@article{2816,
  author = {Xiaojun Wang, Yaru Zhu},
  title = {Curvelet Transformed Medical Image Fusion through Regional Entropy Distribution Characteristics  },
  journal = {Journal of Multimedia Processing and Technologies},
  year = {2019},
  volume = {10},
  number = {3},
  doi = {https://10.6025/jmpt/2019/10/3/87-103},
  url = {http://www.dline.info/jmpt/fulltext/v10n3/jmptv10n3_1.pdf},
  abstract = {Multi-modal digital medical image fusion has drawn extensive attentions for integrating more information into one image. In this article, novel multi-resolution curvelet entropy distribution method is proposed to discriminate the image curvelet-transformed coefficients. Fusion weight determination for individual source information has been emphasized on, which is represented through analyzing the coefficient matrix in curvelet-transformed scale. Shannon entropy has been applied as the essential fusion weight criteria after counting the probabilities of the coefficients through different-sized masks at every scale. Fusion criterion of entropy distribution variance in different sized neighborhood regions is also combined with the entropy idea to preserve more details of the source images. After operating fusion on brain CT/MRI data sets, quantitative evaluation shows that the proposed approach based on 3×3 regional distribution of 3×3 masked entropy can provide more satisfactory fusion effect. And the wider mask or region can sensitively extract more CT information for fusion, though deteriorated the total fusion quality. The proposed algorithm is beneficial for multi-modal medical image comprehension to heighten the clinical diagnosis accuracy.},
}