por
John R. Fischer, Senior Reporter | October 23, 2019
Advanced Intelligent Clear-IQ Engine (AiCE) is now
integrated on the Aquilion Precision CT System
by Canon
The FDA has given Canon Medical Systems USA the go-ahead to integrate and sell its AI-based image reconstruction technology as a component of the Aquilion Precision CT system.
Named Advanced Intelligent Clear-IQ Engine (AiCE), the deep convolutional neural network enables the scanner to suppress noise and enhance signal to produce high quality images at lower doses.
"AiCE integrates seamlessly into routine practice," Dhruv Mehta, senior manager of solutions marketing for CT at Canon Medical Systems, told HCB News. "It offers comparable overall study times relative to Hybrid-IR approaches used in routine practice; can be built into the scan protocols so they are automatically used, and images transferred are consistent with the current standard-of-care; and radiation doses can be set consistent with the standard-of-care protocols since clinical testing demonstrates dose neutrality between the Aquilion Precision super‐high resolution mode (with AiCE) and normal resolution mode (with AIDR)."
Ad Statistics
Times Displayed: 12403
Times Visited: 19 Final days to save an extra 10% on Imaging, Ultrasound, and Biomed parts web prices.* Unlimited use now through September 30 with code AANIV10 (*certain restrictions apply)
Aquilion Precision is the world’s first ultra-high resolution CT scanner and provides users with two times the resolution of a conventional CT platform, revealing details that can usually only be observed in cath labs. The addition of AiCE raises its performance to levels of super-high resolution at doses equivalent to standard resolution CT with traditional hybrid iterative reconstruction techniques.
The basis of learning for the algorithm technology is the high image quality of Model-Based Iterative Reconstruction (MBIR) to reconstruct CT images with improved high contrast spatial resolution. Its rate of reconstruction, however, is three to five times faster than conventional MBIR. It utilizes thousands of features learned during training to distinguish signal from noise for improved resolution, and applies a pre-trained DCNN to enhance spatial resolution and reduce noise at the same time. This allows for reconstruction to take place at a pace fast enough for busy clinical environments.
The integration of AiCE also solves the typical dilemma of iterative reconstruction techniques, which increases quantum noise levels by decreasing the voxel dimensions of a CT image, according to John Boone, former president and current CT subcommittee chair for the American Association of Physicists in Medicine.
“I was frankly impressed to find that this manufacturer has also developed CT reconstruction methods designed to reduce noise based upon artificial intelligence techniques — their deep learning reconstruction algorithm,” he
told HCB News last month. “I’m very excited to see how the DLR algorithm reduces noise in the high-resolution images … for many applications, we were waiting for the DLR algorithm in order to fully assess the system clinically, and we now have our sleeves rolled up to do so.”
The system was
initially cleared for use by the FDA last December on Canon’s Aquilion ONE / GENESIS Edition premium CT during the annual meeting of Radiological Society of North America.
Clinical collaborators are currently investigating the benefits of AiCE in other applications.