Search Results

The Auto Positioning feature provides an alternate, streamlined and efficient workflow and safety checks for the CT technologist in setting up CT examinations of the initiation of the first scout scan.

Auto Positioning acquires 3D spatial information of the individual patient on the table and combines it with information from the selected protocol to automatically calculate and visually display the scout's start and end locations. Concurrently it checks for proper patient orientation, determines the table height for optimum patient centering, and checks for potential contact between the patient and the gantry. Upon acceptance by the technologist the patient is automatically moved to the correct scout start location.

Use of Auto Positioning is intended to provide consistent patient positioning for optimal image quality and automatic exposure control.

Auto Positioning is an optional feature developed for use with GE CT systems. The purpose of this feature is to provide both a streamlined workflow and enhanced quality and safety checks during the exam setup process up to the initiation of the first scout scan. Incorporation of this optional feature does not preclude the technologist from preforming the existing manual workflow on the CT system, if desired.

Auto Positioning uses a fixed, ceiling mounted, off the shelf, 2D/3D video camera that is capable of determining distances to points in its field of view. It displays standard RGB video images on the CT system's existing gantry-mounted touchscreens. Information from the standard output of the camera, precise spatial information of the individual CT system's gantry and table installation geometry, and information contained in the user-selected protocol is used to determine the anatomical landmark location and the start and end locations for the scout scan(s).

Information from the standard output of the camera, precise spatial information of the individual CT system's gantry and table installation geometry, and information contained in the userselected protocol is used to determine the anatomical landmark location and the start and end locations for the scout scan(s).

Addition functionality of Auto Positioning includes performing safety checks for patient orientation and the potential for the patient to come into contact with the gantry while the patient is placed into the gantry and during scanning.

The GE Auto Positioning device is a patient positioning workflow enhancement tool for CT systems. It uses a 2D/3D video camera and deep learning to automate the process of setting the landmark, patient centering, scout's start and end locations, and the scout's start table position. It also performs safety checks for patient orientation and potential patient-gantry collision.

Here's an analysis of its acceptance criteria and the study that proves its effectiveness:

1. Table of Acceptance Criteria and Reported Device Performance

The provided document does not explicitly list a table of acceptance criteria with corresponding performance metrics for the Auto Positioning device. However, it indicates that "All testing met its predefined acceptance criteria." The narrative suggests the criteria would relate to the accuracy of landmark location, scout scan start and end locations, proper patient centering, correct patient orientation detection, and gantry collision prevention.

Based on the information, the reported device performance is that all non-clinical bench testing, which included the evaluation of landmark location and scout scan's start and end location, successfully met its predefined acceptance criteria. The document also states that the device was developed under GE Healthcare's quality system, and all subsystem and system verification testing, including hazard mitigation, demonstrated that the Auto Positioning feature meets its design inputs and user needs.

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

The document explicitly states: "Because Auto Postioning is for the exam setup process up to the initiation of the first scout scan, and does not involve diagnostic imaging or diagnostic evaluation, non-clinical bench testing is appropriate."

And further: "The Auto Positioning can be fully tested on the engineering bench thus no additional clinical testing was required."

Therefore, no clinical test set with patient data (and thus no associated sample size or data provenance) was used for direct performance evaluation for this 510(k) submission. The testing was conducted on an "engineering bench."

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

Since the testing was non-clinical bench testing and no clinical test set was used, there is no mention of experts or their qualifications for establishing ground truth from patient data. The "ground truth" for the non-clinical testing would have been established by engineering specifications and measurements. Engineers or technical experts involved in the design and verification would have assumed this role.

4. Adjudication Method for the Test Set

As there was no clinical test set using human readers, there was no adjudication method employed.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done

No, an MRMC comparative effectiveness study was not done. The document states that no clinical testing was required or performed. The device is focused on workflow enhancement and safety checks, not diagnostic image interpretation.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done

Yes, the primary evaluation was a standalone "algorithm only" performance, given the context of "non-clinical bench testing" and the statement that the device "can be fully tested on the engineering bench." The system's ability to accurately determine landmark locations, scout start/end points, patient orientation, and gantry collision potential was evaluated without human intervention in the positioning process itself. The technologist's role is to accept the automatically calculated positions.

