The Application of Intelligent Data Models for Dementia Classification

Background and Objective: Dementia is a broad term for complex range of conditions that affect the brain, such as Alzheimer’s disease (AD). affects lot people in elderly community, hence there huge demand to better understand this condition by using cost effective quick methods, neuropsychological tests, since pathological assessments are invasive expensive resources. One promising initiatives deals with dementia Mild Cognitive Impairment (MCI) Disease Neuroimaging Initiative (ADNI), which includes cognitive Clinical Rating (CDR) scores. The aim research investigate non-invasive indicators, features, typically diagnosed clinical assessment within ADNI’s data their effect on dementia. Methods: To achieve aim, machine learning techniques have been utilized classify patients into Cognitively Normal (CN), MCI, or having dementia, based sum CDR scores (CDR-SB) besides demographic variables. Particularly, performance Support Vector Machine (SVM), K-nearest neighbors (KNN), Decision Trees (C4.5), Probabilistic Naïve Bayes (NB), Rule Induction (RIPPER) measured respect different evaluation measures, including specificity, sensitivity, harmonic mean (F-measure), among others, large number cases controls from ADNI dataset. Results: results indicate competitive when classifying subjects baseline selected variables technology. Though we observed fairly good across all algorithms utilized, was still variation ability, indicating some algorithms, NB C4.5, suited task status our data. Conclusions: Using CDR-SB scores, attributes pinpoint can be seen less approach could use aid diagnosis This study gives an indication comprehensive tool, CDR, may adequate assessing assigning class patients, upon visit, order speed further procedures.