RESUMO
Background: The determination of brain volumes using visual ratings is associated with an inherently low accuracy for the diagnosis of Alzheimer's disease (AD). A support-vector machine (SVM) is one of the machine learning techniques, which may be utilized as a classifier for various classification problems. This study exploratorily investigated the accuracy of SVM classification models for AD subjects using brain volume and various clinical data as features. Methods: The study was designed as a retrospective chart review. A total of 201 eligible subjects were recruited from the Memory Clinic at Siriraj Hospital, Thailand. Eighteen cases were excluded due to incomplete MRI data. Subjects were randomly assigned to a training group (AD = 46, normal = 46) and testing group (AD = 45, normal = 46) for SVM modeling and validation, respectively. The results in terms of accuracy and a receiver operating characteristic curve analysis are reported. Results: The highest accuracy for brain volumetry (62.64%) was found using the hippocampus as a single feature. A combination of clinical parameters as features provided accuracy ranging between 83 and 90%. However, a combination of brain volumetry and clinical parameters as features to the SVM models did not improve the accuracy of the result. Conclusions: In our study, the use of brain volumetry as SVM features provided low classification accuracy with the highest accuracy of 62.64% using the hippocampus volume alone. In contrast, the use of clinical parameters [Thai mental state examination score, controlled oral word association tests (animals; and letters K, S, and P), learning memory, clock-drawing test, and construction-praxis] as features for SVM models provided good accuracy between 83 and 90%.
RESUMO
BACKGROUND: A strong inverse relationship of functional limitation and socioeconomic status has been established in western ageing society. Functional limitation can be related to chronic diseases, disuse, cognitive decline, and ageing. Among chronic diseases in the Thai population, cerebrovascular diseases, diabetes, and arthritis are common. These factors are known to contribute to disability and poor quality of life in the elder population. Neuropsychiatric problems, cognitive decline, dementia, and cultural issues in elderly people also can alter the quality of life of the elderly. METHODS: The Dementia and Disability Project in Thai Elderly (DDP) aims at comprehensively assessing community dwelling Thai elderly to understand the relationship between disability and motor function, neuropsychiatric symptoms, cognitive function, and chronic diseases. The DDP is the first study to look at the prevalence and etiology of dementia and of mild cognitive impairment (MCI) in Thai elders and to explore the relationship of cognition, disability, small vessel diseases and cortical degeneration with neuroimaging in Thai elderly people. 1998 Thai elders were screened in 2004-2006 and diagnosed as having MCI or dementia. 223 elders with MCI or dementia and cognitively normal elderly had brain magnetic resonance imaging (MRI) or at baseline. 319 elders from the 3 groups had blood tests to investigate the risks and possible etiologies of dementia including genotyping at baseline. RESULTS: The mean age of elders in this study is 69.51(SD=6.71, min=60, max=95) years. 689(34.9%) are men and 1284(65.1%) are women. Mean body weight was 58.36(SD=11.20) kgs. The regression model reveals that performance on gait and balance and serum triglyceride predicts activity of daily living performance (adjusted r2 = 0.280, f=2.644, p=0.003). The majority of abnormal gait in Thai elders was lower level gait disturbance. Only 1.5% (29/1952) had highest level gait disorders. 39.5% of 1964 subjects were free of chronic diseases. Treatment gap (indicating those who have untreated or inadequate treatment) of diabetes mellitus and hypertension in Thai elders in this study was 37% and 55.5% respectively. 62.6% of Thai elders have ApoE3E3 allele. Prevalence of positive ApoE4 gene in this study is 22.85%. 38.6% of Thai elders who had MRI brain study have moderate to severe white matter lesions. CONCLUSION: The large and comprehensive set of measurements in DDP allows a wide-ranging explanation of the functional and clinical features to be investigated in relation to white matter lesions or cortical atrophy of the brain in Thai elderly population. An almost 2 year follow up was made available to those with MCI and dementia and some of the cognitively normal elderly. The longitudinal design will provide great understanding of the possible contributors to disability in the elderly and to the progression of cognitive decline in Thai elders.
