Alzheimer’s disease (AD) is characterized by a prolonged preclinical　phase during which neuropathological changes accumulate without overt　cognitive symptoms. This latency underscores the urgent need for　sensitive and specific biomarkers capable of detecting AD at its　earliest stages. Reliable early biomarkers would improve diagnostic　accuracy and enable timely intervention with disease-modifying　therapies to prevent or delay cognitive decline. Although current AD　biomarkers are useful, their clinical adoption is hindered by challenges such as high costs, invasiveness, and limited　accessibility. In this presentation, we will explore the potential of　dynamical system modeling of fMRI-based connectivity alterations and　AI-driven image synthesis as two non-invasive, scalable strategies for　the early detection of AD.