Leveraging Insights from KNN and LSTM Models for Alzheimer’s Disease Early Detection

In this research, soft computing methods are explored for the early detection of dementia from EEG signal analysis. The study extract features from key frequency bands (Delta, Theta, Alpha, Beta) and uses the machine learning models such as K Nearest Neighbors (KNN) and Long Short Term Memory (LSTM) networks to identify small cognitive changes. Specifically, the LSTM model exhibits exceptional performance, particularly with increased sensitivity and specificity, especially in the Alpha and Beta bands, for early stage dementia detection. This work provides a promising alternate to costly existing methods and contributes to the development of noninvasive, scalable diagnostic tools for dementia. Early detection by observing the synchrony across the brain lobe could help improve diagnosis and contribute to future AI based healthcare applications.