The value of deep learning image reconstruction algorithm to improve the quality of low keV monochromatic portal vein images of energy spectrum CT
10.3969/j.issn.1002-1671.2025.04.028
- VernacularTitle:深度学习图像重建算法改善能谱CT低千伏单能量门静脉图像质量价值研究
- Author:
Li SHEN
1
;
Taiping HE
;
Qian TIAN
;
Nan YU
;
Dong HAN
;
Zhanli REN
;
Yongjun JIA
;
Yangyang YAN
Author Information
1. 陕西中医药大学附属医院影像科,陕西 咸阳 712000
- Publication Type:Journal Article
- Keywords:
energy spectrum CT;
monochromatic imaging;
deep learning;
portal vein;
image quality
- From:
Journal of Practical Radiology
2025;41(4):664-668
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the value of deep learning image reconstruction(DLIR)algorithm to improve the quality of low keV monochromatic portal vein images of energy spectrum CT.Methods Fifty patients who underwent enhanced upper abdominal energy spectrum CT scan were selected.Mixed-model adaptive statistical iterative reconstruction-Veo(50%ASIR-V)algorithm and high-deep learning image reconstruction(DLIR-H)algorithm were used to obtain monochromatic images at 40-70 keV(with intervals of 10 keV).The CT and standard deviation(SD)values of the portal vein trunk,left and right branches,and erector spinae muscle were measured in the transverse position,and the signal-to-noise ratio(SNR)and portal vein contrast-to-noise ratio(CNR)were calculated for objective evaluation.The portal vein image quality between the two algorithms and different energy was subjectively scored by two physicians.Results In terms of objective evaluation:compared with 50%ASIR-V,the CNR and SNR of portal vein in monochromatic DLIR-H images at the same keV between 40-70 keV energy levels were increased while the SD value was decreased(P＜0.05),and the CT value was unchanged;there was no statistical difference in the magnitude of change in CNR between the two algorithms at different energy levels(P＞0.05);there was a statistically significant difference in the magnitude of change in SNR and SD value(P＜0.05)and the magnitude of change was the largest at 40 keV;comparison between different energy levels of DLIR-H,the CNR and SD value of 40 keV DLIR-H were the highest(P＜0.05),and there was no significant difference in the SNR(P＞0.05).In terms of subjective evaluation:there was no significant difference between the subjective scores of the two algorithms at the same keV from 40-70 keV(P＞0.05),and the two reconstruction algorithms at 40 keV and 50 keV had the highest subjective scores between different keV.Conclusion The DLIR algorithm can reduce the noise of low keV monochromatic images,improve the image quality of portal vein.
