Leveraging Machine Learning for Preliminary Early Onset Alzheimer's Disease Classification Using Cost-Effective Data Modalities

<p dir="ltr">This research aims to utilizes cost-effective neuropsychological and demographic data modalities from the ADNI dataset to train machine learning models for the preliminary risk prediction of Alzheimer's disease related dementia. The conventional diagnostics often involve costly and invasive medical assessment procedures, hindering accessibility for socioeconomically disadvantaged populations. The findings of this study highlights the potential of low-cost and noninvasive machine learning approaches for preliminary risk assessment of Alzheimer's disease related dementia. This paper introduces the MiCi preprocessing method to enhance data quality and improve model performance. Among 15 evaluated classifiers, Random Forest demonstrated the highest accuracy, at 92.4% over 1,000 predictions.</p>