Quantification of temporal resolution improvement factor in SMART-RECON based time-resolved C-arm Cone beam computed tomography angiography (TR-CBCTA). (2nd October 2018)

• Record Type: Journal Article
• Authors: Garrett, John W; Li, Yinsheng; Li, Ke; Wu, Yijing; Johnson, Kevin; Schafer, Sebastian; Chen, Guang-Hong
• Is Part Of: Physics in medicine & biology. Volume 63:Number 19(2018:Oct.)
• Publication Date: 2018-10-02
• Subjects: temporal resolution -- image reconstruction -- time-resolved angiography -- Cone beam CT; Biophysics -- Periodicals; Medical physics -- Periodicals; 610.153
• DOI: 10.1088/1361-6560/aadfef
• Languages: English
• ISSNs: 0031-9155

Abstract

Abstract: In a recently published paper by Li et al (2018 Phys. Med. Biol .63 075001), reconstruction parameters were optimized for the so-called synchronized multiArtifact reduction with tomographic reconstruction (SMART-RECON) method to enable time-resolved Cone beam computed tomography angiography (TR-CBCTA) from a single short-scan CBCT data set. However, the paper did not quantitatively address how much temporal resolution can be improved by using the SMART-RECON algorithm in TR-CBCTA. The purpose of this note was to present a method to quantify this temporal resolution improvement factor and to use that method to quantify the improvement in temporal resolution using SMART-RECON and optimized reconstruction parameters. In the method proposed in this note, the potential temporal blurring caused by SMART-RECON was modeled as a convolution between a temporal Gaussian blurring kernel and the ground truth temporal enhancement curve. The width parameter of the resulting Gaussian blurring kernel was used as a surrogate metric for temporal resolution to quantify the achievable temporal resolution for the conventional filtered backprojection (FBP) and SMART-RECON methods. The ratio of the surrogate temporal resolutions for the two reconstruction methods was calculated to quantify the factor of temporal resolution improvement. The quantitative results show that the temporal resolution is improved by a factor of 4.5 using SMART-RECON compared with FBP.