(20 days)
KONICAMINOLTA DI-X1 is a software device that receives digital x-ray images and data from various sources (i.e. R/F Units, digital radiographic devices or other imaging sources). Images and data can be stored, communicated, processed and displayed within the system and/or across computer networks at distributed locations. It is not intended for use in diagnostic review for mammography.
KONICAMINOLTA DI-X1 is a software device that performs image processing and display using X-ray digital images (single-frame images, multi-frame images) generated by various diagnostic imaging modality consoles. It is a standalone software device intended to install onto off-the -shelf Servers and PCs. KONICAMINOLTA DI-X1 receives X-ray digital images, including serial images, processes the received images, as well as displays and sends the resulting images to PACS and other devices. In addition, KONICAMINOLTA DI-X1 can display images through the browser connection with the client that displays and process images, and instruct transmission of images. The personal computer used in KONICAMINOLTA DI-X1 stores the same data in two hard disks in real time using RAID1 mirroring function. Thus, even if one hard disk is defective, operations can be continued with the other hard disk which has the same data. The modifications are made on software to the identified predicate device to add the PC client to connect to the server using the browser on a personal computer to display images for a WEB reference. In addition, additional imaging processing MODES are implemented into the subject device. The subject device also modifies the graphical display to compare the past exam. graphs based on the measurement values in a chronological order.
Let's break down the information about the acceptance criteria and performance study for the KONICAMINOLTA DI-X1 device based on the provided FDA 510(k) summary.
It's important to note that the provided document does not contain a detailed performance study with human readers, specific metrics for AI performance (like sensitivity/specificity), or the methodologies for establishing ground truth for a test set. The submission states that "No clinical studies were required to support the substantial equivalence." This implies that the device's modifications are considered minor enough that extensive clinical validation, as would be expected for a novel AI diagnostic device, was not necessary.
The focus of this submission is on demonstrating substantial equivalence to a predicate device (K182431) for a medical image management and processing system, not a new diagnostic AI algorithm. The "modifications are made on software... to add the PC client to connect to the server using the browser... and additional imaging processing MODES are implemented... The subject device also modifies the graphical display to compare the past exam. graphs based on the measurement values in a chronological order." This suggests the "performance data" refers to validation of these functional additions and software changes, rather than a diagnostic accuracy study.
Therefore, many of the requested points regarding AI performance and human reader studies cannot be precisely answered from this document.
Here's the breakdown based on the provided text, with clarifications where information is absent:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criterion (Inferred from Device Description & Changes) | Reported Device Performance |
|---|---|
| Functional Equivalence to Predicate Device | The device has the same intended use, indications for use, technological characteristics, and principal operations as the predicate device (K182431). |
| Correct Operation of New Features: | |
| * PC client connectivity via browser for WEB reference | Demonstrated to function as intended. (Implied by the statement: "Performance tests demonstrate that the KONICAMINOLTA DI-X1 performs according to specifications and functions as intended.") |
| * New Imaging Processing Modes (PH-MODE, PH2-MODE, LM-MODE) | Demonstrated to function as intended. (Implied by the statement: "Performance tests demonstrate that the KONICAMINOLTA DI-X1 performs according to specifications and functions as intended.") |
| * Modified graphical display for chronological comparison of past exam graphs based on measurement values | Demonstrated to function as intended. (Implied by the statement: "Performance tests demonstrate that the KONICAMINOLTA DI-X1 performs according to specifications and functions as intended.") |
| Data Integrity and Reliability (RAID1 mirroring) | "The personal computer used in KONICAMINOLTA DI-X1 stores the same data in two hard disks in real time using RAID1 mirroring function. Thus, even if one hard disk is defective, operations can be continued with the other hard disk which has the same data." (This is a design feature, its successful implementation and testing would be part of "performance tests."). |
| Meeting all specifications and risk analysis requirements | "All the verification activities required by the specification and the risk analysis for the KONICAMINOLTA DI-X1 were performed and the results demonstrated that the predetermined acceptance criteria were met." |
| No new issues of safety or effectiveness | "The technological differences raised no new issues of safety or effectiveness as compared to its predicate device (K182431)." |
2. Sample Size Used for the Test Set and Data Provenance
The document does not specify a distinct "test set" in the context of a clinical performance study for an AI algorithm. The performance data section states: "All the verification activities required by the specification and the risk analysis for the KONICAMINOLTA DI-X1 were performed and the results demonstrated that the predetermined acceptance criteria were met. No clinical studies were required to support the substantial equivalence."
This indicates that the "testing" was likely functional and verification testing of the software's new features and overall operation, rather than a diagnostic accuracy evaluation on a patient image dataset. Therefore, information regarding data provenance (country of origin, retrospective/prospective) and sample size for such a test set is not provided because such a clinical test set was not deemed necessary for this submission.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications
Not applicable. As noted above, no clinical study requiring expert-established ground truth on a test set for diagnostic accuracy was reported or required for this 510(k) submission.
4. Adjudication Method for the Test Set
Not applicable. No clinical study requiring ground truth adjudication was reported.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No. The document explicitly states: "No clinical studies were required to support the substantial equivalence." Therefore, an MRMC study comparing human readers with and without AI assistance was not performed or reported.
6. Standalone (Algorithm Only) Performance
No. This device is described as a "medical image management and processing system" with enhancements, not a standalone AI diagnostic algorithm performing a specific diagnostic task (like detecting a disease). Its primary function is image handling, processing, and display. Therefore, a standalone performance metric (e.g., sensitivity/specificity for a disease) is not provided or applicable in the context of this submission.
7. Type of Ground Truth Used
Based on the lack of a clinical study, specific "ground truth" for diagnostic accuracy (e.g., pathology, outcomes data, expert consensus) was not established or used for performance evaluation in this 510(k). The "performance tests" focused on verifying the software's functional specifications.
8. Sample Size for the Training Set
Not applicable. This device is described as an image management and processing system, not a device incorporating a machine learning model that requires a "training set" in the typical sense of AI development for diagnostic tasks. The "modifications" were software developments, not AI model training.
9. How the Ground Truth for the Training Set Was Established
Not applicable, as there was no AI model training set as described in typical AI/ML submissions.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(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).