Search Results

INTEGRATED REGISTRATION provides easy means for comparison of three-dimensional (3D) images from Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Emission Tomography (PET or SPECT) and X-Ray Angiography (XA). To help physicians in diagnostic radiology or therapy planning, INTEGRATED REGISTRATION allows 3D registration between volumetric acquisitions that may come from the same acquisition modality or from different acquisition modalities.

The INTEGRATED REGISTRATION tool runs on Advantage Workstation 4.5 or higher versions. This product is an extension to the Volume Viewer application, dedicated to the registration of multi-modality images, and comparison of volumetric datasets from Computed Tomography (CT), Magnetic Resonance (MR), Positron Emission Tomography (PET) or Single Photon Emission Computed Tomography (SPECT), and 3D X-Ray Angiography (XA). Note that the INTEGRATED REGISTRATION licenses control which algorithms are available and which modalities can be saved.

The provided FDA 510(k) Premarket Notification Submission for the GE Healthcare INTEGRATED REGISTRATION device does not include acceptance criteria or a study proving the device meets acceptance criteria in the way typically found for AI/ML-based diagnostic devices.

Instead, the submission focuses on demonstrating substantial equivalence to predicate devices through technical comparisons and a summary of non-clinical tests. Clinical studies were explicitly stated as not required to support substantial equivalence.

Here's an breakdown based on the information provided, explicitly noting where information is missing for the requested categories:

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample Size Used for the Test Set and Data Provenance

• Test Set Sample Size: Not specified.
• Data Provenance: Not specified. (No clinical data was used for substantial equivalence.)
• Retrospective/Prospective: Not applicable, as no clinical studies were performed.

3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications

• Number of Experts: Not applicable. No clinical test set requiring expert ground truth was used.
• Qualifications: Not applicable.

4. Adjudication Method for the Test Set

• Adjudication Method: Not applicable. No clinical test set requiring adjudication was used.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• Was an MRMC study done? No. The submission explicitly states: "The subject of this premarket submission, INTEGRATED REGISTRATION, did not require clinical studies to support substantial equivalence."
• Effect Size of Human Readers Improve with AI vs. without AI: Not applicable, as no MRMC study was conducted.

6. Standalone (Algorithm Only) Performance Study

• Was a standalone study done? No, not in the sense of a clinical performance study with specific metrics. The document mentions "Performance testing (Verification)" as part of non-clinical tests, but no details of such a study or its results are provided. The device itself is an "optimized combination of its predicate devices" with "optimized registration algorithms," implying algorithm-only functionality but without a standalone performance study as typically understood for AI devices.

7. Type of Ground Truth Used

• Type of Ground Truth: Not applicable. No clinical ground truth was established for "substantial equivalence." The basis for equivalence relies on compliance with standards and non-clinical testing.

8. Sample Size for the Training Set

• Training Set Sample Size: Not specified. This device predates the common discussion around AI/ML training sets for medical devices. The "optimized registration algorithms" would have been developed using some form of data, but the submission does not detail any "training set."

9. How the Ground Truth for the Training Set Was Established

• Ground Truth Establishment: Not applicable/Not specified. Given the nature of the device as an "optimized combination" of existing algorithms for image registration, ground truth for algorithm development (if implicitly used) would likely relate to image alignment accuracy, which typically relies on synthetic data or manually established spatial correspondences rather than clinical "ground truth" labels.

Ask a specific question about this device