ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1909.01394
23
30

A Novel Loss Function Incorporating Imaging Acquisition Physics for PET Attenuation Map Generation using Deep Learning

3 September 2019
Luyao Shi
J. Onofrey
E. Revilla
T. Toyonaga
David Menard
Johnny S. Ankrah
R. Carson
Chi Liu
Yihuan Lu
    MedIm
ArXiv (abs)PDFHTML
Abstract

In PET/CT imaging, CT is used for PET attenuation correction (AC). Mismatch between CT and PET due to patient body motion results in AC artifacts. In addition, artifact caused by metal, beam-hardening and count-starving in CT itself also introduces inaccurate AC for PET. Maximum likelihood reconstruction of activity and attenuation (MLAA) was proposed to solve those issues by simultaneously reconstructing tracer activity (λ\lambdaλ-MLAA) and attenuation map (μ\muμ-MLAA) based on the PET raw data only. However, μ\muμ-MLAA suffers from high noise and λ\lambdaλ-MLAA suffers from large bias as compared to the reconstruction using the CT-based attenuation map (μ\muμ-CT). Recently, a convolutional neural network (CNN) was applied to predict the CT attenuation map (μ\muμ-CNN) from λ\lambdaλ-MLAA and μ\muμ-MLAA, in which an image-domain loss (IM-loss) function between the μ\muμ-CNN and the ground truth μ\muμ-CT was used. However, IM-loss does not directly measure the AC errors according to the PET attenuation physics, where the line-integral projection of the attenuation map (μ\muμ) along the path of the two annihilation events, instead of the μ\muμ itself, is used for AC. Therefore, a network trained with the IM-loss may yield suboptimal performance in the μ\muμ generation. Here, we propose a novel line-integral projection loss (LIP-loss) function that incorporates the PET attenuation physics for μ\muμ generation. Eighty training and twenty testing datasets of whole-body 18F-FDG PET and paired ground truth μ\muμ-CT were used. Quantitative evaluations showed that the model trained with the additional LIP-loss was able to significantly outperform the model trained solely based on the IM-loss function.

View on arXiv
Comments on this paper