Despite the widespread use of neuroimaging tools (morphological and functional) in the routine diagnostic of cerebral diseases, the information available by the end user -the clinician- remains largely limited to qualitative visual analysis. This restriction greatly reduces the diagnostic impact of neuroimaging in routine clinical practice and increases the risk of misdiagnosis. In this context, researches are focussing on the development of sophisticated automatic analyses able to extract clinically relevant information from the captured data. The identification of biological markers at early stages of Alzheimer’s disease (AD) contributes to diagnostic accuracy and adds prognostic value. However, in spite of recent developments, results of structural and functional imaging studies on predicting conversion to AD are not uniform. We provide here an overview of analysis methods and approaches, discussing their contribution to clinical assessment.