Cognitive impairment screening using m-health: an android implementation of the mini-mental state examination (MMSE) using speech recognition

Paul Devos; Jari Debeer; Jelle Ophals; Mirko Petrovic

doi:10.1007/s41999-019-00186-0

Cognitive impairment screening using m-health: an android implementation of the mini-mental state examination (MMSE) using speech recognition

Eur Geriatr Med. 2019 Jun;10(3):501-509. doi: 10.1007/s41999-019-00186-0. Epub 2019 Apr 26.

Authors

Paul Devos¹, Jari Debeer², Jelle Ophals², Mirko Petrovic³

Affiliations

¹ Department of Information Technology, Faculty of Engineering and Architecture, Ghent University, Ghent, Belgium. p.devos@ugent.be.
² Department of Information Technology, Faculty of Engineering and Architecture, Ghent University, Ghent, Belgium.
³ Department of Internal Medicine, Section of Geriatrics, Faculty of Medicine and Health Sciences, Ghent University, Ghent, Belgium.

PMID: 34652802
DOI: 10.1007/s41999-019-00186-0

Abstract

Purpose: This paper describes an implementation of an Android application that allows cognitive functions to be tested.

Methods: The application consists of a slightly modified Mini Mental state Examination (MMSE) test that provides examiner assistance in diagnosing cognitive impairments. The application deploys speech recognition techniques to allow easy and automated scoring of the test. The test results and test-related information are stored in a database providing easy access to data for follow-up and analysis, resulting in overall benefits to the examiner workflow. A small-scale pilot study of 5 months duration was conducted in a nursing home where 15 residents were tested with the MMSE app test and with the (paper) MMSE test with the aim of determining the agreement between the two test methods.

Results: The final MMSE test scores, with a maximum score of 30, agree; the differences have a mean of 0.1, a standard deviation of 2.1 and fall in a [- 4, + 4] range as is illustrated in a Bland-Altman analysis. From examiner reflections, the motoric skills of the participant are indicated to contribute strongly to the time benefit of the assessment itself.

Conclusions: The findings of this study suggest that the mobile digital version of the slightly modified MMSE test has the potential to be used as an attractive alternative for the conventional paper version of the test.

Keywords: Android; Dementia; MMSE; Smartphone application; m-Health.