King's College London

Research portal

Diagnosis of Dementia by Machine learning methods in Epidemiological studies: a pilot exploratory study from south India

Research output: Contribution to journalArticle

Sheshadri Iyengar Raghavan Bhagyashree, Kiran Nagaraj, Martin Prince, Caroline H. D. Fall, Murali Krishna

Original languageEnglish
JournalSocial Psychiatry and Psychiatric Epidemiology
Early online date11 Jul 2017
DOIs
Publication statusE-pub ahead of print - 11 Jul 2017

King's Authors

Abstract

BACKGROUND: There are limited data on the use of artificial intelligence methods for the diagnosis of dementia in epidemiological studies in low- and middle-income country (LMIC) settings. A culture and education fair battery of cognitive tests was developed and validated for population based studies in low- and middle-income countries including India by the 10/66 Dementia Research Group.

AIMS: We explored the machine learning methods based on the 10/66 battery of cognitive tests for the diagnosis of dementia based in a birth cohort study in South India.

METHODS: The data sets for 466 men and women for this study were obtained from the on-going Mysore Studies of Natal effect of Health and Ageing (MYNAH), in south India. The data sets included: demographics, performance on the 10/66 cognitive function tests, the 10/66 diagnosis of mental disorders and population based normative data for the 10/66 battery of cognitive function tests. Diagnosis of dementia from the rule based approach was compared against the 10/66 diagnosis of dementia. We have applied machine learning techniques to identify minimal number of the 10/66 cognitive function tests required for diagnosing dementia and derived an algorithm to improve the accuracy of dementia diagnosis.

RESULTS: Of 466 subjects, 27 had 10/66 diagnosis of dementia, 19 of whom were correctly identified as having dementia by Jrip classification with 100% accuracy.

CONCLUSIONS: This pilot exploratory study indicates that machine learning methods can help identify community dwelling older adults with 10/66 criterion diagnosis of dementia with good accuracy in a LMIC setting such as India. This should reduce the duration of the diagnostic assessment and make the process easier and quicker for clinicians, patients and will be useful for 'case' ascertainment in population based epidemiological studies.

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454