LAUSR

In the course of high-intensity focused ultrasound (HIFU) treatment, the capacity to precisely recognize biological tissue that has been denatured is crucial to ensuring the security and availability of HIFU treatment. Multi-scale permutation entropy (MPE) and its variant multi-scale weighted-permutation entropy (MWPE), as common methods to measure the complexity of nonlinear time series, are often used to recognize denatured biological tissue during HIFU treatment. In order to improve the inevitable disadvantages of MPE and MWPE in some cases, a new complexity method called multi-scale phase weighted-permutation entropy (MPWPE) is put forward. The proposed MPWPE improves MPE and MWPE by adding phase information through the Hilbert transform. The simulated analysis result indicates that the MPWPE can detect more dynamic changes in the synthetic signal compared with MPE and MWPE. Finally, based on the key MPWPE feature extraction algorithm, a novel intelligent biological tissue denatured recognition technology combined with the classifier is proposed. The actual HIFU echo signals of biological tissues are employed to verify the effectiveness of the proposed method. The results show that compared with MPE and MWPE, the MPWPE features can distinguish non-denatured and denatured tissues at multi scales with better performance and achieve higher recognition accuracy.