Related Experiment Video
Updated: Jan 9, 2026

A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
A New Instrument for Assessing Cognitive Decline and Dementia: Results on the Classification Accuracy of the MODEMM
Cătălina Şandru1, Iulia Crișan1, Daniela Reisz2
1West University of Timișoara, Romania.
Abstract:
This study addresses the need for screening tests that can discriminate between dementia, mild cognitive impairment (MCI), and normal age-related memory functioning in understudied populations. One hundred sixty-four Romanian patients with dementia, MCI, and community members were assessed with the Memory of Objects and Digits and Examination of Memory Malingering (MODEMM), the MMSE-2 standard version (MMSE-2-SV), and quick mild cognitive impairment (QMCI) screen to determine each instrument's ability to distinguish between diagnostic groups and controls. The integral version of the MODEMM (MODEMM-I) classified diagnostic groups with outstanding accuracies (area under the curve [AUC] = .91-.99, p < .001), similar to QMCI (AUCs = .92-.98, p < .001) and the MMSE-2-SV (AUCs = .89-.99, p < .001). Cutoffs were adjusted for each diagnostic condition according to levels of education. Despite high-accuracy values, the MODEMM subscales were less sensitive to MCI than the integral version. Results support the MODEMM-I as an accurate screening tool for cognitive impairment in the understudied Romanian population.

