摘要
目的:探究低管电压联合人工智能重建技术在头颅CT平扫中的可行性研究。方法:选取2022年8月—11月在济南大学医院因头部不适申请头颅CT检查的62例(男39例,女23例)患者。使用数字随机法分为两组,A组(n=33)采用230 m A s管电流,100 k V管电压进行扫描,并采用滤波反投影法重建;B组(n=29)采用100 m A s管电流,80 k V管电压进行扫描,并采用人工智能技术重建。记录两组患者头颅CT图像灰质CT值、白质CT值、灰质噪声、白质噪声、剂量长度乘积(dose length product,DLP)、体积CT剂量指数(volume CT dosimetry index,CTDIvol)、计算头颅CT图像信噪比(signal-to-noise ratio,SNR)、对比信噪比(contrast-to-noise ration,CNR)、有效剂量(effective dose,ED)以及进行图像评分。结果:A、B两组头颅CT图像灰质CT值、白质CT值、辐射剂量参数差异均显著(P<0.01)。A、B两组头颅CT图像灰质噪声、白质噪声、SNR、CNR、图像评分差异均不具有统计学意义(P>0.05)。结论:低管电压联合人工智能重建技术在头颅CT平扫图像中可改善图像质量,在减少辐射剂量的同时不改变观察者评价,且图像SNR、CNR不变,值得临床推广。
Objective Explore the feasibility of low tube voltage combined with artificial intelligence reconstruction technology in head non-contrast CT.Methods A total of 62 patients(39 male and 23 female)who applied for head CT examination due to head discomfort in Jinan University Hospital from August to November 2022 were selected.The digital random method was used to divide the group into two groups.Group A(n=33)was scanned by 230mAs tube current and 100kV tube voltage,and reconstructed by filtered back projection method;Group B(n=29)was scanned by 100mAs tube current and 80kV tube voltage,and reconstructed by artificial intelligence technology.Gray matter CT value,white matter CT value,gray matter noise,white matter noise,dose length product(DLP),volume CT dosimetry index(CTDIvol),signal-tonoise ratio(SNR),contrast-to-noise ration(CNR)and effective dose(ED)of skull CT images were recorded in the two groups.Results There were significant differences in gray matter CT value,white matter CT value and radiation dose parameters of the head CT images of group A and group B(P<0.01).There was no statistically significant difference in gray matter noise,white matter noise,SNR,CNR and image score of head CT images in group A and group B(P>0.05).Conclusion Low tube voltage combined with artificial intelligence reconstruction technology can improve the image quality in the head non-contrast CT image,reduce the radiation dose without changing the observer's evaluation SNR,CNR remain unchanged,which is worthy of clinical promotion.
作者
景中军
马江凯
JING Zhongjun;MA Jiangkai(Department of Radiology,Hospital of Jinan University,Jinan,Shandong 250022,China)
出处
《影像研究与医学应用》
2023年第4期33-35,共3页
Journal of Imaging Research and Medical Applications
关键词
低管电压
图像质量
人工智能
Low tube voltage
Image Quality
Artificial intelligence