It has been shown that double wave vector diffusion weighting, which employs two gradient pulse pairs of independent directions, can provide information about tissue structure that is not easily available otherwise, such as cell size or shape in a tissue sample. One approach to measure cell size is based on the signal difference between parallel and antiparallel gradient orientations at small mixing times between the two diffusion weightings. A major difficulty for in vivo application is the small size of the signal difference if clinical MR systems with limited gradient hardware are employed. In this study, the method is applied to human brain tissue in vivo, using whole-body gradients. Data are reported for the corticospinal tracts. The characteristics of the observed signal difference between parallel and antiparallel gradient orientations are consistent with both analytical and numerical predictions. As an estimate of pore size, the resulting mean squared radius of gyration of the pores amounts to approximately 4 µm(2) . An analysis that accounts for the finite values of gradient pulse duration and diffusion time yields a volume contribution-weighted mean pore diameter of 13 ?m if a cylindrical pore shape is assumed. The results demonstrate that the technique can be applied in vivo.
It has been shown that double wave vector diffusion weighting, which employs two gradient pulse pairs of independent directions, can provide information about tissue structure that is not easily available otherwise, such as cell size or shape in a tissue sample. One approach to measure cell size is based on the signal difference between parallel and antiparallel gradient orientations at small mixing times between the two diffusion weightings. A major difficulty for in vivo application is the small size of the signal difference if clinical MR systems with limited gradient hardware are employed. In this study, the method is applied to human brain tissue in vivo, using whole-body gradients. Data are reported for the corticospinal tracts. The characteristics of the observed signal difference between parallel and antiparallel gradient orientations are consistent with both analytical and numerical predictions. As an estimate of pore size, the resulting mean squared radius of gyration of the pores amounts to approximately 4 µm(2) . An analysis that accounts for the finite values of gradient pulse duration and diffusion time yields a volume contribution-weighted mean pore diameter of 13 ?m if a cylindrical pore shape is assumed. The results demonstrate that the technique can be applied in vivo.