7. The Type of Ground Truth Used

For the non-clinical bench testing described, the ground truth would have been based on:

• Engineering Specifications: Precisely defined parameters and measurements for expected landmark locations, optimal patient centering, and safe gantry clearance.
• Physical Measurements: Direct measurements of dummy patients, phantoms, or test setups on the engineering bench to represent ideal or boundary conditions. These measurements would then be compared against the device's output.

8. The Sample Size for the Training Set

The document mentions that "Auto Positioning uses Deep Learning CNNs to determine the scout's landmark location and the patient orientation." However, it does not provide any information regarding the sample size of the training set used for these Deep Learning Convolutional Neural Networks (CNNs).

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

The document does not describe how the ground truth for the training set was established for the Deep Learning CNNs. Typically, for such applications, ground truth for training data would involve:

• Manual Annotation by Experts: CT technologists or other medical imaging professionals manually identifying landmarks and patient orientations in a large dataset of patient images or 3D scans.
• Synthetic Data Generation: Creating artificial data with known ground truth based on anatomical models.
• Measurement from Phantoms/Physical Setups: Using a controlled environment with phantoms where the exact positions and orientations are known.

Without further information, the specific method used for the Auto Positioning device's deep learning training ground truth remains unknown from the provided text.

Ask a specific question about this device

The GE Revolution ACT Computed Tomography X-ray system is intended to produce cross- sectional images of the body by computer reconstruction of x-ray transmission data taken at different angles and planes, for patients of all ages, including Axial, Cine, Helical.

These images may be obtained either with or without contrast. This device may include signal analysis and display equipment, patient, and equipment supports, components and accessories.

This device may include data and image processing to produce images in a variety of trans-axial and reformatted planes. Further the images can be post processed to produce additional imaging planes or analysis results.

The GE Revolution ACT CT Scanner System is indicated for head, whole body and vascular X-ray Computed Tomography applications.

The device output is a valuable medical tool for the diagnosis of disease, trauma, or abnormality and for planning, guiding and monitoring therapy.

The multi-slice GE Revolution ACT CT scanner is currently commercially available and in clinical use in various other countries including the EU, Japan, China, and India.

It is a general purpose, 16-slice (detector row) CT scanning system with a z-coverage of 20 mm and a maximum gantry rotation speed of 0.98 seconds. Revolution ACT is designed to help enable greater patient access to CT imaging in facilities that otherwise might not be able to obtain multi-slice CT technology with both current standard and advanced CT features and function.

Revolution ACT uses the same technology, operating principles, features, and functions as the GE Optima CT520 predicate device (K123596) and other cleared GE CT scanners. The system consists of the gantry, patient table, operator console, power distribution unit (PDU), associated accessories, and software options. The Revolution ACT is also available in an 8-detector row (10 mm z-coveragre configuration (Revolution ACTs) using the identical (but depopulated) detector/DAS. The changes from the predicate device do not affect the intended use or patient population.

Becaues the Revolution ACT does not support cardiac or other gated acquisitions, and has a slower rotation time thatn the predicate device, its indications for use were modified by removing cardiac and gated acquisitions and cardiac applications.

The provided text describes the Revolution ACT Computed Tomography X-ray system and its substantial equivalence to the predicate device, Optima CT520. However, the text does NOT contain specific acceptance criteria with numerical values or a direct comparative study that reports device performance against such criteria for AI-related functions.

The document mainly focuses on establishing substantial equivalence to a predicate device for a CT scanner system, emphasizing hardware and imaging performance rather than an AI/ML component with specific performance metrics.

