CN108309350A - The revertive delay computational methods accelerated applied to medical ultrasonic imaging system - Google Patents

Abstract

The invention discloses a reverse delay calculation method applied to the acceleration of a medical ultrasound imaging system. The beamforming algorithm using the reverse delay calculation method in the present invention is realized based on the inverse process of the traditional delay-and-overlay beamforming algorithm, and the delay amount is deduced by the position of the data point in the ultrasonic echo data, and then the position of the data point in the image is derived , and then directly superimposed, avoiding the time-consuming root sign and division calculation. The beamforming algorithm using the reverse delay calculation method in the present invention has lower computational complexity, and uses more abundant original ultrasound data in the calculation process, and can complete the complex high-definition medical ultrasound image algorithm in a very short time. Computing can meet the demand for real-time and high-definition presentation of medical ultrasound images. The invention does not affect the physical structure of the device, and can directly update the algorithm in the original device to achieve acceleration, and improves the imaging definition of the device to a certain extent.

Description

Inverse Delay Calculation Method Applied to Acceleration of Medical Ultrasound Imaging System

technical field

The invention belongs to the field of medical ultrasonic imaging, and in particular relates to a reverse delay calculation method applied to the acceleration of a medical ultrasonic imaging system.

Background technique

In the medical ultrasound imaging system, the physical array element emits ultrasound and receives the echo signal, and then the echo signal data is converted into image data by the imaging algorithm in the computing unit and displayed. At present, in the field of medical detection, medical ultrasound imaging equipment is still widely used to use the delay and stack beamforming algorithm to complete the delay and stack calculation of ultrasound echo data in the central processing unit of ordinary computers to obtain ultrasound images. With the development of medicine, there are further requirements for image clarity and imaging speed. However, the huge imaging data and complex calculation process of the high-definition imaging algorithm make the calculation of the traditional delay-and-add beamforming algorithm in the central processing unit of an ordinary computer unable to meet its needs for high-performance and high-concurrency calculations.

In recent years, the frequency of CPUs has been continuously increasing, and the number of CPU cores on a single chip has continued to increase. Dual-core, quad-core, and even eight-core have appeared. However, there is a bottleneck in the development of multi-core CPUs, and it is difficult to further improve the core frequency and computing throughput. . Some scholars have proposed integrating more computing units to increase the computing power per unit time, but this will increase the size and power consumption of the device, and at the same time increase the cost. On the other hand, the current field-programmable logic gate technology (FPGA) and general-purpose computing graphics processing unit (GPGPU) are developing rapidly, both of which can achieve powerful concurrent computing capabilities, and some scholars have used these two technologies to complete medical research. Acceleration of beamforming algorithms for ultrasound imaging. However, it still takes a long time to transform scientific research results into actual products, and the update of old equipment will also cause waste, and the newly added hardware accelerator will also increase the cost. However, the reverse delay calculation method applied to the acceleration of the medical ultrasound imaging system proposed by the present invention can be directly used in the current medical ultrasound imaging equipment without updating the hardware. Increase the imaging speed.

Contents of the invention

The main purpose of the present invention is to solve the problem of slow speed of current medical ultrasonic imaging equipment. Most of the calculation cores used in current ultrasonic medical imaging equipment are central processing units in ordinary computers. The internal calculation process is simple to implement and can meet the basic requirements of medical ultrasonic imaging. Demand, but the image frame rate is low, and the cost of a single imaging device is high. In order to increase the imaging speed without updating the hardware of medical ultrasound imaging equipment, the present invention improves the delay-and-addition beamforming algorithm itself, and proposes a beamforming algorithm using a reverse delay calculation method. This method greatly reduces the complex calculation operations in the calculation process without reducing the image quality, improves the imaging frame rate of the equipment, and meets the requirement of improving the imaging speed without updating the hardware of the medical ultrasound imaging equipment.

The purpose of the present invention is achieved through the following technical solutions.

The reverse delay calculation method applied to the acceleration of medical ultrasound imaging systems, the beamforming algorithm using the reverse delay calculation method, is based on the inverse process of the traditional delay-and-addition beamforming algorithm, and is deduced by the position of the data point in the ultrasound echo data The amount of delay then deduces the position of the data point in the image, and then superimposes directly, avoiding the time-consuming root sign and division calculation.

Furthermore, the inverse delay calculation method is implemented based on the inverse process of the traditional delay-and-addition beamforming algorithm. First, the delay of the received echo ultrasonic signal is calculated, and then the delayed echo signal is superimposed to obtain an enhanced echo signal output, that is, ultrasonic The brightness value of a pixel in the imaging image.

