Meeting Banner
Abstract #1956

A Machine Learning Approach for Computer-Aided Detection of Cerebral Microbleed Using High-order Shape Features

Amir Fazlollahi 1,2 , Fabrice Meriaudeau 2 , Luca Giancardo 3 , Christopher C. Rowe 4 , Victor L. Villemagne 4 , Paul Yates 4 , Olivier Salvado 1 , Pierrick Bourgeat 1 , and on behalf of the AIBL Research Group 5

1 CSIRO Preventative Health Flagship, CSIRO Computational Informatics, Brisbane, QLD, Australia, 2 Le2i, University of Burgundy, Le Creusot, France, 3 RLE, Massachusetts Institute of Technology, MA, United States, 4 Department of Nuclear Medicine and Centre for PET, Austin Hospital, Melbourne, VIC, Australia, 5 http://www.aibl.csiro.au/, Australia

Since the incidence of cerebral microbleeds (CMBs) have come to attention as a potential biomarker of cerebrovascular disease and dementia, a computer-aided detection scheme to improve screening is required. In this work, a novel approach of CMB detection in SWIs is presented and compared to visual rating. The proposed method (1) identifies potential CMB candidates with their corresponding bounding boxes using a multi-scale Gaussian technique, (2) extracts a set of robust 3D Radon- and Hessian-based shape descriptors within the bounding boxes, as well as 2D Radon features computed on intensity-projection images, and (3) incorporates a cascade of random forests classifiers to reduce false detection rates.

This abstract and the presentation materials are available to members only; a login is required.

Join Here