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

Amir Fazlollahi (1) , Fabrice Meriaudeau (1) , Luca Giancardo (1) , Christopher Rowe (2, 3) , Victor L Villemagne (4, 5) , Paul Yates (2) , Olivier Salvado (6) , Pierrick Bourgeat (6)

1 Le2i - Laboratoire Electronique, Informatique et Image [UMR6306]
2 Department of Nuclear Medicine
3 Department of Medecine
4 Department of Nuclear Medicine
5 The Mental Health Research Institute
6 CSIRO ICT Centre - CSIRO Information and Commuciation Technologies

This paper presents a novel machine learning approach for computer-aided detection of microbleeds in SWI. The major contributions are: identifying microbleed extent in order to extract proper cubic regions-of-interest (ROI) containing the structure, (2) extracting a set of robust 3- dimensional (3D) Radon- and Hessian-based shape descriptors within the ROIs as well as 2D Radon features computed on intensity-projection images of the corresponding ROIs, and (3) incorporating a cascade of random forests (RF) classifiers to iteratively reduce false detection rates while maintaining a high sensitivity.

Domaines: Traitement des images [eess.IV]

Amir Fazlollahi, Fabrice Meriaudeau, Luca Giancardo, Christopher Rowe, Victor L Villemagne, et al.. A Machine Learning Approach for Computer-Aided Detection of Cerebral Microbleed Using High-order Shape Features. ISMRM 2014, May 2014, Milan, Italy. pp.1956. ⟨hal-00989923⟩