Related Experiment Video
Updated: Jun 30, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Digital tools and biomarkers for cognitive screening: A systematic review
Seunghui Hong1, ChanHee Kim2, Seon Heui Lee3
1Department of Nursing, Graduate School, Gachon University, Incheon, Republic of Korea.
Objective:
This systematic review aimed to evaluate and compare three major types of digital cognitive assessment tools-digitized traditional tests, newly developed digital tools, and digital biomarkers. It also explored their validity, feasibility, and potential integration into community-based healthcare, along with clinical utility and implementation challenges.
Methods:
A search of nine electronic databases (July 2024) was conducted following PRISMA guidelines. Eligible studies included those reporting diagnostic accuracy and practical application of digital cognitive assessment tools or biomarkers. Data were narratively synthesized, and diagnostic accuracy metrics such as sensitivity, specificity, and area under the curve (AUC) were extracted. The review was prospectively registered in PROSPERO (CRD42024564512).
Results:
A total of 48 studies involving older adults with or at risk of cognitive impairment were included. These studies evaluated digital cognitive tools or biomarkers and reported diagnostic performance comparison with established tests or clinical diagnosis across clinical, community, and home settings. Approaches were categorized into digitized traditional tests (AUC: 0.65-1.0), newly developed digital cognitive tools (AUC: 0.75-0.94), and digital biomarkers (AUC: 0.59-0.95). Although diagnostic accuracy was generally promising, most studies were conducted in controlled settings with limited longitudinal or real-world validation. Evidence supporting the role of community-based healthcare professionals, such as nurses, remains limited; however, self-administered and remote assessments indicate potential for future community-based implementation. Future validation in real-world, community-based settings is essential before widespread implementation.
Conclusions:
Digital cognitive assessments and biomarkers offer valuable opportunities to enhance early screening and continuous monitoring of cognitive impairment. Based on these findings, we suggest that selection between digitized traditional tests and newly developed tools should be tailored to specific context, with digital biomarkers serving as complementary tools to strengthen cognitive assessment, particularly in community-based nursing practice. Policymakers and healthcare organizations should establish guidelines to safe and effective adoption of digital cognitive tools in community and clinical settings.
