探索智能化神经功能评估在缺血性脑白质病变筛查、诊断中的价值

Exploring the value of intelligent neurological function assessment system in screening and diagnosis of ischemic white matter lesions

目的探索智能化评估系统在缺血性脑白质病的早期筛查、诊断及病情评估中的价值。方法收集脑白质高信号(WMH)组109例,正常对照组50例,通过智能化神经功能评估系统对受试者进行认知功能评估。同时由两位神经内科医生进行Mini-Cog简易认知评价,即分别通过智能及人工两种方式评估两组患者的认知功能水平;分别通过智能化步态评估系统及人工评价定量分析受试者起身时间、转身时间、步长、步速、步频、步宽等步态参数。评价智能化评估系统对不同人群认知障碍及步态障碍的识别率与准确性,采用智能化步态评估比较两组患者的步态差异,采用交叉表计算使用智能化步态评估筛查WMH的灵敏度和特异度;采用spearman秩相关检验分析明确WMH严重程度与步态异常间的关系。结果青浦地区50~80岁的WMH患者其认知障碍的发生率为71.6%、步态异常的发生率为35.8%。WMH组与正常对照组在转身时间、步宽、步长、步速、步频间差异均有统计学意义,P<0.05。WMH严重程度(Fazakas评分)与起身时间(相关系数0.182、P=0.022)、步宽(相关系数0.288、P<0.001)、步长(相关系数-0.209、P=0.008)、步速(相关系数-0.236、P=0.003)相关。结论智能化神经功能评估系统受限于方言等影响因素导致认知评估的准确率低;而智能化步态评估的特异度是82%,但受限于认知评估的影响,智能化平台在临床实际应用中还存在亟需解决的难题。

Objective To explore the value of intelligent evaluation system in early screening,diagnosis and evaluation of ischemic leukdisease.Methods 109 people from white matter group and 50 people from normal control group were collected for cognitive assessment through intelligent neurological assessment system,and Mini-Cog cognitive assessment by neurology specialist to determine impaired cognitive function;gait parameters such as rise time,turning time,step length,step speed,step frequency and step width were analyzed by intelligent gait assessment system.Compare the identification of cognitive impairment by two cognitive assessment methods,compare the gait differences of patients in the two groups,and to evaluate the sensitivity and specificity of hyperintensity,and use spearman rank correlation test to clarify the relationship between the severity of hyperintensity and gait abnormality.Results The incidence of cognitive impairment was 71.6%and 35.8%.The WMH group and the normal control group were statistically different between the turn time,step width,step length,step speed,and step frequency,P<0.05.The severity of WMH(Fazakas score)was correlated with the rise time(correlation coefficient 0.182,P=0.022),step width(correlation coefficient 0.288,P<0.001),step length(correlation coefficient-0.209,P<0.001),and step speed(correlation coefficient-0.236,P=0.003).Conclusion Intelligent neural function assessment system is limited by the low accuracy of cognitive assessment due to the limitation of dialect,while the specificity of intelligent gait assessment is 82%,but under the influence of cognitive assessment,intelligent platform still has some problems to be solved in clinical application.