Abstract: Background DSA is an irreplaceable diagnostic method for cerebral vascular disease. It is sensitive to intracranial macrovascular disease, but not to distal blood perfusion. In order to quantitatively assess the perfusion area of the feeding artery, this study is the first to propose an imaging method of the internal carotid artery (ICA) perfusion area relying on deconvolution algorithm and cross - correlation (CC) algorithm. Methods Four patients with cerebral vascular disease and 2 healthy subjects who underwent DSA examination in Luwan Branch of Ruijin Hospital Affiliated to Shanghai Jiaotong University School of Medicine from June 2020 to May 2021 were enrolled. The arterial input function (AIF) of the ICA was extracted, and a deconvolution algorithm was used to calculate the time to peak (Tmax) for each pixel. Then the CC algorithm was used to obtain the CC map related to the perfusion of the ICA at each pixel point of the whole brain. Area of interest (ROI) of AIF were randomly selected for 2 healthy subjects, calculated the root mean square error (RMSE) of the CC map for reliability analysis. The CC map was drawn for 4 patients with cerebral vascular disease for validity analysis. Results Reliability analysis showed that the RMSE values of Towne's and lateral views of the CC map and Tmax map of the left and right ICA in a healthy subject were 0.008 ± 0.011, 0.022 ± 0.002, 0.015 ± 0.007, 0.004 ± 0.008, and (0.108 ± 0.181) s, (0.181 ± 0.214) s, (0.301 ± 0.230) s, (0.035 ± 0.092) s; another healthy subject were 0.015 ± 0.023, 0.007 ± 0.011, 0.007 ± 0.011, 0.005 ± 0.012, and (0.172 ± 0.275) s, (0.092 ± 0.174) s, (0.087 ± 0.156) s, (0.079 ± 0.153) s. The averaged RMSE value of Tmax map was lower than the temporal resolution of DSA, indicating that the reliability of the CC map of the AIF value box in different ROIs of the ICA was high. Validity analysis showed that the CC map and Tmax map of the patient with acute ischemic stroke treated by intravenous thrombolysis bridging to mechanical thrombectomy can fully display the perfusion changes of the ICA after successful thrombectomy. Postoperative DSA in the patient with subacute ischemic stroke after endovascular therapy showed recanalization of the occluded artery, while the CC map showed no significant change in the perfusion of the ICA. The DSA of the patient with middle cerebral artery (MCA) stenosis without related ischemic events showed severe MCA stenosis, but the CC map showed that the ICA was well perfused, and no endovascular treatment was required. The DSA of the patient with acute ischemic stroke without MCA stenosis showed no MCA stenosis, but only sparse ICA-related perfusion was seen on the CC map. Conclusions The CC map and Tmax map based on DSA images have good reliability in different ROIs of AIF. The CC map can provide quantitative information on the perfusion area of the feeding arteries, which is helpful for accurate diagnosis and appropriate treatment decisions.

Key words: Carotid artery, internal, Cerebrovascular circulation, Angiography, digital subtraction

摘要： 研究背景 DSA是脑血管病不可替代的诊断方法，其对颅内大血管病变较敏感，而对远端灌注信息不全面。为定量评估供血动脉灌注区，本研究首次提出一种依赖去卷积算法和交叉相关（CC）算法的颈内动脉灌注区成像方法。方法 纳入2020年6月至2021年5月在上海交通大学医学院附属瑞金医院卢湾分院行DSA检查的4例脑血管病患者和2例健康受试者，提取颈内动脉输入函数（AIF），采用去卷积算法计算每个像素点达容时间（Tmax），再采用CC算法获得全脑各像素点与颈内动脉灌注相关的CC图。2例健康受试者随机选择AIF兴趣区，计算CC图均方根误差（RMSE），行信度分析； 4例脑血管病患者绘制CC图，行效度分析。结果 信度分析显示，1例健康受试者左和右颈内动脉汤氏位和侧位CC图和Tmax图的RMSE值分别为0.008 ± 0.011、0.022 ± 0.002、0.015 ± 0.007、0.004 ± 0.008和（0.108 ±0.181）s，（0.181 ± 0.214）s、（0.301 ± 0.230）s、（0.035 ± 0.092）s；另1例为0.015 ± 0.023、0.007 ± 0.011、0.007 ± 0.011、0.005 ± 0.012和（0.172 ± 0.275）s、（0.092 ± 0.174）s、（0.087 ± 0.156）s、（0.079 ± 0.153）s；且Tmax图的平均RMSE值低于DSA的时间分辨率，表明颈内动脉兴趣区取值框AIF取值范围的CC图可靠性较高。效度分析显示，1例急性缺血性卒中静脉溶栓桥接机械取栓患者，CC图和Tmax图充分显示成功取栓后颈内动脉相关灌注改变； 1例亚急性缺血性卒中血管内治疗患者，术后常规DSA显示闭塞动脉再通，但CC图显示颈内动脉相关灌注无明显变化； 1例大脑中动脉狭窄但无缺血事件患者，尽管常规DSA显示大脑中动脉重度狭窄，但CC图提示颈内动脉相关灌注良好，无需血管内治疗； 1例急性缺血性卒中但无大脑中动脉狭窄患者，尽管DSA未见大脑中动脉狭窄，但CC图仅见稀疏的颈内动脉相关灌注。结论 基于常规DSA图像的CC图和Tmax图在颈内动脉兴趣区取值框AIF取值范围内具有良好的信度。CC图可提供供血动脉灌注区定量信息，有助于准确诊断和制定适宜的治疗决策。

关键词: 颈内动脉, 脑血管循环, 血管造影术, 数字减影