摘要
Abstract
Objective:To investigate the clinical value of the artificial intelligence iterative reconstruction(AIIR)algorithm and to determine the optimal reconstruction level in low-tube voltage CT enterography.Methods:A total of 56 patients who underwent 80 kV CT enterography were prospectively enrolled.And the CT images were divided into two groups according to the reconstruction protocol.Images in Group A were reconstructed using Karl 3D hybrid iterative reconstruction,whereas images in Group B were reconstructed using AIIR at levels 1-5,yielding five image sets(B1—B5).The CT values and noise of the abdominal aorta,ileal arteries,duodenum wall,jejunum wall,ileum wall and psoas musclel were measured.SNR and CNR were compared among the six groups.Subjective image quality was independently evaluated by two radiologists.Results:The image noise,SNR and CNR among the six groups had statistical differences(all P＜0.001),and as the AIIR reconstruction level increased in Group B,image noise increased,whereas SNR and CNR decreased,with statistically significant differences(all P＜0.05).The subjective image quality scores assigned by the two radiologists were highest in subgroup B5,with the statistical differences compared with other groups(all P＜0.05).The subjective scores given by the two radiologists for each group of images showed good consistency(K=0.582-0.864,all P＜0.01).Conclusions:Low-kV CT enterography combined with the AIIR algorithm can effectively reduce image noise and increase SNR and CNR.For low-dose CT enterography,AIIR level 5 reconstruction is recommended.关键词
小肠造影/深度学习/人工智能/图像重建/辐射剂量/体层摄影术,X线计算机Key words
Enteroclysis/Deep learning/Artificial intelligence/Image reconstruction/Radiation dosage/Tomography,X-ray computed分类
医药卫生
