Validation of models for the diffusion weighted MR signal in brain white matter

The diffusion weighted MR-signal attenuation in brain white matter is not mono-exponential. A bi-exponential model has been proposed since this fits the data well. We have modeled the diffusion weighted MR-signal by Monte Carlo simulations in a fiber geometry similar to brain white matter. Although the bi-exponential model fitted the data very well, the fitted diffusion coefficients and fractions did not converge to the simulated values in the intra and extra cellular space. However, fitting the cumulant expansion form terminated after the Nth order term to the data results in accurate fits for the diffusion coefficient and kurtosis.