Title: Modified dual channel PCNN algorithm with hybrid edge enhancement approach for multimodality brain image fusion

Authors: S. Kavitha; B. Bharathi; P. Sasikala; D. Chandraleka; V. Ashwini

Addresses: Department of CSE, SSN College of Engineering, Kalavakkam, Chennai 603110, India ' Department of CSE, SSN College of Engineering, Kalavakkam, Chennai 603110, India ' BA Continuum India Pvt Ltd., Chennai 600113, India ' Accenture Service Private Limited, Chennai, India ' Accenture Service Private Limited, Bangalore, India

Abstract: Image fusion plays a vital role for many applications in the field of computer vision, remote sensing, image robotics and medical imaging. This paper is focused on the fusion of multimodality brain images, using a Modified Dual Channel Pulse Coupled Neural Network (MDCPCNN) algorithm along with hybrid edge enhancement approach namely Canny and Ant Colony Optimisation (ACO). In general, the fused image derived from PCNN algorithm has better tissue information and contrast even though a loss occurs in edge information. To overcome this drawback, a hybrid edge enhancement approach is proposed and applied along with MDCPCNN fusion algorithm. The proposed model is validated using four datasets of brain images from different modality, with the subjective and objective measures. The fused image constructed from the proposed model consistently retains the edge information, contrast and texture than the existing PCNN's without false information or information loss.

Keywords: image fusion; dual channel pulse coupled neural network; multimodality brain images; Canny algorithm; ant colony optimisation.

DOI: 10.1504/IJBET.2018.094727

International Journal of Biomedical Engineering and Technology, 2018 Vol.28 No.2, pp.120 - 146

Received: 01 Jul 2016
Accepted: 28 Nov 2016

Published online: 30 Aug 2018 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article