Application Research of Multimodal Clinical Biomarker Combined Model in the Early Diagnosis of Alzheimer's Disease

HUANG Sibo; LIANG Yuefei; CHEN Zekai; HE Wenjie; ZHONG Qi; LI Aidong; XIA Jun

doi:10.16458/j.cnki.1007-0893.2025.16.001

Abstract：Objective To explore the differences in olfactory function,eye movement characteristics,and magnetic resonance imaging(MRI)brain structure indicators between patients with Alzheimer's disease(AD)and mild cognitive impairment(MCI),and to analyze the value of combined application of multimodal biomarkers in the early diagnosis of AD.Methods Thirty patients with AD andforty patients with MCI admitted to the Second People's Hospital of Futian District Shenzhen,Shenzhen Luohu People's Hospital,and Shenzhen Second People's Hospital from July 2021 to August 2024 were selected as the AD group and MCI group.Forty healthy individuals who underwent physical examinations at shenzhen Second People's Hospital from January 2022 to August 2024 were recruited as the normal control(NC)group.Olfactory function was evaluated using the Chinese smell identification test(CSIT),eye movement parameters were recorded using artificial intelligence-based eye tracking technology,and brain region volumes were measured using 3.0 T MRI.Group differences,indicator correlations,and diagnostic efficacy were evaluated using analysis of variance,Spearman correlation analysis,and Logistic regression models.Results The volumes of the left and right hippocampus,left and right amygdala,left and right precentral gyrus,left and right prefrontal and left and right cuneus in the AD and MCI group were smaller than those in the NC group,and the volumes of the left hippocampus,left and right amygdala,left and right precentral gyrus,left and right prefrontal lobe,and left and right cuneus in the AD group were smaller than those in the MCI group.The mini-mental state examination(MMSE)and Montreal cognitive assessment(MoCA)scores in the AD and MCI groups were lower than those in the NC group,and the scores in the AD group were lower than those in the MCI group.The CSIT scores,error eye movement correction frequency,and new photo dwell time in the AD group were lower or shorter than those in the NC group,but the indexes of the MCI group were higher or longer than those of the AD group,with statistically significant differences(P<0.05).There were significant positive correlations between MMSE scores,MoCA scores,left hippocampal volume,right hippocampal volume,left amygdala volume,right amygdala volume,left precentral gyrus volume,left cuneus volume,error eye movement correction frequency,and CSIT scores(P<0.05),among which the left hippocampal volume had a strong positive correlation with the CSIT score(r=0.622,P<0.05).In terms of MRI brain structure indicators,the left hippocampal volume was significantly positively correlated with the error eye movement correction frequency(r=0.537,P<0.05),and also had a certain correlation with the new photo dwell time(r=0.320,P<0.05).The right amygdala volume was significantly positively correlated with the new photo dwell time(r=0.397,P<0.05).The left prefrontal lobe volume was significantly positively correlated with the error eye movement correction frequency(r=0.441,P<0.05).The receiver operating characteristic(ROC)curve analysis showed that the area under the curve(AUC)for diagnosing AD by CSIT,MMSE score,left hippocampal volume,and error eye movement correction frequency were 0.799,0.789,0.628,and 0.687,respectively.The AUC value of the logistic regression prediction model constructed with CSIT score,error eye movement correction frequency,left hippocampal volume,and MMSE score as variables reached 0.903,which was significantly higher than that of each individual indicator(P<0.05).Conclusion Multimodal biological markers of olfaction,eye movement,and MRI brain structure show differences among AD,MCI,and NC groups.The diagnostic model constructed by combining olfaction,eye movement,and MRI brain structure can further improve the diagnostic efficacy of early AD,providing a new strategy for early screening of AD.