Search Results

This device is indicated to acquire and display cross sectional volumes of the whole body, to include the head. The Aquilion Serve has the capability to provide volume sets can be used to perform specialized studies, using indicated software/hardware, by a trained and qualified physician. AiCE (Advanced Intelligent Clear-IQ Engine) is a noise reduction algorithm that improves image quality and reduces image noise by employing Deep Convolutional Neural Network methods for abdomen, pelvis, lung, cardiac, extremities, head and inner ear applications.

Aquilion Serve (TSX-307A/1) V1.2 with AiCE-i (Advanced intelligent Clear-IQ Engine-integrated) is a whole body multi-slice helical CT scanner, consisting of a gantry, couch and a console used for data processing and display. This device captures cross sectional volume data sets used to perform specialized studies, using indicated software/hardware, by a trained and qualified physician. This system is based upon the technology and materials of previously marketed Canon CT systems.

The provided text describes the regulatory clearance of the Aquilion Serve (TSX-307A/1) V1.2 with AiCE-i, a Computed Tomography X-ray System, by the FDA. The submission outlines comparisons to a predicate device and various performance tests. However, the document focuses on demonstrating substantial equivalence to a predicate device rather than presenting a formal study with acceptance criteria and a detailed breakdown of results for a specific performance claim of the device's AI component (AiCE) in improving human reader performance.

Several key pieces of information requested, particularly regarding clinical studies focused on AI's impact on human readers, are not explicitly present in the provided text. The document describes technical performance comparisons of the AiCE algorithm (noise reduction, image quality metrics) and clinical verification of diagnostic quality, but not a multi-reader multi-case (MRMC) study on human performance with and without AiCE assistance.

Therefore, the following response will extract the information available and indicate where the requested information is not provided in the supplied text.

Acceptance Criteria and Study Proving Device Meets Criteria

The primary objective of this submission is to demonstrate substantial equivalence of the Aquilion Serve (TSX-307A/1) V1.2 with AiCE-i to a predicate device (Aquilion Lightning (TSX-036A/7) V10.2 with AiCE-i). The "acceptance criteria" can therefore be inferred as demonstrating equivalent or improved performance in various image quality metrics and functional aspects compared to the predicate device, or demonstrating image quality that is diagnostically acceptable. There is no specific acceptance criteria table provided in the context of an "AI effect" on reader performance.

Here's a breakdown of the requested information based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

As there isn't a single "AI acceptance criteria" table for human performance, the table below summarizes the objective image quality performance comparisons the manufacturer conducted to support their claims about AiCE. The "acceptance criteria" is implicitly "equivalent or improved performance" compared to the baseline (FBP, AIDR 3D, or predicate device).

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

• Test Set Sample Size: The document does not specify a distinct "test set" sample size in terms of number of patient cases for the clinical image review. It mentions "Representative body, cardiac, chest, head, and extremity diagnostic images." For phantom studies, specific phantom types (e.g., MITA - FDA LCD Body phantom, 24cm and 32cm water phantoms) are mentioned but not a "sample size" in terms of repeated measurements, though implied through quantitative reporting.
• Data Provenance: Not explicitly stated. The document refers to "volunteer assessment" for automatic scan planning and "clinical images" for diagnostic quality review. The location (country of origin), and whether the data was retrospective or prospective, are not provided.

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

• Number of Experts: For the clinical image review, the document states "reviewed by an American Board-Certified Radiologist." This implies a single expert.
• Qualifications of Experts: "American Board-Certified Radiologist". No mention of years of experience.

4. Adjudication Method for the Test Set

• For the "clinical images," it states "reviewed by an American Board-Certified Radiologist," indicating a single reader, so no adjudication method is described or implied.
• For phantom studies, the ground truth is objective measurements (e.g., from phantoms), not requiring expert adjudication.

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

• No MRMC study is described in the provided text. The document focuses on technical image quality metrics and a single radiologist's review of diagnostic quality, not a comparative study showing human readers' improvement with or without AI assistance.

6. Standalone (Algorithm Only) Performance

• Yes, standalone performance was assessed for the AiCE algorithm through various phantom and objective image quality metric evaluations (e.g., CNR, MTF, NPS, LCD, noise reduction claims). These evaluate the algorithm's output (image quality characteristics) without human interpretation in the loop.

7. Type of Ground Truth Used

• Phantom-based objective ground truth: For most image quality metrics (CNR, CT number accuracy, uniformity, SSP, MTF, SD, NPS, LCD, spatial resolution, dose reduction claims), physical phantoms were used, and the "ground truth" was derived from objective measurements on these phantoms.
• Expert Consensus/Pathology/Outcomes Data:
  • For the "Clinical Images" evaluation, the ground truth was the subjective assessment of diagnostic quality by a single American Board-Certified Radiologist. This is a form of expert assessment, though not explicitly a "consensus" by multiple experts.
  • No pathology or outcomes data is mentioned as ground truth in this submission for the AI (AiCE) component.

8. Sample Size for the Training Set

• Not provided. The document states that AiCE uses "Deep Convolutional Neural Network methods," implying a training phase, but gives no details on the training dataset size.

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

• Not provided. The document does not detail how the ground truth for the training data used by the Deep Convolutional Neural Network (AiCE) was established. This information would typically be found in more detailed technical specifications or a different section of a submission.

Ask a specific question about this device