Further, the input data of the reverse delay calculation method is the three-dimensional array ultrasonic echo data signal, and signal(i,k,d) is defined as the value of the dth signal received by the ith ultrasonic transmission and the kth receiving array element ;The output data is a two-dimensional image image, and image(i,j) is defined as the brightness value of the pixel point in the i-th column and the j-th row in the image.

Further, it specifically includes the following steps:

(1) Set the variable i as the loop variable of the column number, and set the initial value to 0;

(2) Setting variable j as the loop variable of row number, setting initial value 0;

(3) The variable k is set to be a loop variable of the number of array elements, and the initial value is set to 0;

(4) Calculate the delay time according to the variables i, j, k, and the corresponding array subscript d;

(5) Perform superposition, the variable image(i,j) superimposes the signal signal(i,k,d);

(6) Self-incrementing variable k, if the value of variable k is less than the total array element number, then jump to step (3);

(7) Self-increment variable j, if the value of variable j is less than the total number of rows, then jump to step (2);

(8) Self-incrementing variable i, if the value of variable i is less than the total number of columns, then jump to step (1);

(9) The algorithm ends, and the image image is returned.

Further, for the case where the subscript position of the data point in the echo data array is d, assuming that the position of the data point in the image is j, according to the formula:

Available:

Among them, W _pixel and H _pixel are the horizontal width and vertical height corresponding to the pixels of the generated image, respectively. Here, the default emission element and the pixel are strictly aligned in the vertical direction. In summary, the calculation formula of d can be obtained as follows:

In the ultrasonic imaging equipment, f _s , c, W _pixel and H _pixel are all constants; the calculation process includes two multiplications and one division; for dividing d, d is a positive integer.

Compared with the prior art, the advantages of the present invention are mainly reflected in two aspects: the beamforming algorithm using the reverse delay calculation method in the present invention has lower computational complexity, and uses a richer original Ultrasound data can complete complex calculations of high-definition medical ultrasound image algorithms in a very short period of time, and can meet the needs of real-time and high-definition medical ultrasound images. The invention does not affect the physical structure of the device, and can directly update the algorithm in the original device to achieve acceleration, and improves the imaging definition of the device to a certain extent.

Description of drawings

Fig. 1 is a schematic diagram of the ultrasound path from transmission to reception of medical ultrasound imaging equipment.

Figure 2 is an example diagram of an output image calculated using a traditional delay-and-addition beamforming algorithm.

Fig. 3 is an example diagram of the output image calculated by the beamforming algorithm using the reverse delay calculation method.

Detailed ways

The specific implementation of the present invention will be further described below in conjunction with the accompanying drawings and examples, but the implementation and protection of the present invention are not limited thereto. It should be pointed out that, if there are no specific details below, those skilled in the art can refer to the prior art to realize.

The reverse delay calculation method applied to the acceleration of medical ultrasound imaging systems, the beamforming algorithm using the reverse delay calculation method, is based on the inverse process of the traditional delay-and-addition beamforming algorithm, and is deduced by the position of the data point in the ultrasound echo data The amount of delay then deduces the position of the data point in the image, and then superimposes directly, avoiding the time-consuming root sign and division calculation. The reverse delay calculation method is based on the inverse process of the traditional delay-and-addition beamforming algorithm. First, the delay of the received echo ultrasonic signal is calculated, and then the delayed echo signal is superimposed to obtain an enhanced echo signal output, that is, the echo signal output in the ultrasonic imaging image. The brightness value of a pixel.

The input data of the reverse delay calculation method is the three-dimensional array ultrasonic echo data signal, and signal(i,k,d) is defined as the value of the dth signal received by the ith ultrasonic transmission and the kth receiving array element; the output data is a two-dimensional image image, and image(i,j) is defined as the brightness value of the pixel point in the i-th column and j-th row in the image.

The algorithm specifically includes the following steps:

(1) Set the variable i as the loop variable of the column number, and set the initial value to 0;

(2) Setting variable j as the loop variable of row number, setting initial value 0;

(3) The variable k is set to be a loop variable of the number of array elements, and the initial value is set to 0;

(4) Calculate the delay time according to the variables i, j, k, and the corresponding array subscript d;

(5) Perform superposition, the variable image(i,j) superimposes the signal signal(i,k,d);

(6) Self-incrementing variable k, if the value of variable k is less than the total array element number, then jump to step (3);

(7) Self-increment variable j, if the value of variable j is less than the total number of rows, then jump to step (2);

(8) Self-incrementing variable i, if the value of variable i is less than the total number of columns, then jump to step (1);

(9) The algorithm ends, and the image image is returned.

