双能量钙图/虚拟去钙诊断模型在颅内出血与钙化灶鉴别诊断中的价值

Value of Discrimination Model Based on Dual Energy Calcium/Virtual Non-Calcium in Differentiating Intracranial Hemorrhage and Calcified Foci

目的探讨基于双能量钙图/虚拟去钙图获得的判断模型在区分颅内出血灶与钙化灶中的作用,明确其鉴别诊断的价值。方法回顾性收集我院疑似脑卒中157例患者中等密度病灶189例(训练集94例,测试集95例),利用双能量钙图/虚拟去钙图测量值建立分类诊断模型;并与基于CT值的诊断方法对比,分析新方法在诊断效能方面的优势。同时,按照病灶大小分为不同亚组,分析诊断模型对于病灶大小的稳定性。结果在测试集上,基于CT值的模型AUC曲线下面积为0.817(95%CI:0.724~0.890),而利用钙图/虚拟去钙获得AUC曲线下面积为0.994(95%CI:0.951~1.000);主观评估的明确诊断率分别为82.1%、68.4%,且在<5 mm亚组与>5 mm亚组中检出率一致,没有统计学差异(P>0.05)。结论基于双能量钙图/虚拟去钙的诊断模型可以准确鉴别疑似脑卒中患者出血灶与钙化灶的不同表现,并优于CT值的诊断结果。

Objective To explore the role of the discrimination model based on dual energy calcium map/virtual non-calcium image in differentiating intracranial hemorrhage foci from calcification foci,to clarify the value of differential diagnosis.Methods A total of 189 moderate hyperdense foci in 157 patients with suspected stroke in our hospital(94 in the training set and 95 in the test set)were collected retrospectively.A classification model based on dual-energy calcium/virtual non-calcium measurements was established and compared with the diagnoses based on CT values to analyze its benefits in terms of diagnostic efficacy.Meanwhile,subgroups of different lesion size were also studied to analyze the stability of the diagnostic model in large/small lesions.Results In the test set,the AUC of the CT-based model was 0.817(95%CI:0.724~0.890),while the AUC obtained using calcium/virtual non-calcium was 0.994(95%CI:0.951~1.000).Subjective assessment accuracy using calcium/virtual non-calcium and CT images was 82.1%and 68.4%,respectively.Meanwhile,the detection rate was also consistent in the<5 mm or>5 mm subgroup with no statistical difference(P>0.05).Conclusion The diagnostic model based on dual energy calcium/virtual non-calcium can accurately identify the different manifestations of hemorrhagic and calcification foci in patients with suspected stroke and outperforms the diagnostic results of CT values.