Individualized functionnectome for the statistical assessment of white matter circuits underlying task-fMRI activations in glioma patients

The Functionnectome framework enables the projection of task-related fMRI signals onto the underlying white matter pathways, using anatomical priors derived from structural connectivity. However, the existing "standardized" priors are based on averaged tractograms from healthy controls and cannot accurately capture the altered anatomy found in patients with brain lesions. This study extends the Functionnectome framework by generating individualized anatomical priors from subject-specific diffusion MRI tractography, allowing for improved integration of functional and structural data in patients with brain tumors. Twenty-six patients with gliomas (9 females; mean age 43 ± 17 years) underwent 3 T MRI, including a multishell diffusion protocol and task-based fMRI targeting motor and/or language functions. Whole-brain tractography was reconstructed using three methods (TensorDet, PFT, and iFOD2) and converted into voxelwise individualized priors. Functionnectome maps were estimated using both standardized and individualized priors, and Z-statistic activation maps were obtained via GLM analysis. Similarity metrics and atlas-based consistency were used to compare standardized and individualized Functionnectomes. Functionnectomes derived from probabilistic tractography showed moderate correlations with standardized Functionnectomes (average Pearson's r ≈ 0.5), highlighting the influence of individual structural variability while maintaining comparable activation patterns. The PFT-based Functionnectome exhibited superior anatomical plausibility, with consistent overlap with expected motor white matter tracts and identification of relevant bundles in up to 100% of cases, compared to 50-60% with standardized priors. The individualized Functionnectome enhances the anatomical validity and subject specificity of structure-function mapping, advancing precision neuroimaging for clinical and neurosurgical applications.