The beamforming algorithm using the reverse delay calculation method proposed by the present invention will eventually be applied to medical ultrasound imaging systems. Here, a simulated medical ultrasound imaging system is constructed through three modules, which are data simulation module, core calculation module and display image module.

The delay-and-add algorithm consists of three loops, as shown in the algorithm description above. The time complexity of this algorithm is:

Complexity=O(LC×RC×RA),

The variable LC (Line Count) represents the number of vertical image columns in the output image, the variable RC (Rows Count) represents the number of rows in the output image, and the variable PA (Probe Amount) represents the number of probe array elements that receive the echo ultrasonic signal.

In the core code of the traditional delay-and-add beamforming algorithm, the core operation is to add the delayed input data. For a specific pixel point (i, j) in the final generated image, there is a superposition of secondary input data signals, and the superposition formula is as follows:

Among them, i, j and k are the accumulative variables of LC, RC and PA respectively, and d is the subscript value mapped to the ultrasonic echo data array by the delay amount, which can be obtained through the correlation calculation of specific i, j and k variables, Calculated as follows:

d＝t _flight *f _s

Where f _s is the sampling frequency of the equipment, and t _flight is the time spent from transmitting to receiving the corresponding ultrasonic wave in the ultrasonic echo data signal, the calculation formula is as follows:

Among them, d _e and d _r are respectively the physical distances of the corresponding ultrasonic waves in the ultrasonic echo data signal from transmission to reception, as shown in Fig. 1 . d _e and d _r can be obtained from the transmitting position, receiving position and object position, and the calculation formula is as follows:

d _e = H _pixel j

Among them, W _pixel and H _pixel are the horizontal width and vertical height corresponding to the pixels of the generated image, respectively. Here, the default emission element and the pixel are strictly aligned in the vertical direction. In summary, the calculation formula of d can be obtained as follows:

In an ultrasound imaging device, f _s , c, W _pixel and H _pixel are all constants. Furthermore, generating an ultrasound image requires LC×RC×RA calculations, and each calculation needs to execute the above formula once, including multiple square operations and a square root operation, and the overall calculation amount is huge.

For this reason, the present invention proposes a reverse delay calculation method. The traditional delay-and-stack beamforming algorithm derives the delay from the position in the image and then derives the position of the required data point in the echo data array, while the reverse delay calculation method derives the delay from the position of the data point in the echo data array Then the position of the data point in the image is derived. For the case where the subscript position of the data point in the echo data array is d, suppose the position of the data point in the image is j, according to the formula:

Available:

Similarly, f _s , c, W _pixel , and H _pixel are all constants, and the calculation process includes two multiplications and one division. For dividing d, since d is a positive integer within a certain range, this operation can be transformed into a table lookup operation and a multiplication. Furthermore, this scheme avoids the time-consuming calculation of root number and division, and greatly speeds up the calculation speed.

This example adopts the following modules to realize the method.

1. Data simulation module

Use the Field II simulator to simulate the propagation process of ultrasound in ultrasound imaging and obtain simulation data; in the data simulation module, first simulate the corresponding simulation physical data according to the corresponding configuration of the actual ultrasound imaging equipment, create transmitting and receiving array elements, and create simulated detection objects , and then simulate transmitting and receiving echo data one by one according to the scan line.

2. Core computing module

The core imaging module is implemented on an ordinary computer using an ultrasound imaging algorithm. This example uses a computer with Intel i5-4590 16GB RAM and Ubuntu 14.04 installed to configure the core computing module. After the module collects the ultrasonic echo data, it maps the data into the image according to the amount of delay, and finally outputs the image data. The algorithm is implemented using the beamforming algorithm of the reverse delay calculation method.

3. Display image module

In the simulated medical ultrasound imaging system, after the core calculation module obtains the pixel data, the display image module uses the corresponding decoding program to perform operations such as Hilbert change, logarithmic compression, gray scale correction, and image depth and width calculations on the data. Finally, output the image-related data into the corresponding coordinate system, display the image on the screen, or store the image in a file.

In the simulated medical ultrasound imaging system, the traditional delay-and-add beamforming algorithm and the beamforming algorithm using the reverse delay calculation method are respectively used to calculate the same ultrasound echo data for comparison.

