摘要
目的探索人工智能辅助压缩感知(artificial intelligence-assisted compressed sensing,ACS)技术加速颅内血管壁高分辨率三维T1加权黑血成像(three-dimensional T1-weighted black blood imaging,3D T1_HRVWI)的可行性,并与临床常规采用的并行采集技术(parallel imaging,PI)进行对比分析。方法前瞻性纳入47例脑血管疾病患者,行ACS加速(实验组)和PI加速(对照组)的3D T1_HRVWI扫描。客观比较ACS组和PI组的序列扫描时长、图像中血管壁的信噪比以及相对于血管腔的对比噪声比,主观评估两组图像总体质量、颅内血管显示情况及病灶诊断价值,并进行统计分析。结果相较于PI 3D T1_HRVWI,ACS 3D T1_HRVWI的扫描时长缩短43%;ACS 3D T1_HRVWI图像中颈内动脉、基底动脉、椎动脉、大脑前动脉、大脑中动脉、大脑后动脉血管壁的信噪比均显著升高(P<0.001);两组图像的总体质量评分和颅内血管的显示评分均未见显著差异(P>0.05);两组图像的病灶诊断价值相当(P>0.05)。结论ACS技术实现了3D T1_HRVWI序列更快的采集速度,同时保持了较好的成像质量,为开发兼顾扫描效率与诊断精度的脑血管影像扫描方案提供了循证依据,在脑血管疾病的精准诊断中展现出重要价值。
Objective To explore the feasibility of artificial intelligent-assisted compressed sensing(ACS)technology in accelerating high-resolution three-dimensional T1-weighted black blood imaging(3D T1_HRVWI)of intracranial blood vessel walls,in comparison with the parallel imaging(PI)technique.Methods 47 patients with cerebrovascular diseases were prospectively enrolled in this study.ACS accelerated and PI accelerated 3D T1_HRVWI scanning were both performed.The scanning duration,signal-to-noise ratio(SNR)of vessel walls in 3D T1_HRVWI images and the contrast noise ratio(CNR)of vessel walls relative to vascular lumens were objectively compared between the ACS group and the PI group.Also,the overall quality of 3D T1_HRVWI images,intracranial artery visualization and diagnostic value of lesions were subjectively evaluated.Statistical analysis was performed.Results Compared with PI 3D T1_HRVWI,the scanning time of ACS 3D T1_HRVWI was shortened by 43%.In ACS 3D T1_HRVWI images,the SNR of internal carotid artery,basilar artery,vertebral artery,anterior cerebral artery,middle cerebral artery and posterior cerebral artery were significantly higher(P<0.001).There were no significant differences on scores for overall quality and intracranial artery visualization betweenthe two groups(P>0.05).The diagnostic value of the two sequences was equal(P>0.05).Conclusions Compared with PI,ACS can shorten the acquisition time of 3D T1_HRVWI while ensuring the image quality,which provides an evidence-based basis for developing neurovascular imaging scanning protocols and shows important value in the accurate diagnosis of neurovascular diseases.
作者
张月青
赵元宾
张瑞
周鑫
苗成鹏
谭桂蓉
段云云
隋滨滨
董健
柴丽
丁金立
ZHANG Yueqing;ZHAO Yuanbin;ZHANG Rui;ZHOU Xin;MIAO Chengpeng;TAN Guirong;DUAN Yunyun;SUI Binbin;DONG Jian;CHAI Li;DING Jinli(Department of Radiology,Beijing Tiantan Hospital,Capital Medical University,Beijing 100070;College of Chemistry and Life Science,Beijing University of Technology,Beijing 100124;Tiantan Neuroimaging Center of Excellence,China National Clinical Research Center for Neurological Diseases,Beijing Neurosurgical Institute,Beijing Tiantan Hospital,Capital Medical University,Beijing 100070)
出处
《北京生物医学工程》
2025年第5期496-502,共7页
Beijing Biomedical Engineering
基金
国家自然科学基金(62271061)
北京市自然科学基金-海淀原始创新联合基金(L232130)资助。
关键词
颅内血管壁成像
高分辨率3D
T1加权黑血成像
脑血管疾病
人工智能辅助压缩感知技术
磁共振成像
intracranial vascular wall imaging
high resolution 3D T1-weighted black blood imaging
artificial intelligence-assisted compressed sensing
cerebrovascular diseases
magnetic resonance imaging
