A133 - 3D Image Analysis Flash Tutorials: Neuron tracing using Machine Learning

Anne Cantereau becq (anne.cantereau@univ-poitiers.fr)
Sylvain Derossi (sylvain.derossi@mri.cnrs.fr)

This flash workshop is the 6th of 6 different others, presenting 6 different applications of the commercial 3D software Imaris, from visualization, to network description. Morphology of neurons varies during embryonic development, trainings, pathologic situations, pharmacologic treatment… Fine analysis of their shape with data extraction allows statistical studies indispensable to compare and characterize these morphological differences or changes. The goal of this Flash tutorial is to create skeleton of isolated neurons (in that specific case, neurons were stained with fluorescent dyes injected during electrophysiological recording). Participants will follow an analysis pipeline that uses automatic process (using machine learning) in order to create the most similar structure compared to superposed 3D fluorescent image. Once completed, large number of statistical data can be measured to describe the complex shape of the neuron: diameters, lengths, branching points, straightness, spines… The workshop will be a hands-on-session with a computer for each attendant, working with the same datasets (3D confocal images). They will follow a step by step guiding workflow.