肩关节加速MRI应用深度学习重建算法的可行性与临床价值  

Feasibility and clinical value of deep learning reconstruction in accelerated MRI of shoulder

作　　者：王璇 王皓 万云天 陈友芳 聂春焰 杜昱升 何健[3] WANG Xuan;WANG Hao;WAN Yuntian;CHEN Youfang;NIE Chunyan;DU Yusheng;HE Jian(Department of Medical Imaging Department,Jiangbei Branch of Zhongda Hospital,Southeast University,Nanjing,Jiangsu 210048,China;不详)

机构地区：[1]东南大学附属中大医院江北院区医学影像科,江苏南京210048 [2]南京医科大学第四附属医院放射科,江苏南京210031 [3]南京医科大学鼓楼临床医学院,江苏南京210008

出　　处：《中国临床研究》2024年第8期1238-1243,共6页Chinese Journal of Clinical Research

基　　金：江苏省“十三五”强卫工程青年医学重点人才(QNRC2016041);南京医科大学科技发展基金一般项目(NMUB20230037)。

摘　　要：目的 探讨深度学习重建算法(DLR)在肩关节MRI中提高图像质量和缩短扫描时间的可行性与临床价值。方法 前瞻性纳入2023年6月至10月期间在南京医科大学第四附属医院的50例疑似患有肩关节病变的患者,采用1.5T MRI行常规序列扫描图像为Fse_(con)组,使用并行采集加速因子2的扫描图像为Fse_(fast)组,扫描序列包括脂肪抑制质子加权像(PDWI-FS)和T1加权像(T1WI),将Fse_(fast)组传至Subtle MR^(TM )dlr后获得图像Fse_(dlr)组。测量三组图像中的冈上肌、肱二头肌长头肌腱、盂唇软骨、肱骨骨髓的信号噪声比(SNR)及冈上肌/盂唇软骨的对比噪声比(CNR)并进行比较,两名放射科医师双盲采用Likert 4分法分别对Fse_(dlr)组与Fse_(con)组的图像清晰度和伪影进行主观评价,并对这两组的病理异常结构进行诊断效能对比。结果 相对于Fse_(con)组,Fse_(dlr)组扫描时间缩短了44%,且图像清晰度评分、伪影评分均增高,差异有统计学意义(P<0.05),两名医师主观评分组内相关性系数为0.797~0.919。客观评价指标中,Fse_(dlr)组的SNR和CNR均明显高于Fse_(con)组与Fse_(fast)组,差异均有统计学意义(P<0.05)。在两位医师对Fse_(con)组与FSE_(dlr)组病理异常结构的评估中,两组的诊断结果均有较好的一致性(Kappa值:0.675~1.000),在同一名医师的评估中也显示出极好的一致性(Kappa值:0.771~1.000),其中肱骨骨髓、关节滑囊、肱二头肌长头肌腱的Kappa值均高于0.8。结论 将DLR算法应用于肩关节MRI检查中,能够提高图像质量、缩短图像采集时间,并保证诊断效能,提高检查效率,具有较好的临床价值。Objective To investigate the feasibility and clinical value of the deep learning reconstruction(DLR)algorithm in enhancing image quality and reducing scan time in shoulder MRI.Methods Fifty patients suspected of having shoulder joint lesions in the Fourth Affiliated Hospital of Nanjing Medical University from June to October 2023 were prospectively included.Routine sequence scanning images with 1.5T MRI were grouped as Fse con,while scanning images with parallel acquisition acceleration factor 2 were grouped as Fse fast.The scanning sequences included fat-suppression proton density-weighted imaging(PDWI-FS)and T1 weighted imaging(T1WI).The Fse fast group was then transferred to Subtle MRTM dlr to obtain images in the Fse dlr group.The signal-to-noise ratio(SNR)of supraspinatus muscle,long head tendon of biceps brachii,glenoid labrum cartilage,and humerus marrow,as well as the contrast-to-noise ratio(CNR)of supraspinatus muscle/glenoid labrum cartilage,were measured and compared among the three groups.Two radiologists blindly evaluated the image clarity and artifacts of the Fse dlr group and the Fse con group using the Likert 4-point scale,and compared the diagnostic efficacy of pathological abnormal structures between the two groups.Results Compared with the Fse con group,the scan time of the Fse dlr group was shortened by 44%,and the image clarity and artifact scores were higher,with statistically significant differences(P<0.05).The intraclass correlation coefficients(ICC)for subjective scores within the two radiologists'groups were 0.797 to 0.919.Among the objective evaluation indicators,the SNR and CNR of the Fse dlr group were significantly higher than those of the Fse con group and the Fse fast group,with statistically significant differences(P<0.05).In the evaluation of the pathological abnormal structures of the Fse con group and the Fse dlr group by two radiologists,the diagnostic results of the two groups were consistent(Kappa value:0.675-1.000),and also showed excellent consistency in the evaluation of

关 键 词：深度学习重建 肩关节 核磁共振 信号噪声比 对比噪声比 图像质量 

分 类 号：R445.2[医药卫生—影像医学与核医学]