Alzheimer's disease (AD) is a neurodegeneration disease and the most common causes of dementia. Currently over five million in US are living with the disease,and it is estimated that by 2050, the number could reach 16 million and the direct costs of caring for those patients could reach $1.1 trillion annually. However, by now, there is no drug or treatment can cure or even slow the progression of disease. It has been proposed that early diagnosis and prevention might be the only efficient way to stop and/or slow the onset and progression of the disease. Thus there is a pressing and urgent need to find biomarkers that can accurately predict future disease progression and evaluate the effects of disease-modifying agents. However, despite recent efforts and significant progress in research, the reliable detection of preclinical AD and evaluation of drug treatment remains a major challenge. The effectiveness of recent functional magnetic resonance imaging (fMRI, a non-invasive technique to measure neural activity in vivo)-based approaches has been limited by the absence of a clear model of how AD affects neuronal processing, and how these changes in neuronal processing could be detected with fMRI at high sensitivity. Here we propose a novel fMRI biomarker of AD, Hcorr, which we recently developed to quantitatively predict cognitive impairments in individuals with developmental or neurological disorders. The novel technique holds strong promises with a few unique and significant advantages over other techniques, Hcorr can quantify disease progression as a single diagnostic number; is non-invasive and can detect neuronal dysfunction prior to onset of behavioral symptoms, thus making it an appealing technique to assess preventative drug treatment; and is highly sensitive and can detect treatment effects in a very small sample size of patients, with significant potential to increase speed and decrease cost in clinical drug trials.