Related Experiment Video
Updated: Apr 21, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Development of a computerized tool for the chinese version of the montreal cognitive assessment for screening mild
Ke Yu1, Shangang Zhang2, Qingsong Wang1
1Department of Neurology, Chengdu Military General Hospital, No270, Rongdu Avenue, Jinniu District, Chengdu 610083, Sichuan, China.
Background:
The Montreal Cognitive Assessment (MoCA) is used for screening mild cognitive impairment (MCI), and the Beijing version (MoCA-BJ) is widely used in China. We aimed to develop a computerized tool for MoCA-BJ (MoCA-CC).
Methods:
MoCA-CC used person-machine interaction instead of patient-to-physician interaction; other aspects such as the scoring system did not differ from the original test. MoCA-CC, MoCA-BJ and routine neuropsychological tests were administered to 181 elderly participants (MCI = 96, normal controls [NC] = 85).
Results:
A total of 176 (97.24%) participants were evaluated successfully by MoCA-CC. Cronbach's α for MoCA-CC was 0.72. The test-retest reliability (retesting after six weeks) was good (intraclass correlation coefficient = 0.82; P < 0.001). Significant differences were observed in total scores (t = 9.38, P < 0.001) and individual item scores (t = 2.18-8.62, P < 0.05) between the NC and MCI groups, except for the score for "Naming" (t = 0.24, P = 0.81). The MoCA-CC total scores were highly correlated with the MoCA-BJ total scores (r = 0.93, P < 0.001) in the MCI participants. The area under receiver-operator curve for the prediction of MCI was 0.97 (95% confidence interval = 0.95-1.00). At the optimal cut-off score of 25/26, MoCA-CC demonstrated 95.8% sensitivity and 87.1% specificity.
Conclusion:
The MoCA-CC tool developed here has several advantages over the paper-pencil method and is reliable for screening MCI in elderly Chinese individuals, especially in the primary clinical setting. It needs to be validated in other diverse and larger populations.
