[New brain imaging techniques].

Neuroimaging technologies have improved neurology and neurosurgery by providing tools to look inside the brain and investigate its functions and diseases. As for any tool, the users should know the basics of each technique and be aware about their uses and limitations. Here we review these new techniques and illustrate their use with examples from studies at the University Hospital in Geneva. From all the techniques, MRI (Magnetic Resonance Imagery) has the highest spatial resolution. Taking advantage of the magnetic properties of the hydrogen nucleus, it is possible to reach a sub-millimeter resolution in 3D. When MRI images are digitized, they can be treated to perform re-slicing, segmentation and 3D reconstruction of cortical surfaces, as well as to measure anatomical structures (volumetry). Functional MRI (fMRI) is based on blood oxygenation changes when a task is performed or when epileptic activity occurred. Then it can be used to non-invasively show for example language, motor or epileptic network activation. Electromagnetic imaging techniques, based on EEG and MEG, have the power to localize in 3D the electrical activity of the brain with millisecond temporal resolution and then to follow the temporal activation of neuronal networks. These techniques use mathematical models and algorithms to compute 3D tomography from 2D recordings on the scalp. In the case of epilepsy, EEG allows epileptic foci identification among propagation sites when it is recorded with a sufficient number of electrodes (> 100) and when realistic head models are used. The functional imaging techniques from nuclear medicine (PET and SPECT) have become very useful in neuroscience to explore cerebral changes associated with neuronal pathologies as well as cognitive and sensory tasks. Many efforts have been made to develop new cameras and models to increase the range of research and clinical applications. Co-registration of structural and functional images allows us to add functional information to a structural deficit, or conversely to better interpret functional images such as PET, SPECT and EEG in terms of specific anatomy. In the case of SPECT and fMRI, substraction between ictal and interictal exams points out areas involved in epileptic processes.