本文提出了一种基于功率谱信息熵和GK模糊聚类相结合的方法, 应用于生物组织变性识别判定中。利用高强度聚焦超声辐照离体新鲜猪肉来改变其特性, 并用热电偶测量声焦区温度, 同时采集了不同温度时的超声回波信号。通过对采集的信号进行截取, 研究了分段数对功率谱信息熵辨识性能的影响。研究发现当分段数为26至32时, 功率谱信息熵的准确度、灵敏度和特异度均有较高的值。本研究计算了分段数为30时的功率谱信息熵, 此时的变性组织对应的功率谱信息熵比未变性组织对应的功率谱信息熵平均高出约0.094, 大约为7.99%。采用功率谱信息熵作为特征参数时, GK模糊聚类效果优于模糊C均值聚类；采用GK模糊聚类方式, 功率谱信息熵比小波熵具有更高的辨识率。

A method combining power spectrum information entropy and GK fuzzy clustering was proposed to recognize the denatured biological tissue. The characteristics of porcine muscle tissues were changed by high-intensity focused ultrasound irradiation. In this experiment, the thermistor was used to measure the temperature in the acoustic focal zone, and the ultrasonic echo signals at different temperatures were collected. In the data process, the original signal was intercepted and the influence of the segmentation number on the identification performance of power spectral information entropy was discussed. It was found that while the number of segments is 26 to 32, the accuracy, sensitivity and specificity of the power spectrum information entropy are higher. The power spectrum information entropy being calculated as the number of segments is 30 .The average value of the power spectrum information entropy from the signal corresponding to denatured tissues was about 0.094 higher than normal tissues, which was about 7.99%. When power spectrum information entropy is used as the characteristic parameter, the GK fuzzy clustering effect is better than the fuzzy C-means clustering. With GK fuzzy clustering, power spectrum information entropy has higher recognition rate than wavelet entropy.