METHODS FOR QUANTITATIVE ANALYSIS OF BRAIN MICRO-ARCHITECTURE: APPLICATION TO ANIMAL MODELS OF AUTISM AND STUDY OF HUMAN CLAUSTRAL ORGANIZATION

One of the grand challenges of digital imaging in the _eld of neuroanatomy is the ability to extensively quantify anatomical structures and thus investigate the brain's structure-function relationship in great detail. In the light of this challenge, my PhD thesis aims to investigate the brain's micro-structure to obtain faithful and reproducible information on neuron morphology within their native three-dimensional arrangement. A rigorous work-ow was designed, that integrates delipidation methods, advanced imaging techniques and image processing algorithms to better understand neural micro-structure and its contribution to neural function. In particular, the work-ow provides i) the optimization and standardization, through the quantification of non-invasive and macroscopic indices, of the CLARITY2 protocol, a tissue clarification method which eliminates lipids and reduces tissue scattering from thick brain slices, ii) the development of a Smart Region Growing (SmRG) algorithm for single neuron tracing from confocal three-dimensional datasets representing densely packed neurons within the brain, and iii) the implementation of N3MO, an open-source tool for quantitative morphometric extraction and multivariate analysis of neurons. The work-ow was then applied to two case studies. The first aims to investigate sexual dimorphism in animal models of autism, because of the unbalanced incidence of the disorder in males and females. The second study is focused on the assessment of neural organization in the claustrum, giving the basis for distinguishing the di_erent neuronal types with respect to their shape.