Abstract:As for the problems of shifting sensitivity and poor directionality in discrete wavelet transform (DWT), a novel texture retrieval method of medical images by dual-tree complex wavelet transform (DT-CWT) is proposed. In this method, generalized Gaussian statistic model (GCD) is first constructed according to shifting invariability and multidirectional selection of DT-CWT coefficients. Then image features are extracted with the above model. Texture retrieval is carried out with improved the log-likelihood (ILL) similarity measurement algorithm. Experimental results show that the veracity of the proposed method is better than the traditional Gabor and DWT algorithm and up to 82.8%. So it is valuable for aid-diagnosis in the future.

A Novel Medical Image Texture Retrieval Method By Dual-Tree Complex Wavelet Transform