Method for extracting temperature information in ultrasonic radio frequency signal based on convolutional neural network

The invention discloses a method for extracting temperature information in an ultrasonic radio frequency signal based on a convolutional neural network. The method comprises the steps that the collected radio frequency signal is divided into multiple areas, and area temperature characteristics are obtained to reduce the influence of the 'thermoacoustic lens effect' and the 'sound shielding' effect on an ultrasonic echo signal behind a treatment area; the temperature information is obtained from the ultrasonic radio frequency signal by adopting an algorithm of the convolutional neural network, the problem that the 'thermoacoustic lens effect' and the 'sound shielding' effect generate images on the ultrasonic echo signal behind the treatment area is solved, and the accuracy and stability of temperature measurement are improved; the temperature information of tissue in the ultrasonic images during ablation can be obtained in real time by using a non-interventional mode, and the problem that the tissue in an abla