Meningioma microstructure assessed by diffusion MRI: an investigation of the source of mean diffusivity and fractional anisotropy by quantitative histology
BackgroundMean diffusivity (MD) and fractional anisotropy (FA) obtained with diffusion MRI (dMRI) have been associated with cell density and tissue anisotropy across tumors, but it is unknown whether these associations persist at the microscopic level. PurposeTo quantify the degree to which cell density (CD) and structure anisotropy (SA), as determined from histology, account for the intra-tumor variability of MD and FA in meningioma tumors. Furthermore, to clarify whether histological features other than cell density account for additional intra-tumor variability of MD. Materials and MethodsWe performed ex-vivo dMRI at 200 m isotropic resolution and histological imaging on 16 excised meningioma tumor samples. Diffusion tensor imaging (DTI) was used to map MD and FA, as well as the in-plane FA (FAIP). Histology images were analyzed in terms of cell nuclei density and structure anisotropy (obtained from structure tensor analysis) and were used separately in a regression analysis to predict MD and FAIP, respectively. A convolutional neural network (CNN) was also trained to predict the dMRI maps from histology patches. The association between MRI and histology was analyzed in terms of coefficient of determination (R2). Regions showing unexplained variance (large residuals) were analyzed to identify features apart from cell density and structure anisotropy that could influence MD and FAIP. ResultsCell density assessed by histology poorly explained intra-tumor variability at the mesoscopic level (200 m) in MD (median R2 = 0.06, interquartile range 0.01 - 0.29) or FAIP (median R2 = 0.19, 0.09 - 0.29). Samples with low R2 for FAIP exhibited low variations throughout the samples and thus low explainable variability, however, this was not the case for MD. Across tumors, cell density and structure anisotropy were associated with MD (R2 = 0.58) and FAIP (R2 = 0.82), respectively. In 37% of the samples (6 out of 16), cell density did not explain intra-tumor variability of MD when compared to the degree explained by the CNN. Tumor vascularization, psammoma bodies, microcysts, and tissue cohesivity were associated with bias in MD prediction when solely CD was considered. Our results support that FAIP is high in the presence of elongated and aligned cell structures, but low otherwise. ConclusionCell density and structure anisotropy account for variability in MD and FAIP across tumors but cell density does not explain MD variations within the tumor, which means that low or high values of MD locally may not always reflect high or low tumor cell density. Features beyond cell density need to be considered when interpreting MD. HighlightsO_LICell density accounts for MD variability across but not within meningioma tumors. C_LIO_LIStructure anisotropy accounts for in-plane FA variability across and within tumors C_LIO_LIVascularization, psammoma bodies, and microcysts influence the MD. C_LIO_LIHigh and low meningioma tumor cell density can yield similar MD. C_LIO_LIFeatures beyond cell density need to be considered when interpreting MD. C_LI