Deep Learning-Based Diffusion MRI Analysis Software

Diffusion MRI (dMRI) is a powerful tool for characterizing the local properties of microstructures in living organisms, using various parametric signal models including the well-known diffusion tensor imaging (DTI). The parameters have been inferred conventionally through fitting. Recently, machine learning based approaches become popular including the author’s method. Based on the method, a software named “diMaRIA (diffusion Magnetic Resonance Image Analyzer)” was developed for dMRI analysis and visualization. It works with a GUI in the Microsoft Windows environment and is freely available. In this abstract, a diffusion MRI analysis software named diMaRIA is introduced with its technical features such as synthetic Q-space learning, Q-space interpolation by radial basis functions, and runtime neural network.