(53 days)
Advantage Windows Tisue Volume Option is intended to calculate the volume of any three dimensional structure from a set of CT images. The images are displayed and measured on the CT/MR Advantage Windows Diagnostic Workstation (K913770). The Tissue Volume Option will automate manual measurements and estimates that are currently made regarding the measure of volume of anatomy and pathology imaged with computed tomography.
The device is a software package to be used on the Advantage Windows Workstation (K913770).
Design: The design is based on the same software platform as used for Advantage Windows 3D (VoxTool) and will operate on the Advantage Windows Operating System. The workstation hardware required for operation will be the Sun Sparc 20, Mod 40 or higher. The images used to measure volume can be captured by any CT scanner and transferred to Advantage Windows workstation by DICOM or Ethernet.
Energy Source and Exposure Levels: There is no energy source associated with this package in and of itself. However, the energy source used to make the image being analyzed is the same used to take standard CT diagnostic images.
Principals of Operation: The same as Advantage Windows 3D.
Features: To calculate the volume of areas of a CT image selected by the operator.
Accessories: None
This document, K963345, is a Summary of Safety & Effectiveness for the Advantage Windows Tissue Volume Option. It is a pre-market notification (510(k)) and largely focuses on substantial equivalence to a previously cleared device (K913770, Advantage Windows 3D).
Based on the provided information, it is not possible to fully answer all the questions regarding acceptance criteria and studies that prove the device meets them. This document is a 510(k) summary, not a detailed study report. 510(k) submissions primarily focus on demonstrating substantial equivalence, and often don't include detailed performance studies with acceptance criteria in the same way a PMA (Pre-Market Approval) or a clinical trial report would.
Here's a breakdown of what can be gleaned and what cannot:
1. A table of acceptance criteria and the reported device performance
- Cannot be provided. The document states: "The Advantage Windows Tissue volume Option enhances the current Advantage Windows 3D package by automaiting volumetric measurements that are currently being made manually using the measure distance tools of the 3D package. It is substantially equivalent to the Advantage Windows 3D package in design, construction, principle of operation, and features."
This statement indicates a reliance on substantial equivalence to an existing device (Advantage Windows 3D, K913770), rather than presenting new performance data against specific acceptance criteria for a novel functionality. There are no explicit acceptance criteria or reported performance metrics (like accuracy, precision, sensitivity, specificity, etc.) for the new "Tissue Volume Option" presented in this summary.
2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
- Cannot be provided. No information on a specific test set, its size, or data provenance is mentioned. The submission relies on the established performance of the predicate device.
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)
- Cannot be provided. No ground truth establishment for a test set is discussed. The device "automates manual measurements and estimates that are currently made" implying that the manual measurements by an operator (presumably a clinician) would be the reference, but no formal ground truth process is described.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
- Cannot be provided. No test set or adjudication method is described.
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, an MRMC comparative effectiveness study was not explicitly done or reported. The device is described as an "option" to automate current manual measurements. While it might assist human readers by automating volume calculation, a formal MRMC study comparing human performance with and without this specific AI-driven automation is not mentioned.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
- A standalone performance study focused on the algorithm is not explicitly reported in this summary. The document describes the device as a "software package to be used on the Advantage Windows Workstation" and emphasizes that "The oulining is in the hands of the operator who can choose to accept or reject the outlined region." This indicates a human-in-the-loop process where the operator's input (outlining) is crucial, implying it's not a purely standalone algorithm making diagnostic decisions. It assists in measuring after operator input.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
- Implicitly, the ground truth would be the manual measurements and estimates currently made by operators. The device aims to "automate manual measurements and estimates." However, there's no mention of a formal process to establish this ground truth for validation purposes in this document.
8. The sample size for the training set
- Cannot be provided. No information on a training set or its size is mentioned. Given the device's description as automating "manual measurements and estimates," its "training" might be more aligned with software development principles rather than machine learning model training on a large dataset in the modern sense. It's likely a rule-based or image processing algorithm based on established methods, rather than a deep learning model.
9. How the ground truth for the training set was established
- Cannot be provided. No training set or ground truth establishment process is mentioned.
In summary: This 510(k) summary focuses on demonstrating substantial equivalence to a predicate device. It describes a software feature that automates an existing manual task (volume calculation). It does not present detailed performance studies, specific acceptance criteria, or robust validation datasets (test and training sets) as would be expected for a novel AI/ML device submission today. The lack of this information is typical for many 510(k) submissions from this era (1996) that relied heavily on substantial equivalence to existing technologies.
§ 892.1000 Magnetic resonance diagnostic device.
(a)
Identification. A magnetic resonance diagnostic device is intended for general diagnostic use to present images which reflect the spatial distribution and/or magnetic resonance spectra which reflect frequency and distribution of nuclei exhibiting nuclear magnetic resonance. Other physical parameters derived from the images and/or spectra may also be produced. The device includes hydrogen-1 (proton) imaging, sodium-23 imaging, hydrogen-1 spectroscopy, phosphorus-31 spectroscopy, and chemical shift imaging (preserving simultaneous frequency and spatial information).(b)
Classification. Class II (special controls). A magnetic resonance imaging disposable kit intended for use with a magnetic resonance diagnostic device only is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9.