Multi Modal EEG-Based Classification of Alzheimer's Disease and Mild Cognitive Impairment Using Hilbert Huang Transform

This study investigates whether high-resolution brain-functional connectivity, which is measured by using Electroencephalography (EEG) combined with Hilbert Huang Transform (HHT) can be useful in Alzheimer Disease Diagnosis. The major issue with Alzheimer's disease (AD) is that it leads to cognitive decline and affects millions of people worldwide. Early prediction of the disease is mandatory and crucial for timely intervention and effective care. EEG signals, a non-invasive and cost-effective method to assess brain activity, have shown potential in predicting this disease. In this study, we propose an approach that utilizes EEG data combined with Hilbert-Huang Transform (HHT) and machine learning algorithms for classification which is a novelty of the study that enables early prediction of Alzheimer's disease. Our methodology was tested on a dataset comprising EEG signals recorded using 19 electrodes from both healthy individuals and patients with Alzheimer's and the Mild Cognitive Impairment (MCI). The results were promising, demonstrating high accuracy, sensitivity, and specificity. Machine learning techniques can be employed for categorizing EEG signals, but it essential to validate their effectiveness using large and diverse datasets. This research highlights the potential of HHT and EEG data in the prediction of this disease. Principal Component Analysis (PCA) is used for dimensional reduction of the dataset. Notably EEG signals combined with Hilbert Huang Transform and Principal Component Analysis gives better accuracy of 98.6% for KNN. This research highlights the potential of EEG analysis combined with advanced techniques for accurate classification of dementia-related disorders.