Cascaded Convolutional Neural Networks with Perceptual Loss for Low Dose CT Denoising

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Ataei, Sepehr, Alirezaie, Javad, Babyn, Paul
Natura: Preprint
Pubblicazione: 2020
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914932315914240
author Ataei, Sepehr
Alirezaie, Javad
Babyn, Paul
author_facet Ataei, Sepehr
Alirezaie, Javad
Babyn, Paul
contents Low Dose CT Denoising research aims to reduce the risks of radiation exposure to patients. Recently researchers have used deep learning to denoise low dose CT images with promising results. However, approaches that use mean-squared-error (MSE) tend to over smooth the image resulting in loss of fine structural details in low contrast regions of the image. These regions are often crucial for diagnosis and must be preserved in order for Low dose CT to be used effectively in practice. In this work we use a cascade of two neural networks, the first of which aims to reconstruct normal dose CT from low dose CT by minimizing perceptual loss, and the second which predicts the difference between the ground truth and prediction from the perceptual loss network. We show that our method outperforms related works and more effectively reconstructs fine structural details in low contrast regions of the image.
format Preprint
id arxiv_https___arxiv_org_abs_2006_14738
institution arXiv
publishDate 2020
record_format arxiv
spellingShingle Cascaded Convolutional Neural Networks with Perceptual Loss for Low Dose CT Denoising
Ataei, Sepehr
Alirezaie, Javad
Babyn, Paul
Image and Video Processing
Computer Vision and Pattern Recognition
Low Dose CT Denoising research aims to reduce the risks of radiation exposure to patients. Recently researchers have used deep learning to denoise low dose CT images with promising results. However, approaches that use mean-squared-error (MSE) tend to over smooth the image resulting in loss of fine structural details in low contrast regions of the image. These regions are often crucial for diagnosis and must be preserved in order for Low dose CT to be used effectively in practice. In this work we use a cascade of two neural networks, the first of which aims to reconstruct normal dose CT from low dose CT by minimizing perceptual loss, and the second which predicts the difference between the ground truth and prediction from the perceptual loss network. We show that our method outperforms related works and more effectively reconstructs fine structural details in low contrast regions of the image.
title Cascaded Convolutional Neural Networks with Perceptual Loss for Low Dose CT Denoising
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2006.14738