Sign In

Communications of the ACM

ACM TechNews

With Deep Learning Algorithms, Standard CT Technology Produces Spectral Images

Analysis of computed tomography scans.

A deep neural network developed by researchers at Rensselaer Polytechnic Institute, Shanghair First-Imaging Tech, and GE Research produced spectral analyses of computed tomography scans that provide more information, resulting in better diagnoses.

Credit: Wenxiang Cong et al.

Engineers at Rensselaer Polytechnic Institute (RPI), China's Shanghai First-Imaging Tech, and GE Research have developed a deep neural network that produces spectral images almost as well as dual-energy computed tomography (CT) imaging technology.

The team applied a deep learning algorithm trained on dual-energy CT images to a regular single-spectrum X-ray CT scan.

The model yielded high-quality approximations, with a relative error of less than 2%.

RPI's Ge Wang said, "We hope that this technique will help extract more information from a regular single-spectrum X-ray CT scan, make it more quantitative, and improve diagnosis."

From RPI School of Engineering
View Full Article


Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA


No entries found