首页|智能光梭成像在膝关节磁共振检查中的应用价值

智能光梭成像在膝关节磁共振检查中的应用价值

扫码查看
目的:探讨智能光梭成像(AI-assisted compressed sensing,ACS)技术在膝关节i MRI中应用的可行性及优势.方法:纳入华中科技大学同济医学院附属同济医院2021至2022年36例行膝关节MRI检查的参与者作为研究对象,每例参与者均采集膝关节常规对照组图像及ACS组图像.记录2组各序列的图像采集时间.于膝关节冠状位PD图像上选取骨髓、肌肉、前交叉韧带、后交叉韧带作为感兴趣区,计算2组相应感兴趣区的信噪比(signal to noise ratio,SNR)及对比噪声比(contrast to noise ratio,CNR),采用Wilcoxon带符号秩检验比较对照组图像与ACS组图像的客观图像质量.2位诊断医生对图像进行主观评分,采用Kappa检验评价2名医师主观评分结果的一致性.结果:膝关节磁共振图像采集总时间由对照组的461 s缩短至ACS组的245 s,扫描时间缩短了 46.9%;2组兴趣区(骨髓、前交叉韧带、后交叉韧带)的SNR及CNR差异无统计学意义(P>0.05);2名医师主观评分一致性较好(KappaACS组=0.79,Kappa对照组=0.72),组间主观评分差异无统计学意义(P>0.05).结论:ACS技术在膝关节MRI中具有良好的应用价值,在保证图像质量的前提下能显著缩短了图像采集时间,为快速MRI检查提供了新的思路.
The application value of AI-assisted compressed sensing in knee joint MRI
Objective:To evaluate the feasibility and the superiority of AI-assisted compressed sensing(ACS)in knee joint MRI.Methods:Thirty-six participants who underwent both the conventional scanning and ACS scanning of knee joint MRI in Tongji Hospital,Tongji Medical College,Huazhong University of Science and Technology during the period of 2021 to 2022 were included in this study.The acquisition time of each scanning sequence,selected the bone marrow,muscle,anterior cruciate ligament and posterior cruciate ligament as regions of interest on coronal PD images,and calculated the signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)of these region of interst(ROI)were recorded.We evaluated the subjective image quality,and compared the objective image quality.The Wilcoxon signed-rank test was used to compare the objective image quality,and the kappa coefficient was used to statistically analyze the subjective scores.Results:Compared with conventional group,the total acquisition time of knee joint MRI in ACS group was shortened from 461 s to 245 s.There was no significant difference in SNR and CNR of bone marrow,anterior cruciate ligament and posterior cruciate ligament between the two groups(P>0.05).And there was no significant difference in the subjective scores of the two groups(P>0.05).In addition,the subjective measurement result of the two radiologists was in good agreement(Kappaacs=0.79,Kappacontrol=0.72).Conclusion:AI-assisted compressed sensing can significantly shorten the acquisition time without influencing the image quality,which has great potential in knee joint MRI.This technology can provide new ideas for rapid magnetic resonance imaging.

knee jointmagnetic resonance imagingartificial intelligencedeep learning

聂鸿雁、赵延洁、邢文、蔡威

展开 >

华中科技大学同济医学院附属同济医院放射科,湖北武汉 430030

膝关节 磁共振成像 人工智能 深度学习

2024

暨南大学学报(自然科学与医学版)
暨南大学

暨南大学学报(自然科学与医学版)

CSTPCD北大核心
影响因子:0.996
ISSN:1000-9965
年,卷(期):2024.45(6)