eduzhai > Applied Sciences > Engineering >

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

  • Save

... pages left unread,continue reading

Document pages: 5 pages

Abstract: Low Dose CT Denoising research aims to reduce the risks of radiation exposureto patients. Recently researchers have used deep learning to denoise low doseCT images with promising results. However, approaches that usemean-squared-error (MSE) tend to over smooth the image resulting in loss offine structural details in low contrast regions of the image. These regions areoften crucial for diagnosis and must be preserved in order for Low dose CT tobe used effectively in practice. In this work we use a cascade of two neuralnetworks, the first of which aims to reconstruct normal dose CT from low doseCT by minimizing perceptual loss, and the second which predicts the differencebetween the ground truth and prediction from the perceptual loss network. Weshow that our method outperforms related works and more effectivelyreconstructs fine structural details in low contrast regions of the image.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×