FDA clears 1st deep learning-based CT image reconstruction technology
The FDA granted GE Healthcare’s Deep Learning Image Reconstruction (DLIR) platform 501(k) clearance April 18, marking the first time the agency has approved a deep learning-based CT image reconstruction technology.
According to a release from GE, DLIR uses a deep neural network to generate TrueFidelity CT images—another branded technology the company claims improves CT reading confidence across a range of clinical applications like head, whole body and cardiovascular scans. Compared to run-of-the-mill CT tech, GE said TrueFidelity images have superior image quality performance, image sharpness and noise texture without compromising dose performance.
DLIR was approved by the FDA for use on GE’s new Revolution Apex CT device and as an upgrade to the company’s current Revolution CT system.
“We are proud to usher in the next generation of image reconstruction,” said Mike Barber, president and CEO of MICT, GE Healthcare, in the release. “Our Deep Learning Image Reconstruction engine combines the ground truth image quality of filtered back projection with the low-dose capabilities of iterative reconstruction to produce TrueFidelity CT images. These images offer outstanding image quality and restore noise texture to improve radiologists’ confidence in diagnosing a wide range of clinical cases.”
The FDA simultaneously granted 501(k) clearance to three other GE CT applications, including Bone VCAR, Thoracic VCAR with GSI Pulmonary Perfusion and SnapShot Freeze 2. The trio were first introduced last November at RSNA 2018.
After graduating from Indiana University-Bloomington with a bachelor’s in journalism, Anicka joined TriMed’s Chicago team in 2017 covering cardiology. Close to her heart is long-form journalism, Pilot G-2 pens, dark chocolate and her dog Harper Lee.