria and the reported device performance:
The document describes modifications to an existing device (Helios Laser printer, K912073) and asserts that these modifications do not change the technology or safety of the Helios printer and that the new version ("Helios Laser Imager 1417") is without question substantially equivalent to its predecessor and is safe and effective for its intended use.
The "performance" described relates to the system's ability to accept image signals and produce hard copy images, and improvements in throughput and cost efficiency. However, there are no specific, quantifiable acceptance criteria or reported device performance metrics in the way one would expect for an AI algorithm's diagnostic accuracy (e.g., sensitivity, specificity, AUC).
Instead, the closest to "criteria" are general statements about safety and equivalence.
| Acceptance Criteria (Implied) | Reported Device Performance (Implied) |
|---|---|
| Maintain Safety | "The results of the hazard analysis, combined with the appropriate preventive measures taken indicate the device is of minor level of concern..." "These modifications do not change the technology or safety of the Helios printer." |
| Maintain Effectiveness for Intended Use | "The Helios 'C' is without question substantially equivalent to its predecessor and is safe and effective for its intended use." "The device does not impact the quality or status of the original acquired image data." |
| Compatibility with Image Signals and Production of Hard Copy | "The device accepts electrical image signals and produces hard copy images." "The Helios Laser Imager uses the information in the image signals digitally record diagnostic images and patient data on a proprietary product specific medical imaging media." (No specific metrics on image quality are provided, but the statement implies successful image creation.) |
| Increased Throughput | "First is the increased number of lasers to allow the simultaneous printing of 2 lines of image data; thus increasing the throughput." (No specific numerical increase in throughput is provided, but the increase is stated.) |
| Improved Cost Efficiency and Reliability | "The mechanical sheet feeder has been improved for increased cost efficiency and reliability." (No specific metrics on cost reduction or reliability improvement are provided, but the improvement is stated.) |
| Compliance with Regulatory Standards | "The device complies with the relevant international and national Safety Standards. It has been manufactured in compliance with ISO9000 and the Quality System Regulation [21 CFR 820]." (This refers to general manufacturing and safety standards, not AI performance criteria.) |
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective):
This document describes a hardware device modification, not an AI algorithm. Therefore, there is no concept of a "test set" of patient data for evaluating an AI's performance. The "testing" referred to would be internal engineering verification and validation of the hardware and software functionality, not medical image analysis performance.
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):
Not applicable. As there's no AI algorithm being evaluated for diagnostic accuracy, there's no "ground truth" to establish from medical images or expert consensus. The function of the device is to print images accurately, not to interpret them. The document mentions that "The output of the device is evaluated by additional trained professionals allowing sufficient review to afford identification and intervention in the event of a malfunction," but this is about quality control of the printed output, not diagnostic ground truth.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set:
Not applicable for the reasons mentioned above.
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:
Not applicable. This device is an imager, a hardware component for displaying medical images, not an AI-powered diagnostic tool. An MRMC study is not relevant here.
6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done:
Not applicable. This is not an AI algorithm.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc):
Not applicable. There is no diagnostic ground truth established for this device, as its function is to print images, not to provide diagnostic interpretations.
8. The sample size for the training set:
Not applicable. This is a hardware device, not an AI algorithm requiring a training set of data.
9. How the ground truth for the training set was established:
Not applicable.
§ 892.2040 Medical image hardcopy device.
(a)
Identification. A medical image hardcopy device is a device that produces a visible printed record of a medical image and associated identification information. Examples include multiformat cameras and laser printers.(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). The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9.