Soft tissue elasticity and viscoelasticity characterization method and device based on ultrasound elastography

Abstract

The invention discloses a soft tissue elasticity and viscoelasticity characterization method and device based on ultrasonic elastography, wherein the method includes: simulating ultrasonic elastography by finite element method to obtain training data of artificial neural network; Data pre-processing, so that the training data contains effective information that meets preset conditions, and obtain a pre-processed feature map; train the pre-processed feature map to obtain a representation that satisfies the preset neural network effect Neural Networks of Soft Tissue Elasticity and Viscoelasticity. This method has the advantages of high characterization accuracy, good robust performance, and large optimization space, and has good application prospects and room for improvement.

Description

technical field [0001] The invention relates to the technical field of medical imaging, in particular to a soft tissue elasticity and viscoelasticity characterization method and device based on ultrasonic elastography. Background technique [0002] Many diseases of the human body, such as liver cirrhosis and tumors, are often accompanied by changes in the mechanical properties of human tissue. Characterizing the mechanical properties of human tissues in vivo is of great significance for the diagnosis and treatment of many diseases. Ultrasound elastography is an emerging means of characterizing the mechanical properties of materials. Due to its non-destructive and non-invasive characteristics, it has received extensive attention in disease diagnosis in recent years. Due to the complexity of human tissue, the specific implementation of this method in some application scenarios remains to be studied. [0003] Machine learning is a relatively general method of using computers ...

AI Technical Summary

Problems solved by technology

This also means that methods that increase effective data utilization have the potential to greatly improve the accuracy of representations

[0012] 2. The characterization results are not stable enough

Ultrasound elastography often has a certain level of random noise in the actual use process. These noises may have unpredictable effects on some characterization methods, resulting in unstable characterization results and large fluctuations in repeated measurement results for the same part.

[0013] 3. The scope of application of the method is narrow

The above characterization methods are often aimed at the constitutive relationship of a specific material and specific form, and it is difficult to migrate to the characterization of materials with other properties

Images