Artificial intelligence iterative reconstruction for improving image quality of 80 kVp low-dose CT enterography for Crohn disease
10.13929/j.issn.1003-3289.2025.04.006
- VernacularTitle:深度学习全模型迭代算法改善80kVp低剂量小肠CT造影显示克罗恩病图像质量
- Author:
Rui GUO
1
;
Wanhui ZHOU
;
Pengzhi HU
;
Song PENG
;
Qi LIANG
;
Pengfei RONG
Author Information
1. 中南大学湘雅三医院放射科,湖南 长沙 410013
- Publication Type:Journal Article
- Keywords:
Crohn disease;
intestine,small;
tomography,X-ray computed;
artificial intelligence
- From:
Chinese Journal of Medical Imaging Technology
2025;41(4):535-538
- CountryChina
- Language:Chinese
-
Abstract:
Objective To observe the value of artificial intelligence iterative reconstruction(AIIR)for improving image quality of 80 kVp low-dose CT enterography(CTE)for Crohn disease(CD).Methods Totally 59 patients with CD who underwent 80 kVp low-dose CTE were retrospectively enrolled,and CTE images were reconstructed with hybrid iterative reconstruction(HIR)(HIR group)and AIIR(AIIR group),respectively.Then subjective and objective scores of image quality were evaluated and compared between groups,and the value of AIIR was analyzed.Results Compared with HIR group,AIIR group clearly displayed the intestinal wall,intestinal lumen,mesenteric vessels and peri-intestinal soft tissue,and displayed typical CD signs more obviously,with higher subjective scores of imaging quality(all P＜0.001).The standard deviation(SD)was lower,while signal-to-noise ratio(SNR)of intestinal walls without disease,SNR and contrast-to-noise ratio(CNR)of diseased intestinal walls in AIIR group were all higher than those in HIR group(all P＜0.001).The effective dose of 80 kVp CTE examination was(2.82±0.54)mSv.Conclusion AIIR was helpful for improving image quality of 80 kVp low-dose CTE for CD.
