Skull removal in MR images using a modified artificial bee colony optimization algorithm

Article type: Research Article

Authors: Taherdangkoo, Mohammad

Affiliations: Department of Artificial Intelligence, Tehran Business School, Tehran, Iran. E-mail: [email protected]

Abstract: Removal of the skull from brain Magnetic Resonance (MR) images is an important preprocessing step required for other image analysis techniques such as brain tissue segmentation. In this paper, we propose a new algorithm based on the Artificial Bee Colony (ABC) optimization algorithm to remove the skull region from brain MR images. We modify the ABC algorithm using a different strategy for initializing the coordinates of scout bees and their direction of search. Moreover, we impose an additional constraint to the ABC algorithm to avoid the creation of discontinuous regions. We found that our algorithm successfully removed all bony skull from a sample of de-identified MR brain images acquired from different model scanners. The obtained results of the proposed algorithm compared with those of previously introduced well known optimization algorithms such as Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO) demonstrate the superior results and computational performance of our algorithm, suggesting its potential for clinical applications.

Keywords: Skull bone region, particle swarm optimization (PSO), ant colony optimization (ACO), artificial bee colony (ABC), MRI segmentation

DOI: 10.3233/THC-140845

Journal: Technology and Health Care, vol. 22, no. 5, pp. 775-784, 2014