Radial glia (RG) are the principal neural stem cells of the embryonic brain, giving rise to neural cell types and serving as mechanical scaffold for neuron migration. Ln2PMMA is an RG biomimetic substrate, of poly methyl methacrylate with 2μm linear topography, that reproduced the surface properties and mechanical anisotropy of the embryonic RG niche. Mechanical signals from ln2PMMA induce cultured astrocytes to dedifferentiate into functional RG. However, the mechanotransducive components underlying this process are largely unknown. We used mouse cortical glial cultures grown on control and ln2PMMA, as a biomechanical tool for in vitro analysis of mechanotransducive mechanisms involved in RG linage differentiation. We developed a unified image segmentation analysis, based on MATLAB algorithms, for the extraction of morphology parameters from subcellular structures. Linked to biochemically identified cells from microscopy images. Second, we developed a multivariate logistic regression model (RGM model) for fitting RG probability based in nuclear morphology parameters and cell density. The application of this model to our experimental data revealed the existence of intrinsic RG nuclear constraints, and that nuclear deformation and changes in nuclear lamins ratio precede the expression of NSC/RG markers. The application of the RGM model to mouse and human neural cells image datasets suggested the evolutionary conservation of RG nuclear constraints.The RGM model application demonstrated that only glial cells expressing nestin/GFAP respond to RG-mimetic mechanical signals, dedifferentiating into RG in ln2PMMA. The evolutive conservation of functional RG nuclear constraints between species made the RGM model an unvaluable tool for improved RG identification.