(78 days)
This device is indicated to acquire and display cross-sectional volumes of the whole body, to include the head.
The Aquilion Prime SP has the capability to provide volume sets. These volume sets can be used to perform specialized studies, using indicated software/hardware, by a trained and qualified physician.
The Aquilion Prime SP TSX-303B/1 is an 80-row CT System that is intended to acquire and display cross-sectional volumes of the whole body, including the head. This system is based upon the technology and materials of previously marketed Toshiba CT systems.
The provided text describes a 510(k) submission for the Toshiba Aquilion Prime SP, TSX-303B/1, v8.4. It outlines modifications to a previously cleared CT system. While the document mentions various performance evaluations and studies, it does not contain specific acceptance criteria tables nor detailed study designs that definitively "prove" the device meets acceptance criteria in the format of a typical peer-reviewed clinical study. Instead, it focuses on demonstrating substantial equivalence to a predicate device through engineering and performance testing.
However, I can extract and infer information about the testing and performance as described in the document.
Missing Information:
- A clear table of acceptance criteria for specific performance metrics. The document describes improvements but doesn't explicitly state "acceptance criteria" values met.
- Detailed sample sizes for all tests.
- Specific data provenance for all tests (e.g., country of origin, retrospective/prospective).
- Number and qualifications of experts for all ground truth establishment.
- Adjudication methods.
- MRMC comparative effectiveness study details (effect size of human reader improvement with AI).
- Standalone algorithm performance (the device is a CT system, not an algorithm in the AI sense).
- Sample size for the training set.
- How ground truth for the training set was established.
Based on the provided text, here's what can be extracted and inferred:
1. A table of acceptance criteria and the reported device performance
The document does not explicitly present a table of acceptance criteria. Instead, it describes performance improvements and that the modified system "demonstrates equivalent or slightly improved image quality characteristics." The performance evaluations are primarily focused on physical parameters and dose reduction, not diagnostic accuracy in the way an AI algorithm might be assessed against clinical endpoints.
| Performance Metric | Reported Device Performance (Aquilion Prime SP, TSX-303B/1, v8.4) | Implied Acceptance Criterion (relative to predicate) |
|---|---|---|
| Spatial Resolution | Evaluated; demonstrated equivalent or slightly improved image quality. | Equivalent or improved |
| Axial Slice Thickness/Slice Sensitivity Profile | Evaluated; demonstrated equivalent or slightly improved image quality. | Equivalent or improved |
| CT Number Magnitude/Uniformity | Evaluated; demonstrated equivalent or slightly improved image quality. | Equivalent or improved |
| Noise Properties | Evaluated; demonstrated equivalent or slightly improved image quality. | Equivalent or improved |
| Low Contrast Detectability (LCD) | Evaluated; demonstrated equivalent or slightly improved image quality. | Equivalent or improved |
| Contrast-to-Noise Ratio (CNR) | Evaluated; demonstrated equivalent or slightly improved image quality. | Equivalent or improved |
| Dose Reduction (with AIDR 3D Enhanced) | 51% to 75% dose reduction supported while preserving LCD and high contrast spatial resolution. | Not explicitly stated, but demonstrated within range |
| Dose Reduction (with PURE ViSION Optics) | 20%-31% quantitative dose reduction. | Not explicitly stated, but demonstrated within range |
| LCD Improvement (Head, PURE ViSION Optics) | Range 13%-19% improvement. | Not explicitly stated, but demonstrated improvement |
| LCD Improvement (Body, PURE ViSION Optics) | Range 15%-22% improvement. | Not explicitly stated, but demonstrated improvement |
| Noise Reduction (PURE ViSION Optics) | 13% noise reduction at the same dose. | Not explicitly stated, but demonstrated improvement |
| Diagnostic Quality of Images | Produces images of diagnostic quality for head, chest, abdomen, and peripheral exams. | Diagnostic quality maintained |
2. Sample size used for the test set and the data provenance
- Sample Size for Physical Performance Tests: Not explicitly stated. The tests involved "model observer studies" using MITA-FDA LCD Head and MITA-FDA LCD Body phantoms, implying a phantom-based test set rather than patient data.
- Sample Size for Image Review: "Representative diagnostic images" were obtained. The exact number is not specified.
- Data Provenance: Not specified. Phantoms for performance tests. Clinical images for diagnostic quality assessment (implicitly from a clinical setting, but no country of origin or retrospective/prospective status is mentioned).
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
- Number of Experts: One.
- Qualifications of Expert: An "American Board Certified Radiologist." Further details on experience (e.g., years) are not provided.
- Role: This radiologist "reviewed" the "representative diagnostic images" to confirm they were of "diagnostic quality."
4. Adjudication method for the test set
- Adjudication Method: Not applicable or not specified in detail. The document states a single American Board Certified Radiologist reviewed images. There is no mention of consensus or multi-reader adjudication for this informal review of diagnostic quality. For the quantitative performance metrics (dose reduction, LCD, noise), these were based on phantom studies and model observer analysis, not human reader adjudication.
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 Study: No. The document does not describe a MRMC comparative effectiveness study. This submission is for a CT system itself, not an AI-assisted diagnostic tool in the typical sense of showing improved human reader performance. The "AI" mentioned (AIDR 3D Enhanced, SEMAR) refers to image processing algorithms within the CT system to improve image quality or reduce artifacts, not a separate AI application for diagnosis or interpretation assistance that would warrant an MRMC study comparing human readers with and without its aid.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
- Standalone Performance: Yes, in a way. The "performance testing" of the modified system, including spatial resolution, CT number, noise properties, LCD, and CNR, as well as the quantitative dose reduction and LCD/noise improvement studies using phantoms and model observers, represent a standalone evaluation of the system's technical image quality parameters. These are inherent algorithmic and hardware performance metrics of the CT scanner, not dependent on human interpretation for their measurement.
7. The type of ground truth used
- For Quantitative Performance: Model observer studies using MITA-FDA LCD Head and MITA-FDA LCD Body phantoms. These phantoms represent a controlled, objective ground truth for physical image quality parameters.
- For Diagnostic Quality: The subjective assessment of an "American Board Certified Radiologist" confirming images were of "diagnostic quality." This is expert opinion/consensus for a qualitative judgment rather than a definitive "ground truth" like pathology.
8. The sample size for the training set
- Training Set Sample Size: Not applicable / Not provided. This document describes a 510(k) submission for a CT scanner, not a machine learning algorithm that requires a "training set" in the conventional sense. While there might be internal development and validation data, it's not discussed as a distinct "training set" within this regulatory context.
9. How the ground truth for the training set was established
- Ground Truth Establishment for Training Set: Not applicable / Not provided, as there is no described training set for an AI algorithm in the context of this submission.
§ 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.