Aquilion Precision (TSX-304A/1 and /2) V8.8 with AiCE

The Aquilion Precision (TSX-304A/1 and /2) V8.8 is an ultra-high resolution whole-body CT scanner that introduces AiCE (Advanced intelligent Clear-IQ Engine), a Deep Convolutional Neural Network-based algorithm for image reconstruction. This technology is engineered to significantly reduce image noise, enhance low contrast detectability by 13%, and improve spatial resolution in abdomen and pelvis imaging, while maintaining dose neutrality compared to previous reconstruction methods like AIDR 3D.

The Canon Aquilion Precision (TSX-304A/1 and /2) V8.8 is an ultra-high resolution, whole-body, multi-slice helical Computed Tomography (CT) scanner designed to acquire and display cross-sectional volumes of the entire body, including the head. It features a 160-row, 0.25 mm detector, a 5.7-MHU large-capacity tube, and 0.35 s scanning capabilities, enabling wide-range scanning with short acquisition times. The system incorporates two primary image reconstruction technologies: FIRST 3.0, an iterative algorithm aimed at reducing radiation exposure and enhancing high-contrast spatial resolution for applications including the abdomen, pelvis, chest, cardiac, extremities, and head. A key modification is the introduction of AiCE (Advanced intelligent Clear-IQ Engine), a novel noise reduction algorithm specifically for abdomen and pelvis applications, which utilizes Deep Convolutional Neural Network (DCNN) methods to improve image quality and reduce noise in 1024x1024 high-resolution/super-high resolution images. The AiCE algorithm's methodology involves DCNNs that are trained to differentiate between anatomical structures and image noise by analyzing the statistical properties of the signal, thereby producing fast and high-quality CT reconstructions. Performance evaluations, conducted through phantom testing, demonstrated significant improvements with AiCE compared to the earlier AIDR 3D algorithm. Specifically, for super-high resolution body imaging at equivalent radiation doses, AiCE achieved a 13% improvement in low contrast detectability and a 42% reduction in image noise. The system also maintained dose neutrality between its super-high resolution mode with AiCE and normal resolution mode with AIDR, while still offering superior low contrast detectability. Furthermore, AiCE provided an enhanced high-contrast spatial resolution, claiming an improvement of 8.8 lp/cm at 10% of the Modulation Transfer Function (MTF) for abdomen and body imaging. The diagnostic quality of AiCE-reconstructed images for abdomen and pelvis applications was confirmed through review by a board-certified radiologist.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the performance of the AiCE Deep Convolutional Neural Network vary across different patient body mass indexes in abdomen and pelvis applications?; What are the specific quantitative improvements in diagnostic accuracy or changes in clinical decision-making observed when using AiCE-reconstructed images compared to AIDR 3D for challenging abdominal and pelvic pathologies?; Beyond the documented noise reduction and improved Low Contrast Detectability, what is the impact of AiCE on the conspicuity and delineation of subtle or complex anatomical structures within ultra-high resolution CT images of the abdomen and pelvis?; What are the computational demands and reconstruction times for AiCE when processing 1024x1024 HR/SHR images, and how do these compare to FIRST 3.0 and AIDR 3D in a typical clinical workflow?; For which specific clinical scenarios or suspected pathological findings in the abdomen and pelvis does the 13% improvement in low contrast detectability with AiCE offer the most significant diagnostic advantage or potential to alter patient management?