Image tracking method and device based on multi-neighborhood-aided two-dimensional ultrasonic deformed microstructure

Abstract

The invention designs an image tracking method based on multi-neighborhood-aided two-dimensional ultrasonic deformed microstructure, which comprises the steps as follows: a two-dimensional ultrasonic cardiogram of a tested heart object is collected under a high frame frequency condition and used for offline analysis firstly; then any visible image is used as a selected initial frame of a region of interest (ROI) so that the ROI includes cardiac muscular tissues, simultaneously the interferences of grey scale signals is eliminated; after the ROI of the initial frame on an initial image I is completely selected, the original ROI is set as a center; moreover, the ROI and eight areas coating the surrounding of the original ROI and having same sizes are used as a new searching source region, the tracking vectors of all subregions with the sizes the same as the size of the original ROI in the searching area are traversed through an SAD (Sum of Absolute Differences) arithmetic; and weighted average is implemented on each tracking vector finally, thus obtaining the tracking result of the original ROI. Compared with the traditional SAD tracking method, the neighborhood tracking result of the ROI is sufficiently considered; therefore, the image tracking method based on multi-neighborhood-aided two-dimensional ultrasonic deformed microstructure has excellent noise immunity, and is especially beneficial for movement estimation and parameter calculation on movement issues.

Description

technical field [0001] The invention relates to an ultrasonic image tracking technology in the field of medical equipment, in particular to a multi-neighborhood-assisted two-dimensional ultrasonic deformation tissue image tracking method and device. Background technique [0002] Tissue tracking of ultrasonic cardiac images is a new technology based on tissue Doppler imaging. It can evaluate the systolic and diastolic functions of the myocardium by measuring the atrioventricular plane displacement. In addition to evaluating the overall function of the heart, it is mainly used to evaluate Therefore, it is of great significance in the early detection of myocardial ischemia, the evaluation of the viability of infarcted myocardium, and the prediction of the prognosis of heart disease. [0003] Existing two-dimensional tissue tracking techniques mainly include fast matching method, pixel recursive method, frequency-domain-based method, feature-based matching method, etc. Block mat...

AI Technical Summary

Problems solved by technology

Some scholars propose a multi-layer search algorithm to speed up the search, and some use Fourier transform-based analysis methods. However, the above methods use a basic source region of interest (ROI) as the tracking area, and do not specifically consider the neighborhood impact problem

However, the actual echocardiogram is noisy. Even if the tracking of the source ROI area is accurate, due to the influence of noise, the precise tracking deviates from the actual tissue movement. Therefore, there are major defects in the tracking of a single area, so It must be combined with the tracking results of its neighbors to reduce the impact of noise

Images