摘要
目的 基于隐结构模型及数据挖掘评估MRI多序列联合检查对阿尔茨海默病(AD)患者认知功能障碍程度的诊断价值。方法 选取2021年1月—2024年8月绍兴市第七人民医院诊治的203例AD患者(AD组)和102例非AD患者(对照组)作为研究对象,并根据认知功能障碍程度将AD患者分为轻度组(86例)、中度组(72例)、重度组(45例),观察患者MRI多序列联合检查情况。结果 AD组海马体积和SN相位偏移值低于对照组,而ADC值较高(P<0.05)。隐结构模型显示,ADC、海马体积、黑质结构相位偏移值与AD具有密切关联。Modeler数据挖掘结果显示,“高风险脑部病变特征聚类群”患者MMSE评分较高,但存在潜在病变风险;“显著脑部病变特征聚类群”患者病变特征更明显,MMSE评分较低。重度组患者的左侧、右侧海马体积、SN相位偏移值低于轻度、中度组,ADC高于轻度、中度组(P<0.05)。AD患者的认知功能与海马体积、SN相位偏移值呈负相关关系,与ADC呈正相关关系(P<0.05)。ROC曲线分析显示,海马体积、ADC、SN相位偏移值及多序列联合诊断AD患者重度认知功能障碍的AUC值分别为0.748、0.673、0.802、0.892。结论 MRI多序列联合检查对评估AD患者认知功能障碍有一定价值,且结合隐结构模型及数据挖掘技术有利于为AD早期诊断和病情评估提供重要参考依据。
Objective Evaluate the diagnostic value of MRI multi-sequence combined examination for the degree of cognitive dysfunction in patients with Alzheimer's disease(AD)based on hidden structure models and data mining.Methods A total of 203 AD patients(AD group)and 102 non-AD patients(control group)diagnosed and treated at Shaoxing 7th People's Hospital from January 2021 to August 2024 were selected as the subjects in this study.The AD patients were divided into the following three groups according to their degree of cognitive dysfunction:mild,moderate and severe.Results The hippocampal volume and SN phase shift values in the AD group were lower than those in the control group,while the ADC value was higher(P<0.05).The implicit structure model demonstrates a close relationship between AD,ADC,hippocampal volume,and the phase shift value of substantia nigra structure.The results of the Modeler data mining show that patients in the"high-risk brain lesion feature cluster"have a higher MMSE score,however,there is a potential risk of lesion.Patients in the"significant brain lesion feature cluster"exhibit more pronounced lesion features and lower MMSE scores.The severity group exhibited reduced volumes of the left and right hippocampus and the SN phase shift value when compared to the mild and moderate groups,and the ADC value in the severity group was higher than observed in the mild and moderate groups(P<0.05).The cognitive function of AD patients was found to be negatively correlated with hippocampal volume and SN phase shift value,and positively correlated with ADC(P<0.05).ROC curve analysis demonstrated that the AUC values of hippocampal volume,ADC,SN phase shift,and multi-sequence combined diagnosis of severe cognitive dysfunction in AD patients were 0.748,0.673,0.802 and 0.892.Conclusion MRI multi-sequence combined examination has been shown to have a certain value in the evaluation of cognitive dysfunction in AD patients,and the combination of hidden structure models and data mining techniques has been demonstrated to be conducive to providing an important reference basis for the early diagnosis and condition assessment of AD.
作者
戚洪阳
林秀琴
周杰
QI Hongyang;LIN Xiuqin;ZHOU Jie(Radiology Department,Shaoxing 7th People's Hospital,Shaoxing,Zhejiang 312000,China;不详)
出处
《中华全科医学》
2025年第11期1920-1924,共5页
Chinese Journal of General Practice
基金
浙江省医药卫生科技计划项目(2023KY1271)。
关键词
阿尔茨海默病
隐结构模型
数据挖掘
磁共振
磁敏感加权成像
扩散加权成像
Alzheimer's disease
Hierarchical structure model
Data mining
Magnetic resonance
Magnetic sensitive weighted imaging
Diffusion-weighted imaging
