Cerebral perfusion computed tomography deconvolution via structure tensor total variation regularization. Issue 5 (5th April 2016)

• Record Type: Journal Article
• Authors: Zeng, Dong; Zhang, Xinyu; Bian, Zhaoying; Huang, Jing; Zhang, Hua; Lu, Lijun; Lyu, Wenbing; Zhang, Jing; Feng, Qianjin; Chen, Wufan; Ma, Jianhua
• Is Part Of: Medical physics. Volume 43:Issue 5(2016)
• Page Start: 2091
• Page End: 2107
• Publication Date: 2016-04-05
• Subjects: blood -- brain -- computerised tomography -- data acquisition -- dosimetry -- haemodynamics -- haemorheology -- image denoising -- image enhancement -- iterative methods -- mean square error methods -- medical disorders -- medical image processing -- neurophysiology -- spatiotemporal phenomena; Computed tomography -- General statistical methods -- Blood‐brain barrier -- Stroke -- Noise -- Edge enhancement; Computerised tomographs -- Biological material, e.g. blood, urine; Haemocytometers -- Methods or arrangements for processing data by operating upon the order or content of the data handled -- Data acquisition and logging -- Digital computing or data processing equipment or methods, specially adapted for specific applications -- Image data processing or generation, in general -- Image enhancement or restoration, e.g. from bit‐mapped to bit‐mapped creating a similar image -- Scintigraphy; cerebral perfusion computed tomography -- low‐mAs -- deconvolution -- structure tensor total variation -- regularization; Deconvolution -- Computed tomography -- Medical image noise -- Brain -- Tensor methods -- Medical image reconstruction -- Image reconstruction -- Optical inspection -- Data acquisition; Medical physics -- Periodicals; Medical physics; Geneeskunde; Natuurkunde; Toepassingen; Biophysics; Periodicals; Periodicals; Electronic journals; 610.153
• DOI: 10.1118/1.4944866
• Languages: English
• ISSNs: 0094-2405

Abstract

Abstract : 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.