Brain MRI Tissue Classification Based Fuzzy Clustering with Competitive Learning

作者：Zhang, Yuanpeng; Tian, Fei; Wu, Huiqun; Geng, Xingyun; Qian, Danmin; Dong, Jiancheng; Jiang, Kui^*

来源：Journal of Medical Imaging and Health Informatics, 2017, 7(7): 1654-1659.

DOI：10.1166/jmihi.2017.2181

• Brain MRI Tissue Classification
• Fuzzy Clustering
• Competitive Learning
• Anti-Noise
• Multi-Medoid

Brain MRI tissue classification aims to partition it to the gray matter, white matter and cerebrospinal fluid. MRIs often contain noises due to the complexity of the human anatomy, the irregularity of the soft tissue and other factors. Noises can have the negative influences on brain tissues classification, so how to develop a brain tissues classification approach with anti-noise capability and promising accuracy is very important. In this study, we propose a novel fuzzy clustering algorithm based on multi-medoid representative strategy and improved fuzzy partitions. The multi-medoid representative strategy can help capture the inner cluster structure more perfectly and the improved fuzzy partitions can guarantee the capacity of anti-noise. Experimental results on texture images and brain MRIs show that the proposed algorithm performs better than other comparison approaches.

• 出版日期2017-11
• 单位南通大学