Mapping brain tumor microstructure: A multimodal study of diffusion MRI, intraoperative fluorescence, and neuropathology in navigated biopsies.

High-grade glioma characteristics such as heterogeneity and diffuse growth present a major diagnostic and therapeutic challenge, making accurate imaging essential for diagnosis and surgical planning. Diffusion MRI (dMRI) shows promise in tissue identification through a negative correlation between the dMRI apparent diffusion coefficient and tumor cellularity. Further, tissue disorganization due to tumor growth is correlated with decreased fractional anisotropy (FA) from diffusion tensor imaging (DTI). Q-space trajectory imaging (QTI) through free gradient waveform encoding during dMRI acquisition has been suggested as a framework for dMRI scalar map generation, enabling disentangled measures of shape, size, and orientation. We aimed to extend a clinically integrated workflow for optical guidance in frameless navigated brain tumor biopsies to include DTI and QTI scalars for multimodal analysis. Diffusion scalars were compared to tumor indications on tissue fluorescence, conventional imaging, and neuropathology in navigated brain tumor biopsy procedures. In seven high-grade glioma patients, the biopsied tissue volume was associated with decreased dMRI features (anisotropy, kurtosis, and order parameters) and increased diffusivity in DTI when compared with contralateral white matter. Principal components of diffusion scalars depend on microstructural (QTI) and diffusivity (DTI) parameters, respectively. Redundancy analysis between the dMRI scalars revealed scalar pairs that offer novel information for tissue separation that could be of interest for fine-tuning of the MRI protocol before further evaluation of QTI for tumor tissue identification in the clinical setting.