(118 days)
The ASIR-V Reconstruction Option is intended to produce cross-sectional images of the head and body by computer reconstruction of X-ray transmission data taken at different angles and planes, including Axial, Helical (Volumeric), and Cardiac acquisitions for all ages.
When used, it allows for an alternate reconstruction method designed to reduce image noise and streak artifact, increase resolution and improve low contrast detectability in images produced using raw Computed Tomography data from the GE CT scanner. The ASIR-V Reconstruction Option can be used to reduce noise in images and also to reduce the dose required for diagnostic CT imaging, including scans of the head, chest, heart, abdomen and pelvis. The ASIR-V Reconstruction Option may also improve the image quality of low dose non-diagnostic Filtered Back Projection images such that they become diagnostic.
ASiR-V reconstruction option is for use with the GE CT Scanners.
The GE ASIR-V Reconstruction Option is an alternate CT image reconstruction option for GE CT Systems to Filtered Back Projection, the advanced GE iterative reconstruction software Veo Reconstruction Option (K103489) and the ASiR reconstruction option.
This reconstruction technique is the advanced GE iterative reconstruction method intended to be used when higher image quality and/or lower dose acquisitions are desired for all routine cases to improve image performance such as Low Contrast Detectability, Image Noise, Spatial Resolution, artifact reduction, etc. Image quality improvements and dose reduction depend on the clinical task, patient size, anatomical location, and clinical practice.
The GE ASIR-V Reconstruction Option when used with the CT System performs as well as or better than Filtered Back Projection and uses operating principles derived from ASIR and Veo.
Here's a summary of the ASiR-V device's acceptance criteria and the study that proves it meets them, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
The acceptance criteria for the ASiR-V Reconstruction Option, relative to Filtered Back Projection (FBP), are outlined below along with the reported performance from phantom-based tests.
| Acceptance Criterion (Relative to Filtered Back Projection) | Reported Device Performance (Relative to FBP) |
|---|---|
| Lower radiation dose at the same image quality | 50% to 82% dose reduction |
| Improved Low Contrast Detectability (LCD) at the same dose | 59% to 135% LCD improvement |
| Reduced Image Noise at the same dose | Up to 91% image noise reduction |
| Improved Spatial Resolution at the same image noise | Up to 2.07X (107%) spatial resolution improvement |
| Reduced low signal artifacts (e.g., streak artifact) | Possesses the capability of low signal artifact reduction (e.g., streak artifact) |
2. Sample Size and Data Provenance (Test Set)
- Sample Size for Clinical Image Evaluation: 96 patient exams
- Data Provenance: The document does not explicitly state the country of origin for the patient data, nor does it explicitly state if it was retrospective or prospective. It refers to "sample clinical images" and "retrospective clinical read," suggesting the data was pre-existing and evaluated retrospectively. The study was conducted by GE Healthcare, which has manufacturing locations in Japan and the USA, and China, but this doesn't specify data origin.
3. Number and Qualifications of Experts for Test Set Ground Truth
- Number of Experts: Four radiologists.
- Qualifications: The document does not specify the years of experience or specific subspecialties of the radiologists.
4. Adjudication Method (Test Set)
The radiologists rated the diagnostic image quality (IQ) using a 5-point Likert scale for 192 images (96 exams, each reconstructed with FBP and ASiR-V). After this individual rating, the "total score of each recon algorithm is compared." This suggests individual assessments were aggregated, but no formal adjudication method like a 2+1 or 3+1 consensus process between the radiologists is explicitly described. They simply "rated" the images and then "total scores" were compared.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
A multi-reader multi-case (MRMC) comparative effectiveness study was conducted for diagnostic image quality.
- Effect Size of Human Readers with AI vs. without AI assistance: The document states that ASiR-V "demonstrated the clinical diagnostic quality of the ASiR-V images across various CT platforms and patient anatomies" and that "the total score of each recon algorithm is compared." It concludes that ASiR-V provides "equivalent or better performance (no loss of diagnostic quality...)" compared to FBP. However, it does not provide a specific quantitative effect size for how much human readers improved with ASiR-V assistance compared to reading FBP images alone. The evaluation focused on the diagnostic quality of the ASiR-V images themselves, not on the delta improvement in reader performance.
6. Standalone (Algorithm Only) Performance
Yes, standalone (algorithm only) performance was done. The majority of the acceptance criteria (dose reduction, LCD improvement, noise reduction, spatial resolution improvement, artifact reduction) were demonstrated through phantom-based tests, which assesses the algorithm's direct impact on image characteristics without human interaction in the loop for initial assessment. The model observer study for LCD also represents a standalone assessment.
7. Type of Ground Truth Used
- For performance metrics (dose, LCD, noise, spatial resolution, artifact reduction): Phantom-based measurements and model observer analysis were used, which are objective quantitative assessments based on known phantom characteristics.
- For clinical image evaluation: Expert Likert scale ratings of diagnostic image quality by radiologists were used. This is a form of expert consensus/opinion on diagnostic quality.
8. Sample Size for Training Set
The document does not specify a sample size for the training set. The descriptions focus on the evaluation of the ASiR-V algorithm, not its development or training data.
9. How Ground Truth for Training Set Was Established
As the document does not provide information about a training set or its sample size, it also does not specify how ground truth for a training set was established.
§ 892.1750 Computed tomography x-ray system.
(a)
Identification. A computed tomography x-ray system is a diagnostic x-ray system intended to produce cross-sectional images of the body by computer reconstruction of x-ray transmission data from the same axial plane taken at different angles. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II.