Revolution Apex CT system
FDA clears GE’s AI-based CT image reconstruction technology
April 25, 2019
by Lisa Chamoff
, Contributing Reporter
GE Healthcare has received FDA clearance for the Deep Learning Image Reconstruction engine using artificial intelligence (AI) for use on its new Revolution Apex CT device and as an upgrade to its Revolution CT devices in the U.S. The company unveiled Revolution Apex at last year’s RSNA as part of its new Edison platform.
The application, which is available as an upgrade for the Revolution Apex CT system, uses algorithms based on previously reconstructed images, resulting in what the company calls TrueFidelity CT images — clearer, low-dose images with improved quality and more natural noise texture.
“One of the challenges that we have when we’re talking about reconstruction has been the look and feel of the image quality,” Saad Sirohey, GE’s general manager of global MICT clinical applications, told HCB News. “What we’re able to reproduce with the deep learning reconstruction is giving (clinicians) the same image texture they’ve been used to and using this technique to reduce the noise and improve the image quality.”
Sonia Sahney, global product marketing director for GE’s premium CT segment, said the TrueFidelity images will require no additional hardware and will have no impact on the reconstruction time.
“It will just fit into clinical workflow,” Sahney told HCB News.
This is the first CT application on the Edison platform, designed to harness machine learning to improve efficiency and patient care, that has been FDA cleared.
Three other GE products — Bone VCAR, SnapShot Freeze 2 and Thoracic VCAR with GSI Pulmonary Perfusion — have also received FDA 510(k) clearance.
Bone VCAR uses a deep-learning algorithm to automatically identify and label the vertebrae in a spine image, while SnapShot Freeze 2 is a heart motion correction algorithm for cardiac imaging. Both are Edison applications.
Thoracic VCAR with GSI Pulmonary Perfusion automates segmentation and measurements in assessing and following up on thoracic diseases, according to the company. It also targets pulmonary perfusion by identifying areas that have relative deficits of perfusion in the lungs, for use in diagnosing issues such as pulmonary embolism or chronic obstructive pulmonary disease (COPD).