Improvement of diagnostic performance of hyperacute ischemic stroke in head CT using an image-based noise reduction technique with non-black-boxed process

To date, a variety of image-based noise reduction (INR) techniques for computed tomography (CT) have been reported. They are based on iterative processes or deep learning techniques. Some INR products using iterative processes are already commercially available and have been reported to exhibit noise reduction performance better than hybrid-type iterative reconstruction (HIR) in abdominal CT [1,2]. However, the details of the processes inside these commercial INR systems are usually proprietary and cannot be readily replicated outside.

Source: Physica Medica: European Journal of Medical Physics - August 5, 2023 Category: General Medicine Authors: Tetsuya Hirairi, Katsuhiro Ichikawa, Atsushi Urikura, Hiroki Kawashima, Takasumi Tabata, Tamaki Matsunami Source Type: research