Can we trust computers to assess the cognition of stroke patients? A systematic review

Qi Zhang; Jia-Hang Wei; Xue Fu; Xin Liu; Xin-Yi Li; Wei Liu; Zhong-Liang Liu; Xiao-Qin Duan; Bin Zheng

doi:10.3389/fneur.2023.1180664

Can we trust computers to assess the cognition of stroke patients? A systematic review

Front Neurol. 2023 May 25:14:1180664. doi: 10.3389/fneur.2023.1180664. eCollection 2023.

Authors

Qi Zhang^#¹, Jia-Hang Wei^#², Xue Fu^#³, Xin Liu^{4

5}, Xin-Yi Li¹, Wei Liu¹, Zhong-Liang Liu¹, Xiao-Qin Duan^{1

5}, Bin Zheng⁵

Affiliations

¹ Department of Rehabilitation Medicine, Jilin University Second Hospital, Changchun, China.
² Jilin University, Changchun, China.
³ Changchun University of Chinese Medicine, Changchun, China.
⁴ School of Computer and Communication Engineering, University of Science and Technology Beijing, Beijing, China.
⁵ Surgical Simulation Research Lab, Department of Surgery, University of Alberta, Edmonton, AB, Canada.

^# Contributed equally.

Abstract

Purpose: To summarize the classification of computerized cognitive assessment (CCA) tools for assessing stroke patients, to clarify their benefits and limitations, and to reveal strategies for future studies on CCA tools.

Methods: A literature review was performed using PubMed, Embase, Scopus, JAMA Network, Cochrane Library and PsycINFO databases from January 1st, 2010, to August 1st, 2022. Two authors independently screened the literature following the same criteria, evaluated the study quality, and collected data from the articles.

Results: A total of 8,697 papers were acquired from the six databases. A total of 74 potentially eligible articles were selected for review. Of these, 29 articles were not relevant to this research, 3 were reviews, 2 were not written in English, and 1 was on an ongoing trial. By screening the references of the reviews, 3 additional articles were included in this study. Thus, a total of 42 articles met the criteria for the review. In terms of the CCA tools analyzed in these studies, they included five types: virtual reality (VR)-based, robot-based, telephone-based, smartphone-based, and computer-based cognitive assessments. Patients' stages of the disease ranged from the subacute phase and rehabilitation phase to the community phase. A total of 27 studies supported the effectiveness of CCA tools, while 22 out of 42 articles mentioned their benefits and 32 revealed areas for future improvement of CCA tools.

Conclusions: Although the use of CCA tools for assessing the cognition of post-stroke patients is becoming popular, there are still some limitations and challenges of using such tools in stroke survivors. More evidence is thus needed to verify the value and specific role of these tools in assessing the cognitive impairment of stroke patients.

Keywords: cognitive assessment screening instrument; cognitive impairments; computer-aided design; review; stroke.

Publication types

Systematic Review

Grants and funding

This work was supported by the Fundamental Research Funds from the Science and Technology Department of Jilin Province (grant number 20210101359JC, 2021) and Jilin University in China (grant number 2020B41, 2020).