Optimization of the Data Pattern and Analysis Algorithm for the T2-based Water Suppression Diffusion MRImaging (T2wsup-dMRI) Technique.

We have proposed a T2-based free water suppression diffusion MRI (T2wsup-dMRI) technique to address parameter quantification issues due to cerebrospinal fluid (CSF) partial volume effects (PVEs), using a closed form (CF) algorithm. This study optimizes data patterns in (TE, b-value) space and analyzes algorithms for enhanced accuracy and precision. We simulated noise-added numerical, phantom, and brain MRI data to evaluate relative error and coefficient of variation in quantitative parameters using various data patterns and analysis algorithms (CF and least squares [LSQ] fitting). With 4 minimum data points applied to healthy brain tissue with T2 < 100 ms, the CF algorithm with water volume separation was optimal. For more than 4 points, a smaller b-value with shorter TE combined with 2d single- and bi-exponential LSQ fitting provided the best results. The T2wsup-dMRI technique reduces CSF-PVE artifacts in tissue-specific parameter quantification, enhancing approaches for patient needs, data acquisition, and computing costs.