This study introduces a deep-learning framework aimed at improving the early detection of Alzheimer’s Disease (AD) through the analysis of eye movement patterns. We developed a Bidirectional Long Short-Term Memory (Bi-LSTM) network enhanced with an attention mechanism, utilizing a dataset consisting of eye movement data from both early-stage AD patients and a control group. The model captures temporal dynamics and ocular characteristics that may indicate early cognitive decline. Our empirical results show that the Bi-LSTM network with attention mechanism performs better than traditional models in metrics such as accuracy, precision, recall, F1 score, and the area under the Receiver Operating Characteristic (ROC) curve. These findings suggest that eye movement data could be a useful, non-invasive tool for early AD detection. The study highlights the potential for more accessible and timely diagnostic methods, which could support earlier intervention and better patient outcomes.