Based on the provided text, here’s a breakdown of the requested information, highlighting where the information is absent:

Acceptance Criteria and Study for Revolution ACT CT Scanner System (as per provided document)

The document primarily establishes substantial equivalence for a CT scanner system, Revolution ACT, to a predicate device, Optima CT520 (K123596). It focuses on the device's ability to produce cross-sectional images for diagnostic purposes. The "acceptance criteria" discussed are largely tied to compliance with standards and demonstration of equivalent performance to the predicate device in terms of image quality and safety.

1. A table of acceptance criteria and the reported device performance:

The document explicitly states that the Revolution ACT "maintains virtually the same image quality specifications and dose performance as its predicate" and that "verification testing along with additional engineering testing demonstrated Revolution ACT's equivalent performance to currently marketed the predicate and other cleared GE CT devices and is therefore as safe and effective."

However, a quantitative table with specific acceptance criteria (e.g., minimum spatial resolution, maximum noise level, sensitivity, specificity, or accuracy metrics) and corresponding reported performance values for the Revolution ACT is not provided in the given text. This section of the document describes general compliance and equivalence.

The text mentions a comparison of:

• CT number accuracy
• CT number uniformity
• Image noise (standard deviation)
• Modulation Transfer Function (MTF)
• Noise Power Spectrum
• Slice thickness

It states these comparisons were "provided" but does not present the actual values or the acceptance criteria for these metrics.

2. Sample size used for the test set and the data provenance:

• Test Set Sample Size: Not explicitly stated for specific quantitative performance tests. The document refers to "engineering bench testing" and "images from both a uniform phantom and one with embedded LCD objects."
• Data Provenance: Phantoms were used for engineering bench testing. For clinical verification, "sample clinical images" were reviewed. The origin of these clinical images (e.g., country) is not specified.
• Retrospective or Prospective: Not specified for the clinical images used for review. Phantom studies are inherently controlled.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• Number of Experts: "A board certified radiologist." (Singular)
• Qualifications of Experts: "Board certified radiologist."

4. Adjudication method for the test set:

• Adjudication Method: "Sample clinical images reviewed by a board certified radiologist." This implies a single expert review, so no multi-reader adjudication method (e.g., 2+1, 3+1) 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:

• MRMC Comparative Effectiveness Study: No, an MRMC comparative effectiveness study involving human readers and AI assistance is not described in the provided text. The device is a CT scanner, and while it has "signal analysis and display equipment" and "data and image processing," the document does not focus on an AI-assisted diagnostic function. The primary focus is on the scanner's core imaging performance and safety. The "SmartPlan" feature mentioned is for workflow enhancement (initial scan setup parameters) and not for diagnostic assistance that would typically be evaluated in an MRMC study for improved reader performance.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• Standalone Performance Study: The document does not describe a standalone performance study for an AI algorithm. The device itself (the CT scanner) undergoes performance testing against engineering specifications. The "ASiR" (Adaptive Statistical Iterative Reconstruction) algorithm is mentioned as being ported to new hardware, and its performance is verified through clinical image review by a radiologist, but this is not presented as a standalone AI algorithm evaluation.

7. The type of ground truth used:

• For engineering bench testing: Physical phantoms with known properties (uniform phantom, phantom with embedded LCD objects).
• For clinical image review: Implied clinical diagnosis/reference for the "sample clinical images reviewed by a board certified radiologist." However, the explicit nature of this ground truth (e.g., pathology, outcomes data) is not detailed. It is likely derived from standard clinical practice and radiologist interpretation.

8. The sample size for the training set:

• Training Set Sample Size: Not applicable/not provided. The document describes a CT scanner and its underlying technologies. While reconstruction algorithms like ASiR might involve models, the document does not detail their training or associated datasets. The focus is on the hardware platform and its imaging capabilities.

9. How the ground truth for the training set was established:

• Ground Truth for Training Set: Not applicable/not provided, as no training set for an AI/ML diagnostic algorithm is described.

Ask a specific question about this device