Aquilion Precision (TSX-304A/4) V10.14 with AiCE

The Aquilion Precision (TSX-304A/4) V10.14 is an ultra-high resolution CT system designed to provide superior diagnostic images of the whole body by leveraging advanced hardware, including a 0.25mm detector and fast scanning. Its core science involves innovative image processing through FIRST iterative reconstruction for dose reduction and AiCE, a Deep Convolutional Neural Network, for significant noise reduction and image quality enhancement, now extended to extremities, head, and inner ear applications.

The Aquilion Precision (TSX-304A/4) V10.14 with AiCE is an ultra-high resolution, whole-body multi-slice helical CT scanner designed to acquire and display cross-sectional volumes for comprehensive diagnostic imaging. It integrates a 160-row detector with 0.25 mm elements, a 5.7-MHU large-capacity X-ray tube, and rapid 0.35-second scanning capabilities. These features enable wide-range scanning with short acquisition times, capturing detailed volume datasets essential for specialized studies across the entire body, including the head, when operated by trained physicians using the indicated software and hardware components. The system employs two primary advanced software technologies: FIRST, an iterative reconstruction algorithm intended to reduce patient radiation exposure dose while maintaining or enhancing high-contrast spatial resolution for applications including abdomen, pelvis, chest, cardiac, extremities, and head. Complementing this is AiCE (Advanced intelligent Clear-IQ Engine), a sophisticated noise reduction algorithm that utilizes Deep Convolutional Neural Network (DCNN) methods. AiCE is designed to improve image quality and reduce noise in 1024x1024 high-resolution (HR) and super-high resolution (SHR) images across abdomen, pelvis, lung, cardiac, and with this version, expanded applications to extremities, head, and inner ear. Bench and clinical testing demonstrated substantial equivalence, confirming image quality metrics (like CNR, MTF, SD, NPS, LCD) and quantitative spatial resolution improvements, with reconstructed images deemed of diagnostic quality.

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 AiCE in the newly added scan regions (extremities, head, inner ear) compare to its performance in previously established regions in terms of quantitative noise reduction and perceived image quality improvement?; What specific Deep Convolutional Neural Network architecture is utilized in AiCE, and what were the key considerations or challenges in training and validating this network for the diverse anatomical structures of extremities, head, and inner ear?; The document mentions a spatial resolution improvement of 5.37 lp/cm for Bone, 7.50 lp/cm for Inner Ear, and 11.31 lp/cm for Brain CTA in HR mode with AiCE compared to hybrid iterative reconstruction. What are the corresponding dose levels for these comparisons, and how do these resolution figures translate to tangible diagnostic benefits in clinical scenarios for these specific anatomies?; How does the FIRST iterative reconstruction algorithm balance exposure dose reduction with the preservation of low-contrast detectability, particularly in abdominal and pelvic imaging, and are there any specific trade-offs clinicians should be aware of?; Given the ultra-high resolution capabilities (0.25 mm detector, 1024x1024 matrix), what is the typical reconstruction time for complex datasets using AiCE, and how does this impact clinical workflow, especially in high-throughput environments?