Fully Automated Dynamic Contrast Enhanced MRI Image Processing Pipeline Reduces Variance in Ktrans

Motivation: Processing dynamic contrast-enhanced (DCE) MRI involves several preprocessing steps, making it very time-consuming and challenging to achieve consistency and reproducibility. Goal(s): To develop an automated software pipeline for processing DCE MRI images, enhancing both the speed and consistency of DCE data processing. Approach: We developed a DCE pipeline with preprocessing steps for motion correction, bias field removal, automatic arterial input function (AIF) selection, image alignment, z-slice normalization, segmentation, and quality control report generation. Results: The pipeline successfully processed 400 out of 407 cases, showing significant reductions in population standard deviations for Ktrans values in gray and white matter, and AIF quality scores. Impact: This software will facilitate more reproducible and consistent DCE processing across a variety of imaging sites and research labs.