Problem: Established methods for the monitoring of the local fluid distribution in human tissues are either complicated and cost-intensive or poorly reproducible. None of them are continuously applicaple in the brain. However, the non-invasive monitoring of brain edema represents an unsolved problem of intensive care medicine. Method: Multichannel MIS represents a promising alternative for the continuous monitoring of local fluid shifts and correlated pathologies in the brain, e. g. edema, hemorrhages or epileptic events. MIS allows measurements of the electrical tissue conductivity via the perturbation of an alternating magnetic field at multiple frequencies. The conductivity is closely related to water fractions in the intra- and extracellular space. MIS requires: (1) hardware which measures magnetic field changes by 1E7 lower than the excitation field. (2) solution of an inverse eddy-current problem for the reconstruction of the hydration parameters. This requires a realistic electromagnetic model of the head and efficient inverse solvers. Results: A highly sensitive prototype hardware, employing planar gradiometers, low-noise amplifiers and a digital synchronous demodulator. The feasibility of MIS has been demonstrated for homogeneous and inhomogeneous biological objects. Current activities: (1) Hardware improvement (2) Implementation of electromagnetic models for a human head and for a head phantom. The model forms the basis for solving the inverse problem.