摘要
目的探究深度学习图像重建(DLIR)在缺血性心肌病CT血管造影(CTA)图像质量评价中的应用。方法选取2020年1月至2023年4月于郑州市第七人民医院就诊的103例缺血性心肌病患者为研究对象,重建出自适应统计迭代重建(ASIR-V)50%和80%,DLIR(低、中、高强度)5组图像。通过比较5组图像的CT值、噪声值、信噪比(SNR)和对比度噪声比(CNR)对不同图像质量进行客观评价,并由2名放射诊断科医生采用双盲法对图像质量进行主观评价。结果5组重建图像的CT值差异无统计学意义(P>0.05),5组图像的噪声值从低到高依次为:DLIR高强度、DLIR中强度、ASIR-V80%、DLIR低强度、ASIR-V50%,差异有统计学意义(P<0.05)。5组图像质量的主观评分差异有统计学意义(P<0.05)。图像质量由低到高依次为ASIR-V50%、DLIR低强度、ASIR-V80%、DLIR中强度、DLIR高强度。结论基于深度学习的图像重建算法能降低噪声,提高缺血性心肌病CTA图像质量。
Objective To investigate the application effect of deep learning image reconstruction(DLIR)in the evaluation of CT angiography(CTA)image quality in ischemic cardiomyopathy.Methods A total of 103 patients with ischemic cardiomyopathy who received treatment at the Seventh People’s Hospital of Zhengzhou from January 2020 to April 2023 were selected as the research subjects.Five sets of images were reconstructed,including adaptive statistical iterative reconstruction-V(ASIR-V)50%and 80%,and DLIR(low,medium,and high intensity).Objective evaluation of different image quality was conducted by comparing CT values,noise values,signal-to-noise ratio(SNR),and contrast to noise ratio(CNR)of 5 sets of images,and subjective evaluation of image quality was conducted by two radiologists using a double-blind method.Results No statistical difference was reported in CT attenuation values(P>0.05).The image noise was the lowest in DLIR-H,followed by DLIR-M,ASIR-V80%and DLIR-L,and was the largest in ASIR-V50%,with statistical difference(P<0.05).Subjective quality rating demonstrated statistical difference among five reconstructions,and the score was the lowest in ASIR-V50%,followed by DLIR-L,ASIR-V80%and DLIR-M,and was the highest in DLIR-H.Conclusion Application of DLIR can effectively reduce the image noise and improve the image quality of CTA image in ischemic cardiomyopathy.
作者
王建
李真真
李静
WANG Jian;LI Zhenzhen;LI Jing(Department of Medical Imaging,Zhengzhou Seventh People’s Hospital,Zhengzhou 450016,China;Department of Endocrinology and Gerontology,Zhengzhou Seventh People’s Hospital,Zhengzhou 450016,China)
出处
《河南医学研究》
2026年第7期1274-1278,共5页
Henan Medical Research
关键词
缺血性心肌病
CT血管造影
深度学习
图像质量
ischemic cardiomyopathy
CT angiography
deep learning image reconstruction
image quality
