Meeting Banner
Abstract #0949

Classification of Cortical Layers at Sub-Pixel Resolution

Shlomi Lifshits1, Daniel Barazany2, Saharon Rosset1, Yaniv Assaf2

1Department of Statistics and Operations Research, Tel-Aviv University, Tel-Aviv, Israel; 2Department of Neurobiology, Tel-Aviv University, Tel-Aviv, Israel

We train a classification model for prediction of the cortical layer based on inversion recovery data. We show how partial volume artifact can be minimized by enhancing the resolution using the generated probability maps.

Keywords

ability acquisition addition adjacent aged allow architecture artifact artifacts axon became beck belongs border brain calculating cellular characterize class classes classification color component composition compute computed conditions cortex cortical covering definition described detect detecting development deviation diffusion distinct distribution divided edges enables enhance enhanced enhancement entire example explore extreme extremely fall features field fitting fraction free function gray hard hemisphere histogram identifies identifying implementation imply imposing in vivo included increasing indeed inspecting introduce introduced inversion layer layers learning lengthy limited logistic majority maps masked micro microns microscope minimize minimizing mixture model need neurobiology operations optimization original overcomes partial pattern popular post posterior predicated predicts preserves principal probability problem process processing prominent protocol quantification quantify recently reconstruct recovery reflected regression regularization regularized represents resolution resolved reveal revealed revealing scale scanner seems segmentation setup short solved solving statistical statistics structures subjects sufficient surfer tissue training trans turns underwent variation varied virtual visualization volume water width