Correlation Tensor Imaging Enhances Histological Differences Between Mouse Glioblastoma Subtypes

Correlation Tensor Imaging (CTI) has been recently proposed as a novel methodology for resolving kurtosis sources without relying on strong multi-gaussian component assumptions. Here, CTI is harnessed to decipher kurtosis sources in two different mouse glioblastoma subtypes. Our results reveal that CTI’s kurtosis source separation resolves the underlying microstructural differences between the glioblastoma subtypes and faithfully represents their specific histopathological features. In comparison to previous diffusion MRI approaches, CTI kurtosis maps present enhanced sensitivity towards tumor differences. These first results show the potential of CTI in providing more sensitive and specific future biomarkers for monitoring tumor progression and therapy outcome.