人工智能斑块定量分析在冠心病中的预测价值  

The predictive value of artificial intelligence plaque quantitative analysis in coronary heart disease

作　　者：罗焕 高律萍 曹成瑛 杨陵武 朱友义 LUO Huan;GAO Lüping;CAO Chengying;YANG Lingwu;ZHU Youyi(Department of Radiology,Qinghai Province Cardiovascular and Cerebrovascular Diseases Specialist Hospital,Xining 810012,China)

机构地区：[1]青海省心脑血管病专科医院放射科,青海西宁810012

出　　处：《实用放射学杂志》2024年第6期898-902,共5页Journal of Practical Radiology

基　　金：青海省卫生健康委员会卫生科研课题项目(2022-WJZDX-53)。

摘　　要：目的探讨人工智能(AI)斑块定量分析联合CT血流储备分数(CT-FFR)、冠状动脉周围脂肪衰减指数(FAI)对青海地区冠心病(CHD)的预测价值。方法选取118例疑似CHD患者,根据其血管狭窄率是否>50%分为狭窄组(76例)和非狭窄组(42例),测定并对比2组患者斑块体积、负荷、CT-FFR值和冠状动脉周围FAI,分析其在CHD中的预测价值。结果狭窄组斑块总体积、钙化及非钙化斑块体积、脂质斑块体积及对应的体积负荷均显著高于非狭窄组(P<0.05),餐巾环征百分比和冠状动脉周围FAI显著高于非狭窄组,CT-FFR显著低于非狭窄组(P<0.05)。logistic回归分析发现,钙化斑块及脂质斑块的体积负荷、餐巾环征、CT-FFR、冠状动脉周围FAI为CHD独立危险因素(P<0.05)。AI斑块定量分析与CT-FFR、冠状动脉周围FAI联合诊断具有较高的敏感度和特异度,其在CHD诊断方面价值较高。结论AI斑块定量分析联合CT-FFR、冠状动脉周围FAI在CHD诊断方面具有较高的价值,值得临床推广应用。Objective To explore the predictive value of artificial intelligence(AI)plaque quantitative analysis combined with CT derived fractional flow reserve(CT-FFR)and pericoronary fat attenuation index(FAI)in coronary heart disease(CHD)in Qinghai.Methods A total of 118 suspected CHD patients were selected,and were divided into a stenosis group(n=76)and a non-stenosis group(n=42)based on whether their vascular stenosis rate was>50%.The plaque volume,load,CT-FFR value,and pericoronary FAI of the two groups were measured and compared,and their predictive value in CHD was further analyzed.Results The plaque total volume,calcified and non-calcified plaque volume,lipid plaque volume,and corresponding volume load of the stenosis group were significantly higher than those of the non-stenosis group(P<0.05),the percentage of napkin ring sign and pericoronary FAI were significantly higher than those of the non-stenosis group,and the CT-FFR was significantly lower than that of the non-stenosis group(P<0.05).Logistic regression analysis found that the volume load of calcified and lipid plaques,napkin ring sign,CT-FFR,and pericoronary FAI were independent risk factors for CHD(P<0.05).The combination of AI plaque quantitative analysis,CT-FFR,and pericoronary FAI had high sensitivity and specificity,with high value in CHD diagnosis.Conclusion The combination of AI plaque quantitative analysis,CT-FFR,and pericoronary FAI has high value in the diagnosis of CHD,and may be worthy of clinical promotion application.

关 键 词：钙化斑块 脂质斑块 CT血流储备分数 冠心病 人工智能 

分 类 号：R814.42[医药卫生—影像医学与核医学] R541.4[医药卫生—放射医学]