Scenario and Performance Verification:

In the experiment, the input signal data of 10 medical ultrasound images were loaded into the simulated medical ultrasound imaging system, and the input signal data of each image corresponded to one ultrasound image. The core calculation module is called, and the traditional delay-and-addition beamforming algorithm and the beamforming algorithm using the reverse delay calculation method are respectively used to complete the calculation of the image. The images obtained after calculation are shown in Figure 2 and Figure 3. The time from the beginning of the calculation to the end of the calculation of the two algorithms is actually measured, and the average value of each time parameter is obtained through multiple experiments. The final result is that the traditional delay-and-addition beamforming algorithm takes 105ms, and the beamforming using the reverse delay calculation method The algorithm takes 25ms.

From the above data, it can be concluded that the reverse delay calculation method applied to the acceleration of the medical ultrasound imaging system in this example achieves a speedup ratio of more than 4 times over the traditional delay-and-addition beamforming algorithm without degrading the image quality.

This example describes the design and evaluation of the acceleration of the medical ultrasound imaging process by the reverse delay calculation method applied to the acceleration of the medical ultrasound imaging system in the simulated medical ultrasound imaging system. The experimental evaluation results show that the beamforming algorithm using the reverse delay calculation method is more than 4 times faster than the traditional delay-and-stack beamforming algorithm, and the theoretical value of the calculation frame rate reaches 40 frames per second. The demand for high frame rate of real-time imaging.

The method of the present invention is applied to a conventional medical ultrasound imaging system. By improving the algorithm itself, the ultrasound imaging process can be accelerated, and the imaging frame rate of the medical ultrasound imaging equipment can be improved without changing the physical equipment and without reducing the imaging quality. It is widely used in the currently widely used medical imaging equipment based on the calculation of the central processing unit of ordinary computers, the medical imaging equipment based on the calculation of the ARM processor on the embedded mobile device, and other medical imaging systems based on GPU or FPGA acceleration.

Claims (5)

1. the revertive delay computational methods accelerated applied to medical ultrasonic imaging system, it is characterised in that calculated using revertive delay The beamforming algorithm of method, the inverse process based on traditional delay stack beamforming algorithm are realized, by data point super Position in sound echo data derives retardation and then derives the position of the data point in the picture, and then is directly overlapped, Evaded take open radical sign and division calculation.

2. the revertive delay computational methods according to claim 1 accelerated applied to medical ultrasonic imaging system, feature It is that inverse process of the revertive delay computational methods based on traditional delay stack beamforming algorithm is realized, calculates receive first Echo ultrasonic signal retardation, and then the echo signal after delay is superimposed to the echo signal output enhanced, i.e., it is super The brightness value of a pixel in acoustic imaging image.

3. the revertive delay computational methods according to claim 2 accelerated applied to medical ultrasonic imaging system, feature It is that the input data of revertive delay computational methods is three-dimensional array ultrasonic echo data signal, defines signal (i, k, d) For ith ultrasound emission, k-th of value for receiving d-th of signal that array element receives；Output data is two dimensional image image, definition Image (i, j) is the brightness value of the pixel of the i-th row jth row in image.

4. the revertive delay computational methods according to claim 3 accelerated applied to medical ultrasonic imaging system, feature It is to specifically comprise the following steps：

(1) variable i is set as the cyclic variable of columns, sets initial value 0；

(2) variable j is set as the cyclic variable of line number, sets initial value 0；

(3) variable k is set as the cyclic variable of array number, sets initial value 0；

(4) according to variable i, j, the time of k computing relays and corresponding array index d；

(5) superposition, variable i mage (i, j) superposed signal signal (i, k, d) are executed；

(6) from variable k is increased, if the value of variable k is less than total array number, step (3) is jumped to；

(7) from variable j is increased, if the value of variable j is less than total line number, step (2) is jumped to；

(8) from variable i is increased, if the value of variable i is less than total columns, step (1) is jumped to；

(9) algorithm terminates, and returns to image image.

5. the revertive delay computational methods according to claim 3 accelerated applied to medical ultrasonic imaging system, feature It is the case where subscript position for data point in echo data array is d, it is assumed that the position of the data point in the picture For j, according to formula：

It can obtain：

Wherein W_pixelAnd H_pixelThe corresponding horizontal width of image slices vegetarian refreshments and vertical height are respectively generated, the battle array of acquiescence transmitting here Member and pixel close alignment in vertical direction.The calculation formula that d can to sum up be obtained is：

In supersonic imaging apparatus, f_s、c、W_pixelAnd H_pixelIt is constant；Calculating process includes multiplication and a division twice； For removing d, d is positive integer.