US-based Canon Medical Systems has introduced deep convolutional neural network (DCNN) image reconstruction technology of CT scans.
Canon Medical’s Advanced Intelligent Clear-IQ Engine (AiCE), which yet to get 510(k) clearance, makes use of deep learning technology to differentiate signals from noise generated during CT image reconstruction. It can suppress noise while boosting signal.
The algorithm, with the ability to learn from the high image quality of Model-Based Iterative Reconstruction (MBIR), facilitates in reconstructing CT images with improved spatial resolution. This process is 3-5x times faster than traditional Model-Based Iterative Reconstruction (MBIR), thereby ushering a new era for CT image reconstruction.
Equipped with AiCE’s deep learning approach, several features learned during training help differentiate signal from noise in order to provide improved resolution.
AiCE makes use of a pre-trained DCNN to improve spatial resolution, whilst reducing noise with reconstruction speeds in order to quickly cater to busy clinical environments.
Canon Medical Systems CT, PET/CT, and MR business units senior director Dominic Smith said: “As a leader in deep learning reconstruction technology for CT images, Canon Medical is committed to forging new ground for CT imaging in order to meet our customers’ evolving needs. With AiCE technology, we haven’t just raised the bar, we’ve set a new standard for image reconstruction in CT.”
Canon Medical intends to exhibit the AiCE technology at Radiological Society of North America (RSNA) meeting to be held in Chicago from 25 to 30 November.
Headquartered in Tustin, California, Canon Medical Systems markets, sells and distributes radiology and cardiovascular systems, including ultrasound, CT, X-ray, MR, and interventional X-ray equipment.