3D-TOF MRA序列筛查脑微出血在脑梗死出血性转化中的预警价值

Cerebral Microbleeds Screening with Original Image of 3D-TOF MRA Sequences Predicting Hemorrhagic Transformation in the Patients with Cerebral Infarction

摘要: 目的：探讨三维时间飞跃法磁共振血管成像(Three dimensional time-of-flight magnetic resonance angiography, 3D-TOF MRA)原始图像评估脑梗死患者脑微出血(Cerebral microbleeds, CMBs)病灶的可行性，进一步探讨3D-TOF序列在预测脑梗死出血性转化中的应用价值。方法：选取192例脑梗死患者为研究对象，分别进行磁敏感加权成像(Susceptibility weighted imaging, SWI)和3D-TOF序列扫描，评估SWI及3D-TOF MRA原始轴位图像显示CMBs的优劣。观察脑梗死出血性转化在CT、T1WI、DWI、SWI及3D-TOF MRA序列原始图像中的影像学表现及CMBs的分布和信号特点，明确基于3D-TOF序列CMBs病灶数量分级与出血性转化的关系，探讨3D-TOF序列在预测脑梗死出血性转化中的应用价值。结果: 因影像学资料显示不清排除12例，实际纳入180例脑梗死患者，SWI共发现微出血病灶264个，CMBs数目分级：28例为轻度，8例为中度，4例为重度。3D-TOF MRA序列原始图像发现微出血病灶188个，其中24例为轻度，4例为中度，4例为重度。两种序列对微出血病灶的检出率比较差异有统计学意义(皮质和皮质下、丘脑、小脑，p < 0.05)；两者对基底节和脑干的微出血病灶检出率差异无统计学意义(p < 0.05)。180例患者中，24例发生出血性转化(13.3%)，T1WI、DWI、SWI及3D-TOF MRA序列检测出血性转化的敏感度分别为25.0%、62.5%、100%及83.3%。CMBs病灶数量与出血性转化的发生率呈正相关。结论：在缺乏SWI序列的情况下，3D-TOF MRA原始图像可有效显示CMBs病灶，可作为预测脑梗死发生出血性转化的有效手段之一。

Abstract: Objective: To investigate the feasibility of the original image of the three-dimensional time of flight MR angiography (Three dimensional time-of-flight magnetic resonance angiography, 3D- TOF MRA) assessing cerebral microbleeds (CMBs) in the patients with cerebral infarction, and further investigate the application value of 3D-TOF sequence in predicting hemorrhagic transformation from cerebral infarction. Methods: 192 cases of patients with cerebral infarction were enrolled into this study, and susceptibility weighted imaging (SWI) and 3D-TOF sequence scanning were performed to assess the ability of SWI and original axial images of 3D-TOF MRA showing CMBs. We investigated the imaging signal distribution of hemorrhagic transformation with cerebral infarction in the original image of CT, T1WI, DWI, SWI and 3D-TOF MRA sequences and the characteristics of CMBs to clarify the relationship between the number grade of lesions and hemorrhagic transformation a clear sequence of CMBs based on 3D-TOF sequence, and explored the application value of 3D-TOF predicting hemorrhagic transformation with cerebral infarction. Results: 12 cases were excluded for the unclear imaging data. SWI found 264 lesions of CMBs among 180 cases with cerebral infarction. Classification of CMBs: 28 cases of mild, 8 cases of moderate, and 4 cases of severe. The original sequences of 3D-TOF MRA found 188 lesions of CMBs including 24 cases of mild, four cases of moderate, and four cases of severe. There was statistically significant difference in the CMBs lesions detection rate between these two sequences (cortical vs. subcortical, thalamus, cerebellum, p < 0.05). There was no significant difference in CMBs lesions detection rate in the basal ganglia and brainstem between them (p < 0.05). Among 180 patients, there were a total of 24 cases of hemorrhagic transformation (13.3%). The detection sensitivities of hemorrhagic transformation of T1WI, DWI, SWI and 3D-TOF MRA sequence were 25.0%, 62.5%, 100% and 83.3%. CMBs lesions number was positively correlated with the incidence of hemorrhagic transformation. Conclusion: In the absence of SWI sequence, the original image of 3D-TOF MRA sequence can effectively display CMBs lesions, which can be used as an effective means of predicting hemorrhagic transformation of cerebral infarction.