Real-time 3D ultrasonic scanning conversion method based on GPU platform

Abstract

Then invention discloses a real-time 3D ultrasonic scanning conversion method based on a GPU platform. The method comprises the steps of presetting a 3D image model after scanning; determining an XY plane through an X axis and a Y axis, establishing a series of proxy planes in parallel with a Z axis perpendicular to the XY plane, and dividing the plane into n*n boxes; obtaining the rectangular coordinates of four vertexes of each box on each proxy plane; conducting GPU texture coordinate conversion on the rectangular coordinates of four vertexes and midpoints of each box to obtain the texture coordinates of four vertexes and midpoints of each box; conducting ultrasonic acquisition on the texture coordinates of each box, and mapping the acquired ultrasonic echo strength into brightness and contrast; drawing all proxy planes with the GPU, and meanwhile starting an alpha module of the GPU to conduct hybrid rendering and finish all images. No additional specialized hardware is needed, so this method saves CPU computing resource.

Description

technical field [0001] The invention relates to the technical field of ultrasonic diagnosis, in particular to a real-time 3D ultrasonic scan conversion method based on a GPU platform. Background technique [0002] Scan conversion in 2D mode consumes a lot of computing resources in traditional ultrasound, because scan conversion is performed every time a point is drawn. Some manufacturers use special hardware such as DSP or even FPGA to complete scan conversion. There are some devices that use the CPU's fast algorithm, and even then, a very high-end CPU is required to complete it. Real-time 3D mode, also known as 4D mode, requires scan conversion of volume data, which is very difficult for traditional CPUs, DSPs and even FPGAs. Both are huge challenges. Traditional scan transformation algorithms cannot perform scan transformation on 3D volume data in real time. Contents of the invention [0003] Based on the deficiencies of the prior art, the present invention provides a...

AI Technical Summary

Problems solved by technology

[0002] Scan conversion in 2D mode consumes a lot of computing resources in traditional ultrasound, because scan conversion is performed every time a point is drawn

Both are huge challenges. Traditional scan transformation algorithms cannot scan and transform 3D volume data in real time

Images