Characterization of Brain Tissues by the Field Dependence of Their Longitudinal Relaxation Rates

Abstract

Since the very beginning of the development of medical NMR techniques, one of the major goals has been the differentiation between normal and pathological tissue, with special interest in tumors. MRI techniques are now widely used, and their sensitivity in respect of pathological alterations is often high. However, their specificity is still low, i.e., identification of a tissue by its NMR data is still a difficult task. The main reason lies in the extensive overlap of the main parameters T1, T2, and Q between different normal and pathological tissues. This feature is complicated by the strong field dependence of T1, making comparisons of results obtained at different field strengths even more difficult. Futhermore, the NMR parameters of tissues are difficult to measure with satisfactory accuracy on clinical imagers, and systematic errors are often encountered.