Related Experiment Video
Updated: Oct 10, 2025
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Comparison of the Mini-Mental State Examination and Computerized Brief Cognitive Screening Test as Cognitive
Che-Sheng Chu1, I-Chen Lee2, Chuan-Cheng Hung3
1Department of Psychiatry, Kaohsiung Veterans General Hospital, Kaohsiung,Taiwan.
Background:
The aim of this study was to establish the validity and reliability of the Computerized Brief Cognitive Screening Test (CBCog) for early detection of cognitive impairment.
Methods:
One hundred and sixty participants, including community-dwelling and out-patient volunteers (both men and women) aged ≥ 65 years, were enrolled in the cross section study. All participants were screened using the CBCog and Mini-Mental State Examination (MMSE). The internal consistency of the CBCog was analyzed using Cronbach's α test. Areas under the curves (AUCs) of receiver operating characteristic analyses were used to test the predictive accuracy of the CBCog in detecting mild cognitive impairment (MCI) in order to set an appropriate cutoff point.
Results:
The CBCog scores were positively correlated with the MMSE scores of patients with MCI-related dementia (r = 0.678, P < .001). The internal consistency of the CBCog (Cronbach's α) was 0.706. It was found that the CBCog with a cutoff point of 19/20 had a sensitivity of 97.5% and a specificity of 53.7% for the diagnosis of MCI with education level ≥ 6 years. The AUC of the CBCog for discriminating the normal control elderly from patients with MCI (AUC = 0.827, P < 0.001) was larger than that of the MMSE for discriminating the normal control elderly from patients with MCI (AUC= 0.819, P < .001).
Conclusion:
The CBCog demonstrated to have sufficient validity and reliability to evaluate mild cognitive impairment, especially in highly educated elderly people.
