[Possibilities of technical and methodological optimization of functional magnetic resonance tomography]

The techniques of in vivo functional magnetic resonance imaging (fMRI) have been established over the past few years. Signal changes resulting from small differences in the magnetic resonance signal caused by variations in the oxygenation state of the venous vasculature are used to map neuronal activity of the brain. While many of the advantages for using MRI for functional neuroimaging are quite obvious (high spatial and temporal resolution, direct anatomic correlation, noninvasiveness), the underlying mechanisms and problems are not fully known. This article focuses on selected topics of fMRI techniques and their problems, such as field strength B0, different types of sequences and their parameters, 2D and 3D data acquisition, different statistical methods and motion artifacts. The authors favor the use of a 1.5 T clinical scanner with EPI capabilities, 2D FLASH sequences with a TR < 60 ms, TE = 30-40 ms, alpha < 20 degrees for high resolution and sensitivity, fMRI and multislice EPI for high temporal and overview and fMRI as well as cross-correlation for post-processing.