Biexponential diffusion attenuation in various states of brain tissue: Implications for diffusion‐weighted imaging

TLDR

Biexponential diffusion attenuation curves have implications for diffusion‑weighted imaging experiments. The study presents and evaluates possible reasons for the discrepancy between estimated magnitude components and physiological values. Diffusion‑weighted experiments up to b = 10 000 s mm⁻² revealed biexponential attenuation in normal and ischemic brain, with f1≈0.8–0.9 and f2≈0.1–0.3, and changes in these fractions tracked extracellular and intracellular space fractions during cell swelling and recovery from excitotoxic edema.

Abstract

Diffusion-weighted single voxel experiments conducted at b-values up to 1 x 10(4) smm-2 yielded biexponential signal attenuation curves for both normal and ischemic brain. The relative fractions of the rapidly and slowly decaying components (f1, f2) are f1 = 0.80 +/- 0.02, f2 = 0.17 +/- 0.02 in healthy adult rat brain and f1 = 0.90 +/- 0.02, f2 = 0.11 +/- 0.01 in normal neonatal rat brain, whereas the corresponding values for the postmortem situation are f1 = 0.69 +/- 0.02, f2 = 0.33 +/- 0.02. It is demonstrated that the changes in f1 and f2 occur simultaneously to those in the extracellular and intracellular space fractions (fex, f(in)) during: (i) cell swelling after total circulatory arrest, and (ii) the recovery from N-methyl-D-aspartate induced excitotoxic brain edema evoked by MK-801, as measured by changes in the electrical impedance. Possible reasons for the discrepancy between the estimated magnitude components and the physiological values are presented and evaluated. Implications of the biexponential signal attenuation curves for diffusion-weighted imaging experiments are discussed.