Related Experiment Video
Updated: Nov 20, 2025
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Community-based Model for Dementia Risk Screening: The Beijing Aging Brain Rejuvenation Initiative (BABRI) Brain
Yiru Yang1, Chenlong Lv2, He Li3
1State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China; Beijing Aging Brain Rejuvenation Initiative (BABRI) Centre, Beijing Normal University, Beijing, China.
Objectives:
To address the condition that community-based geriatric services for the assessment and promotion of older adults' cognitive ability systemically aimed at delaying or preventing dementia is lacking in China.
Design:
A community-based model including cognitive assessment and training, geriatric health guidance and long-term support was designed based on a prospective cohort study.
Setting And Participants:
Participants (N = 5593) were all from an ongoing cohort study, the Beijing Aging Brain Rejuvenation Initiative (BABRI) study.
Methods:
We conducted receiver operating characteristic, stepwise logistic regression and branch-and-bound algorithm analyses to select the most effective tests from the BABRI neuropsychological test battery. Canonical discriminant analysis was conducted to extract the first canonical variable as a composite index of the tests. In addition, we developed comprehensive surveys and computerized cognitive trainings targeting every cognitive domain.
Results:
The BABRI brain health system (BABRI-BHS) was designed to include SCREEN, ASSESS, and DIAGNOSE sessions. When distinguishing cognitively impaired older adults from cognitively healthy older adults, the canonical variable extracted from tests in the SCREEN session achieved an area under the curve (AUC) of 0.730 [95% confidence interval (95% CI) 0.671-0.789], with a sensitivity of 0.630 and a specificity of 0.780; in the ASSESS session, the AUC was 0.906 (95% CI 0.894-0.917), the sensitivity was 0.809, and the specificity was 0.854. A stepwise screening pathway is recommended when using the BABRI-BHS in communities to divide older adults into subtypes and to provide targeted interventions and long-term geriatric health guidance.
Conclusions And Implications:
The BABRI-BHS is an effective and efficient geriatric health care solution that is suitable for community-based dementia risk screening, providing stepwise cognitive assessments and helping older adults acquire tailored interventions and guidance conveniently.
