Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Cerebral perfusion computed tomography deconvolution via structure tensor total variation regularization

Journal Article · · Medical Physics
DOI:https://doi.org/10.1118/1.4944866· OSTI ID:22620871
; ; ; ; ; ; ; ; ;  [1];  [2]
  1. Department of Biomedical Engineering, Southern Medical University, Guangzhou, Guangdong 510515, China and Guangdong Provincial Key Laboratory of Medical Image Processing, Southern Medical University, Guangzhou, Guangdong 510515 (China)
  2. Department of Radiology, Tianjin Medical University General Hospital, Tianjin 300052 (China)

Purpose: Cerebral perfusion computed tomography (PCT) imaging as an accurate and fast acute ischemic stroke examination has been widely used in clinic. Meanwhile, a major drawback of PCT imaging is the high radiation dose due to its dynamic scan protocol. The purpose of this work is to develop a robust perfusion deconvolution approach via structure tensor total variation (STV) regularization (PD-STV) for estimating an accurate residue function in PCT imaging with the low-milliampere-seconds (low-mAs) data acquisition. Methods: Besides modeling the spatio-temporal structure information of PCT data, the STV regularization of the present PD-STV approach can utilize the higher order derivatives of the residue function to enhance denoising performance. To minimize the objective function, the authors propose an effective iterative algorithm with a shrinkage/thresholding scheme. A simulation study on a digital brain perfusion phantom and a clinical study on an old infarction patient were conducted to validate and evaluate the performance of the present PD-STV approach. Results: In the digital phantom study, visual inspection and quantitative metrics (i.e., the normalized mean square error, the peak signal-to-noise ratio, and the universal quality index) assessments demonstrated that the PD-STV approach outperformed other existing approaches in terms of the performance of noise-induced artifacts reduction and accurate perfusion hemodynamic maps (PHM) estimation. In the patient data study, the present PD-STV approach could yield accurate PHM estimation with several noticeable gains over other existing approaches in terms of visual inspection and correlation analysis. Conclusions: This study demonstrated the feasibility and efficacy of the present PD-STV approach in utilizing STV regularization to improve the accuracy of residue function estimation of cerebral PCT imaging in the case of low-mAs.

OSTI ID:
22620871
Journal Information:
Medical Physics, Journal Name: Medical Physics Journal Issue: 5 Vol. 43; ISSN 0094-2405; ISSN MPHYA6
Country of Publication:
United States
Language:
English

Similar Records

Deblurring of breathing motion artifacts in thoracic PET images by deconvolution methods
Journal Article · Sun Oct 15 00:00:00 EDT 2006 · Medical Physics · OSTI ID:20853580

Low-dose computed tomography image restoration using previous normal-dose scan
Journal Article · Sat Oct 15 00:00:00 EDT 2011 · Medical Physics · OSTI ID:22098650

Bayesian SPECT lung imaging for visualization and quantification of pulmonary perfusion
Journal Article · Mon Nov 30 23:00:00 EST 1998 · IEEE Transactions on Nuclear Science · OSTI ID:323711