容积CT数字减影血管造影与人工智能技术对头颈CTA减影效果的比较

Comparison of Volume Computed Tomographic Digital Subtraction Angiography and Artificial Intelligence Technology in the Subtraction Effect of the Head and Neck CTA

目的研究容积CT数字减影血管造影(Volume Computed Tomographic Digital Subtraction Angiography,VCTDSA)与人工智能技术(Artificial Intelligence Technology,AI)在正常人头颈CTA的减影效果差异。方法回顾性分析2018年1月至8月我院70名行头颈CTA检查的正常人的影像资料。同一操作者分别使用VCTDSA法与AI对其原始图像进行后处理,计算其减影所需时间及辐射剂量,并对减影后的图像进行质量评价。结果减影时间:VCTDSA法为342(257,441)s,AI为(3.17±0.49)s,差异有统计学意义(P<0.001)。AI的辐射计量明显小于VCTDSA法的。在颈总动脉和颈内动脉颅内段以及椎动脉颅外段的显示上,AI优于VCTDSA法,在上矢状窦、横窦以及乙状窦的显示上,VCTDSA法优于AI,差异均有统计学意义(P<0.05)。结论对头颈CTA的减影后处理,AI更省时省力。若仅对颅内动脉有需求,AI可做优先选择;若对颅内动静脉均有需求,则需选择VCTDSA。

Objective To comparison of the subtraction effect of volume computed tomographic digital subtraction angiography(VCTDSA)and artificial intelligence technology(AI)in the head and neck CTA of healthy individuals.Methods From January to August 2018,70 healthy individuals who underwent head and neck CTA examinations in our hospital were analyzed retrospectively.The same operator used VCTDSA and AI to postprocessing the original image.Time needed and radiant quantity for subtraction was calculated,and the quality of picture after subtraction was evaluated.Results Subtraction time:VCTDSA method was 342(257,441)s,AI was(3.17±0.49)s.The difference between the two was statistically significant(P<0.001).The radiant quantity of AI was less than VCTDSA.AI was better than VCTDSA in the display of the common carotid artery and internal carotid artery intracranial segment and extracranial vertebra arteries.VCTDSA was superior to the AI in the display of the sagittal sinus and transverse sinus and sigmoid sinus,the difference had statistical significance(P<0.05).Conclusion For postprocessing of head and neck CTA,AI saves time and effort.If only the intracranial arteries showed is demand,AI can do preference.If the display has the demand of intracranial arteriovenous,VCTDSA need to choose.