Aquilion Precision (TSX-304A/4) V10.10 with AiCE

The Aquilion Precision (TSX-304A/4) V10.10 with AiCE is an advanced computed tomography system designed for whole-body, ultra-high resolution imaging, utilizing a Deep Convolutional Neural Network (AiCE) and iterative reconstruction (FIRST) to enhance image quality, reduce noise, and improve spatial resolution across diverse clinical applications including cardiac and lung imaging. This technology aims to provide superior diagnostic images at potentially optimized radiation doses by intelligently differentiating anatomical structures from noise during the reconstruction process.

The Aquilion Precision (TSX-304A/4) V10.10 with AiCE is an ultra-high resolution, whole-body multi-slice helical CT scanner designed to acquire and display cross-sectional volumes for various diagnostic applications, including head, abdomen, pelvis, chest, cardiac, and extremities. It incorporates a 160-row, 0.25 mm detector and rapid scanning capabilities to capture detailed volume data sets. A key technological advancement is its AiCE (Advanced intelligent Clear-IQ Engine) feature, which utilizes Deep Convolutional Neural Network (DCNN) methods for image reconstruction. This DCNN-based approach is engineered to reduce image noise and enhance image quality by differentiating between anatomical structures and noise, aiming for fast, high-quality CT reconstructions, particularly for 1024x1024 high-resolution (HR) and super-high-resolution (SHR) images. The system also includes the FIRST iterative reconstruction algorithm for dose reduction and improved high-contrast spatial resolution. This version (V10.10) builds upon its predicate by expanding the application of AiCE to lung and cardiac imaging, in addition to the existing abdomen and pelvis applications, and enhances gantry tilt operability with AiCE. The AiCE algorithm was retrained for these new indications. Performance testing demonstrated AiCE's substantial equivalence to the predicate, with improvements in noise texture, making it more natural than FIRST and comparable to filtered backprojection. Quantitative studies confirmed significant high-contrast spatial resolution improvements (e.g., >10 lp/cm across AiCE regions) at equivalent or dose-neutral levels compared to conventional reconstruction methods. Clinical image reviews affirmed the diagnostic quality of images produced by the device across various anatomical regions, including low-dose chest scans and studies of diseased patients. The system also showed improved performance for lung cancer screening applications.

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 retraining of the AiCE DCNN for lung and cardiac applications specifically address the unique imaging challenges (e.g., motion artifacts, tissue density variations) in these anatomical regions compared to abdominal/pelvic imaging?; What is the quantitative impact of AiCE on low-contrast detectability in the newly added lung and cardiac applications, especially at reduced radiation doses, beyond the stated 'equivalent LCD'?; Could the 'more natural noise texture' achieved by AiCE compared to FIRST and MBIR translate to improved diagnostic confidence or reduced interpretation time for radiologists, and are there studies planned to assess this?; Given the significant improvement in high-contrast spatial resolution (e.g., 16.5 lp/cm for Body HR), what are the practical implications for visualizing finer anatomical details or smaller pathologies, particularly in challenging cases?; With the console now integrating the AiCE/FIRST processing (CCRS-003B), what are the improvements in workflow efficiency, reconstruction times, and data handling compared to the previous separate I-REC BOX system?