Imaging Technology
Study on the effect of virtual grid on chest X-ray image quality
Kang Tianliang, Liu Yunfu, Zhang Yongxian, Guo Senlin, Ma Wentao, Niu Yantao
Published 2023-05-10
Cite as Chin J Radiol, 2023, 57(5): 547-552. DOI: 10.3760/cma.j.cn112149-20230213-00094
Abstract
ObjectiveTo explore the image quality and its evaluation method using virtual grid under different tube voltages in the clinical chest X-ray exam.
MethodsAccording to the conditions of chest X-ray photography commonly used in clinical practice, the corresponding thickness of plexiglass (20 cm, including CDRAD phantom) was determined as the experimental object. With a fixed tube loading of 4 mAs and the tube voltage from 60 to 125 kV, the experimental object was imaged in three ways: physical grid, none grid and virtual grid. The common physical parameters (CNR, σ, C, SNR), texture analysis (Angular second moment, texture Contrast, Correlation, Inverse difference moment, Entropy) and CDRAD phantom score (IQFinv) were evaluated. Two-way ANOVA test was used for each group of common physical parameters, and further pairwise comparisons were made. At the same time, applying virtual grids on the obtained images with chest anthropomorphic model and texture indexing the images with and without virtual grids, then rank sum test of paired sample can be conducted.
ResultsThere were differences in image quality among the three groups of grid mode(P<0.05), and the physical grid delivered the best image quality. The tube voltage had an impact on all image quality evaluation indexes (P<0.05). The tube voltage was positively correlated with CNR, SNR, angular second moment, inverse difference moment and IQFinv (P<0.05), and negatively correlated with σ, C, texture contrast and entropy (P<0.05). There was no significant correlation between the tube voltage and Correlation (P>0.05). The chest anthropomorphic model images were used to evaluate the virtual grids, and the texture indexes (Angle second moment, Contrast, Correlation, Inverse difference moment, Entropy) were statistically significant (P<0.05).
ConclusionsThe virtual grid can improve the image quality of chest X-ray photography, and the image texture analysis method can be a useful supplement to the image quality evaluation parameters.
Key words:
Radiography, thoracic; Grid; Virtual grid; Image quality; Texture analysis
Contributor Information
Kang Tianliang
Department of Radiology, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China
Liu Yunfu
Department of Radiology, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China
Zhang Yongxian
Department of Radiology, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China
Guo Senlin
Department of Radiology, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China
Ma Wentao
Department of Radiology, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China
Niu Yantao
Department of Radiology, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China
