Early Digital Cognitive Screening For Individuals Across The Dementia Continuum
This validation study uses cognitive tasks and machine learning to identify individuals with early signs of dementia.
Introduction
•By 2030, 1 in 6 people will be aged 60 or above (1).
•Mild cognitive impairment (MCI) is a critical transitional stage between normal aging and dementia, yet early detection remains challenging, partly due to the limitations of current screening tools
•In this study we are testing the feasibility and clinical validity of a novel self-administered digital online cognitive screener based on 2 tasks from the Creyos assessment platform to improve the early detection of dementia across diverse populations of aging individuals.
Creyos Platform
•Online neuropsychological battery of cognitive tasks.
•Normative database comprising ~85,000 participants.
•No manual administration required; practical and cost-effective for primary care.
The Creyos Dementia Protocol (CDP)
By leveraging the Creyos platform’s large normative database and internal datasets, we trained a machine learning (ML) model capable of distinguishing cognitively normal individuals from those with suspect cognitive impairments, including those resulting from MCI, Alzheimer's disease (AD), and age-related cognitive decline.
The two Creyos tasks included in the screener assess visuospatial working memory and attention. The screener algorithm produces a binary outcome (+ or -) to flag people for impairments.
The CDP further classifies cognitive impairment using tasks assessing:
•Verbal short-term memory
•Episodic working memory
•Response inhibition
•Mental rotation
+ GAD-7, PHQ-9, IADL, IQ-CODE
Preliminary Screener Results
In a sample of 14 patients with AD:
Sensitivity: all 14 patients were identified as impaired by the Creyos screener (100% sensitivity). 13/14 of these patients had an MMSE score below the threshold of 25.
Specificity: In a case-matched sample of control subjects, randomly drawn from a normative database of 9,350 individuals and matched for age, gender, and level of education, 2 of 14 had a SCR+ result (86% specificity). (2)
Discriminant validity: when comparing groups of people with pre-existing health conditions to case-matched healthy controls, we observed:
•Significantly lower scores in one or more cognitive domains (reddish cells with asterisks) for other conditions.
•No difference in screener flag rates P(SCR+)—i.e., no asterisks in the rightmost column.
•Only the AD patients (top row) had a significantly higher probability of positive screener results P(SCR+) than their matched controls – indicating that the screener can effectively discriminate between signs of dementia and other health conditions. (2)
NEXT STEP: Large-Scale, Longitudinal Validation Study
Objectives:
1.Assess the feasibility and validity of the CDP in distinguishing healthy aging from early signs of pathological cognitive decline.
2.Evaluate risk and protective factors associated with increased risk of cognitive decline over 12 months.
Methods
400 Participants with either:
•Healthy Cognition
•Subjective Cognitive Decline
•MCI
•Dementia
Procedure:
Baseline Testing Session (35 minutes)
•CDP Creyos tasks (screener + 4 tasks)
•Comprehensive Demographics Questionnaire
•Functional and Mental Health Scales
•System Usability Scale
6-month Testing Session (30 minutes)
•CDP Creyos tasks (screener + 4 tasks)
•Comprehensive Demographics Questionnaire
•Functional and Mental Health Scales
12-month Testing Session (30 minutes)
•CDP Creyos tasks (screener + 4 tasks)
•Comprehensive Demographics Questionnaire
•Functional and Mental Health Scales
Demographic Variables:
•Sex
•Ethnicity
•Menopause status
•Education level and occupation
•Medical comorbidities
•Socialization
•Exercise
•Substance abuse (alcohol, cannabis, cigarettes)
Project Progress
•Recruitment began: 09/06/2026
•51 participants recruited, 28 baseline datasets collected
Predicted Results
•Screener tool will be usable by older adults and feasible for remote administration given the successful use of Creyos by older adults (>3000) and previous reports of online cognitive tests in older adults (3).
•Following preliminary validation analysis in AD patients (2), we expect the screener and full CDP assessment to demonstrate high accuracy in identifying individuals on the dementia continuum.
•Cognitive trajectories will vary with risk/protective factors such as education level, socialization, and exercise.
Conclusion
•Novel, scalable solutions for the early detection of dementia are crucial to transition from a reactive, treatment-based model to a proactive, prevention-focused approach.
•This approach can facilitate timely pharmacological and lifestyle interventions, improving patient outcomes and optimizing healthcare resources.
•Preliminary validation analyses demonstrate that the Creyos MCI screener has high accuracy in identifying cognitive markers of AD. This longitudinal, large-scale study will assess the screener as a scalable and effective solution to triaging older adults.
References:
1.World Social Report: Leaving No One Behind In An Ageing World. (2023). DESA Publications.
2.Urian et al. (2025). Early Cognitive Screening For Individuals On The Dementia Continuum. Journal of Alzheimer’s Disease.
3.Rentz Et Al. (2016). The Feasibility Of At-home iPad Cognitive Testing For Use In Clinical Trials. The Journal of Prevention of Alzheimer’s Disease.
