Use, misuse, and abuse of apparent diffusion coefficients

The notion of effective, time-dependent or, equivalently, apparent diffusion coefficient (ADC) and its use for characterization of porous media are revisited. It is argued that the dynamic ADC, quantifying the mean-square displacement of spin-bearing particles, should not be confused with its counterpart measured by a pulsed gradient spin echo technique. The former is an intrinsic characteristic of the medium, independent of the applied magnetic field. In contrast, the spin-echo ADC depends on the experimental setup (e.g., gradient intensity and temporal profile), raising potential ambiguities in the interpretation of diffusion-weighted measurements, which may be strongly misleading when the Gaussian phase approximation (GPA) does not hold. The oversimplified use of a single b-value is criticized. Several fitting models beyond the GPA are discussed.