Atlas-Based MRI Spatial Phenotyping in Pediatric Low-Grade Glioma

Atlas-based voxel-wise probabilistic mapping of preoperative MRI in 349 pediatric low-grade glioma patients shows tumor distribution differs by age, molecular subtype, and extent of resection.

• Why it matters: Clinical pLGG studies often reduce tumor location to a single manual category, missing spatial extent across multiple brain regions and limiting quantitative links to biology and surgical outcomes.
• What they did: Preoperative MRIs from 349 patients were registered to the Haskins pediatric template using EasyReg, then voxel-wise probabilistic maps and atlas-derived spatial and morphometric features were used to model resection status.
• The result: Atlas-derived spatial features distinguished gross total/near total resection from subtotal resection or biopsy/no-surgery with replication AUC 0.864, and replacing manual location with 107 automated overlap features did not significantly change discrimination (smallest P = 0.12).

AI summary of the abstract · may contain errors · not medical advice · read the paper.