Automated Pipeline Development for Multi-Compartmental Volumetric Glioblastoma Segmentation and Advanced MRI Parameter Extraction

Motivation: Automatic volumetric approach for accurate glioblastoma segmentation combined with perfusion and diffusion data offers significant advantages for quantitative image analysis over two-dimensional manual methods commonly used, enabling precision medicine, improved prognostication, treatment planning, and follow-up. Goal(s): Develop and implement an automated pipeline for multi-compartmental glioblastoma segmentation and physiologic MRI parameter extraction. Approach: Utilizing our validated deep-learning segmentation algorithm and Olea-Sphere software, we created a pipeline generating coregistered anatomic, diffusion, and perfusion MRI sequences with overlaid segmentation masks and descriptive statistics for sub-compartment volumes visible in PACS. Results: The pipeline outputs comprehensive MRI data with segmented compartments and quantitative metrics. Impact: This automated pipeline can enhance clinical decision-making and personalized treatment for glioblastoma patients. Its development will facilitate new research on imaging biomarkers, ultimately improving patient outcomes and advancing neuroimaging practices.