基于人工智能MRI脑区容积比变化对早期帕金森病的诊断价值

Diagnostic value of artificial intelligence-based MRI brain volume ratio changes in early Parkinson's disease

目的探讨基于人工智能MRI脑区容积比变化对早期帕金森病(PD)中的诊断价值。方法前瞻性收集2021年11月至2023年8月在浙江省台州医院神经内科就诊的早期PD患者60例,以及同期本院健康体检者88例作为对照。使用自动分割软件联影智能精确计算脑内各区域的容积比。比较两组研究对象的不同脑区容积比,选择差异性脑区及感兴趣脑区进行单因素及多因素logistic回归及联合诊断分析。结果早期PD患者与对照组相比,海马、苍白球、壳核、丘脑、脉络丛等脑区容积比的差异均有统计学意义(均P<0.05)。以上这5个脑区容积比联合统计形成的诊断体系用于早期PD诊断的效能较好(AUC=0.939)。结论通过人工智能软件联影智能脑容积自动分割可辅助早期PD的临床诊断。

Objective To explore the diagnostic value of MRI-based brain volume automatic analysis software in early Parkinson's disease(PD).Methods Sixty early-stage PD patients who visited the Department of Neurology at Taizhou Hospital of Zhejiang Province from November 2021 to August 2023,as well as 88 healthy individuals who underwent physical examinations during the same period,were selected as the study subjects.The new automatic segmentation software uAI Sphere was used to accurately calculate the volume ratio of various brain regions.The volume ratios of different brain regions between two groups of research subjects were compared,and differential brain regions and regions of interest were selected to conduct univariate and multivariate logistic regression and joint diagnostic analyses.Results Compared with the healthy control group,there were statistically significant differences in the volume ratio of brain regions such as hippocampus,globus pallidus,putamen,thalamus,and choroid plexus in early PD patients(all P<0.05).The diagnostic system formed by combining the volume ratios of these five brain regions can be used for the diagnosis of early PD,with an AUC of 0.939.Conclusion The use of uAI Sphere brain volume automatic analysis technology can assist in the clinical diagnosis of early PD.