A Novel Nature Inspired Fuzzy Tsallis Entropy Segmentation of Magnetic Resonance Images

Abstract

Medical imaging comprises a large number of non-invasive techniques to assist diagnosis or treatment of different medical conditions. Magnetic resonance imaging (MRI) is a noninvasive imaging technique which has a wide range of applications such as neuroimaging, cardiac magnetic resonance imaging, spinal imaging, liver and gastrointestinal imaging, functional magnetic resonance imaging etc. Neuroimaging helps the physician to investigate functions of the brain or neurological disorder. In this paper, a novel quantum optimization technique referred as Improved Quantum Particle Swarm Optimization (IQPSO) for optimizing three levels Fuzzy Tsallis Entropy (FTE) is proposed for segmenting brain MR images. Fuzzy Tsallis Entropy (FTE) is a fuzzy based threshold method segments the image based on entropies. The proposed method is validated on the data sets obtained from the Chennai scan center and is compared with other standard optimization methods such as Particle Swarm Optimization (PSO) and Quantum Particle Swarm Optimization (QPSO). The analysis shows that FTE optimized using Improved QPSO produces maximum entropy of 33.7623, 32.9868, 36.0231, 36.1231, 40.9789, 40.9789, 40.9789, 41.231 , 41.2314, 43.6994 for the brain MRI slices 13,12,11,10,9,8,7,6,5 and 1 of patient 1 and 13.006, 13.126, 13.126, 12.673, 12.295, 12.202, 11.229 for the brain MRI slices 1, 2, 3, 4, 5, 6 and 7 of patient 2. The obtained results are better than traditional Particle Swarm Optimization, Quantum Particle Swarm Optimization and Fuzzy Tsallis Entropy (FTE).