A general framework for experiment design in diffusion MRI and its application in measuring direct tissue-microstructure features

This article introduces a new and general framework for optimizing the experiment design for diffusion MRI of samples with unknown orientation. An illustration then uses the framework to study the feasibility of measuring direct features of brain-tissue microstructure in vivo. The study investigates the accuracy and precision with which we can estimate potentially important new biomarkers such as axon density and radius in white matter. Simulation experiments use a simple model of white matter based on CHARMED (composite hindered and restricted model of diffusion). The optimization finds acquisition protocols achievable on modern human and animal systems that consist of 120 measurements with fixed maximum gradient strengths. Axon radii in brain tissue are typically in the range 0.25-10 mu m. Simulations suggest that estimates of radii in the range 5-10 mu m have highest precision and that a maximum gradient strength of 0.07 T m(-1) is sufficient to distinguish radii of 5, 10, and 20 pm. Smaller radii are more difficult to distinguish from one another but are identifiable as small. A maximum gradient strength of 0.2 T m(-1) distinguishes radii of 1 and 2 mu m. The simulations also suggest that axon densities and diffusivity parameters in the normal range for white matter are recoverable. The experiment-design optimization has applications well beyond the current work to optimize the protocol for fitting any model of the diffusion process.