MRI联合临床特征对自身免疫性脑炎的诊断价值OA

Objective To investigate the diagnostic value of MRI combined with clinical features in autoimmune encephalitis(AE).Methods A total of 180 patients diagnosed and treated in the Jinan City People's Hospital and Ji'nan Integrated Traditional Chinese and Western Medicine Hospital from May 2022 to May 2024 were enrolled as subjects.They were divided into autoimmune encephalitis(AE)group(n=68)and viral encephalitis(VE)group(n=112)based on their final clinical diagnosis.General data,MRI results,including conventional scans,diffusion-weighted imaging(DWI),arterial spin labeling(ASL)perfusion imaging,proton magnetic resonance spectroscopy(1H-MRS),and clinical characteristics of both groups were collected and analyzed.Univariate and multivariate Logistic regression analyses were conducted to identify factors related to AE diagnosis,and receiver operating characteristic(ROC)curves were drawn according to the multivariate Logistic regression analysis results.Results Among the 68 AE patients,there were 26 cases(38.24％)of anti-NMDAR encephalitis and 16 cases(23.53％)of anti-LGI1 enphalitis,these two types accounted for the highest proportion.Univariate analysis revealed significant differences between the two groups regarding onset type,epilepsy,fever,headache,psychiatric behavioral abnormalities,lesion location,apparent diffusion coefficient(ADC)in lesions,relative ADC(rADC),regional cerebral blood flow(rCBF)in lesions,relative regional cerebral blood flow(rrCBF),choline(Cho)/N-acetyl-aspartate(NAA)ratio in lesions,Cho/creatine(Cr)ratio in lesions,and NAA/Cr ratio in lesions(P＜0.05).Multivariate Logistic regression analysis indicated that subacute onset,absence of fever,high ADC in lesions,low rCBF in lesions,and high NAA/Cr in lesions are influencing factors for diagnosing AE(P＜0.05).Based on the multivariate Logistic regression results,a diagnostic model was constructed.The Hosmer-Lemeshow goodness-of-fit test showed satisfactory calibration between the observed and expected values(x2=1.579,P=0.136).ROC analysis demonstrated the area under the curve(AUC)of 0.823,sensitivity of 85.44％,specificity of 79.44％,positive predictive value of 80.62％,negative predictive value of 82.91％,and accuracy rate of 83.06％.Conclusion Subacute onset,absence of fever,high ADC in lesions,low rCBF in lesions,and high NAA/Cr in lesions are influencing factors for diagnosing AE.A diagnostic model developed by combining MRI findings with clinical features can effectively improve the diagnostic accuracy for AE.