Quantitative comparison of reconstruction methods for intra-voxel fiber recovery from diffusion MRI

Validation is arguably the bottleneck in the diffusionMRI community. This paper evaluates and compares 20algorithms for recovering the local intra-voxel fiber structurefrom diffusion MRI data and is based on the results of the“HARDI reconstruction challenge” organized in the context ofthe “ISBI 2012” conference. Evaluated methods encompass amixture of classical techniques well-known in the literature suchas Diffusion Tensor, Q-Ball and Diffusion Spectrum imaging,algorithms inspired by the recent theory of compressed sensingand also brand new approaches proposed for the first timeat this contest. To quantitatively compare the methods undercontrolled conditions, two datasets with known ground-truthwere synthetically generated and two main criteria were used toevaluate the quality of the reconstructions in every voxel: correctassessment of the number of fiber populations and angularaccuracy in their orientation. This comparative study investigatesthe behavior of every algorithm with varying experimental conditionsand highlights strengths and weaknesses of each approach.