Related Experiment Video
Updated: Jan 3, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Introducing Multiple Screener: An unsupervised digital screening tool for cognitive deficits in MS
L van Dongen1, B Westerik1, K van der Hiele2
1Department of Anatomy & Neurosciences, Amsterdam Neuroscience, MS Center Amsterdam, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, Netherlands.
Background:
Cognitive deficits affect up to 70% of all patients with Multiple Sclerosis (MS) and have a significant impact on quality of life. Cognitive assessments need to be performed by a neuropsychologist and are often time-consuming, hampering timely identification and adequate monitoring of cognitive decline in MS.
Objective:
To develop a time-efficient, unsupervised, digital tool to screen for cognitive deficits in MS.
Methods:
A digital (adjusted) version of the Brief International Cognitive Assessment for MS, including the Symbol Digit Modalities Test (SDMT, information processing speed), the California Verbal Learning Test (CVLT-II, verbal memory) and the Spatial Recall Test (SPART, visuospatial memory) was developed: Multiple Screener (intellectual property of Sanofi Genzyme). Firstly, the clarity and feasibility of the tool was confirmed by 16 patients with MS (mean age 50.9 years (SD 9.4, range 37-68). Next, in 60 healthy controls (HCs, mean age 44.5 years (SD 14.0, range 18-67), intraclass correlation coefficients (ICC) were calculated to describe how strongly the digital version resembled the paper and pencil-based assessment. Finally, 236 HCs (mean age 42.8 years (SD 12.8, range 18-69) were included to obtain norm scores for each test.
Results:
ICCs between digital and paper and pencil-based assessment were excellent to good (SDMT (ICC 0.79, confidence interval (CI) 0.67-0.87); CVLT-II (ICC 0.77, CI 0.64-0.85); SPART (ICC 0.61, CI 0.42-0.75)). For each test, a regression-based correction for the effect of age was applied on the raw scores before converting them to norm Z-scores. Additionally, the SDMT scores needed correction for education and the CVLT-II for education and sex (subgroups were created).
Conclusions:
Performance on an adjusted, digital version of the BICAMS correlates highly with the standard paper-and-pencil based test scores in HCs. Multiple Screener is an unsupervised, digital tool, with available norm scores, ultimately allowing for easy monitoring of cognitive decline in patients with MS.
