Contributed talk

Speaker: Burcin Ozcan (Arizona State University)

Location and Time: SAS 221A, Saturday, 2:20–2:50 PM

Title: Image Processing Using Directional Multiscale Representations

Abstract: The automated reconstruction of neuronal morphology is a fundamental task for investigating several problems associated with the nervous system. Furthermore, automated identification of the primary components of a neuron is an essential step in many quantitative studies of neural synaptic networks. However, reconstruction of neurons imaged by optical or fluorescent microscopy is a challenging problem due to the effect of noise and signal degradation. Recent advances in applied harmonic analysis, especially in the area of multiscale representations, offer a variety of techniques for the reconstruction and the extraction of geometric features of neurons. Motivated by the properties of directional multiscale representations, in this talk, we present a multiscale quantitative measure capable of distinguishing isotropic from anisotropic structures and showcase the application of these methods for the detection of somas and of dendritic arbors and axons of neurons.