CN114129130A

CN114129130A - Photoacoustic image back projection reconstruction method based on single address LUT table

Info

Publication number: CN114129130A
Application number: CN202111268589.XA
Authority: CN
Inventors: 王冬芳; 皮明超; 余宁梅
Original assignee: Xian University of Technology
Current assignee: Xian University of Technology
Priority date: 2021-10-29
Filing date: 2021-10-29
Publication date: 2022-03-04
Anticipated expiration: 2041-10-29
Also published as: CN114129130B

Abstract

The invention discloses a photoacoustic image back projection reconstruction method based on a single address LUT table, which specifically comprises the following steps: pre-constructing an LUT lookup table of the first probe according to a back projection algorithm according to the adopted frequency of the first probe, the total effective signal time, the propagation speed of the ultrasound in the biological tissue and the size or resolution of a reconstructed image; reconstructing LUT lookup tables of other remaining probes according to the distance between the two probes in the linear array ultrasonic detector, and reconstructing a total LUT lookup table; when the photoacoustic image is reconstructed, a lookup table corresponding to the probe is derived according to data on the probe to be reconstructed; and finding out the coordinates of the pixel points of the corresponding reconstructed image according to the lookup table corresponding to each probe, and then accumulating the amplitude of the ultrasonic signal of the point on the corresponding probe to the pixel value of the point of the reconstructed image to reconstruct the required photoacoustic image. The invention solves the problems of large calculation amount and large storage amount of a plurality of LUT lookup tables in the prior art.

Description

Photoacoustic image back projection reconstruction method based on single address LUT table

Technical Field

The invention belongs to the technical field of photoacoustic imaging methods, and relates to a photoacoustic image back projection reconstruction method based on a single-address LUT (look-up table).

Background

The basic principle of Photo Acoustic Imaging (PAI)) is that when non-ionizing radiation (e.g. laser or microwave pulses) is used to irradiate a tumor in the human body, the tumor absorbs energy and is converted into thermal energy, and the temperature rises instantaneously. Because heat conduction is much slower and negligible relative to temperature rise, megahertz-level ultrasound, i.e., photoacoustic waves, are excited inside the tissue. The photoacoustic wave carries the electromagnetic absorption distribution characteristics of the tissue, and after the ultrasonic transducer detects the photoacoustic wave, the electromagnetic absorption distribution image in the tissue can be calculated by adopting a corresponding image reconstruction algorithm. The photoacoustic effect is discovered as early as 1880, but only in the 90 s of the 20 th century, the photoacoustic effect is tried to be applied to medical imaging, and in nature 2003, Wang and the like report that pulse lasers are utilized to clearly detect the distribution of cerebral vessels of living mice by adopting a photoacoustic imaging technology, clear imaging of brain parenchyma lesion is obtained, and the development of the nanosecond laser pulse induced photoacoustic imaging technology is promoted.

Due to the fact that photoacoustic imaging is different from optical imaging, the photoacoustic imaging mainly depends on that biomolecules absorb laser pulses to instantly generate thermoelastic expansion to generate sound waves, and scattering of sound signals in human tissues is 2-3 orders of magnitude lower than that of light, so that the photoacoustic imaging has imaging depth and spatial resolution equivalent to that of ultrasonic imaging, for example, the resolution is about 0.8mm under the detection depth of 5 cm. Can realize multi-scale biomedical imaging with different resolutions from micron to millimeter from cell level to small animal brain, human lymph node cancer, breast cancer, prostate cancer and the like.

The relatively classical photoacoustic imaging algorithm is a back projection algorithm, as shown in fig. 1, the left side is a reconstructed image, the right side is biological tissue, a tumor tissue point M is arranged inside the biological tissue, and at least more than three probes of an ultrasonic detector are arranged on the boundary of the biological tissue. (note: common ultrasonic detectors are linear array detectors, and the number of probes is more than 100.)

As can be seen from fig. 1: the distances from the three probes to the tissue point are different, the approximate position of the tissue point can be determined by the intersection point of the three arc lines, and the distance from each probe to the tissue point can be obtained by corresponding calculation. And in the back projection process, the distance between the detector and the biological tissue is calculated according to the arrival time of each point on the detector, so that the convolution process is carried out on the reconstructed image according to the distance.

In order to provide quality of a reconstructed photoacoustic image, signals are generally acquired and reconstructed by adopting a plurality of detectors, for example, a linear array detector consisting of 216 probes is used for acquiring photoacoustic signals to reconstruct the position of a biological tissue, if a classical back projection algorithm is used for calculation, each probe needs to be calculated once, each calculation has square and square operations, and the position of one point can be determined by performing 216 times in total.

Disclosure of Invention

The invention aims to provide a photoacoustic image back projection reconstruction method based on a single-address LUT table, which solves the problems of large calculation amount and large storage amount of a multi-LUT lookup table in the prior art.

The invention adopts the technical scheme that a photoacoustic image back projection reconstruction method based on a single-address LUT (look-up table) is implemented according to the following steps:

step 1, acquiring point number n on the side of a linear array ultrasonic detector according to the number m of probes of the linear array ultrasonic detector, namely the total number n of acquisition time points of a single probe;

step 2, determining the resolution of the reconstructed image as m x n according to the number m of the probes of the linear ultrasonic detector and the total number n of the acquisition time points of a single probe;

step 3, pre-constructing an LUT lookup table of the first probe according to the adopted frequency of the first probe, the total effective signal time, the propagation speed of the ultrasound in the biological tissue and the size or resolution of a reconstructed image and a back projection algorithm; the size of the lookup table is consistent with the resolution of the reconstructed image and is m x n, the lookup table information is a data table which mutually corresponds the coordinates of the reconstructed image and the coordinates of the ultrasonic signals, the coordinates of the ultrasonic signals corresponding to the coordinates of the reconstructed image can be directly found by utilizing the information of the lookup table, and the corresponding amplitude values are superposed to the coordinates of the reconstructed image;

step 4, reconstructing LUT lookup tables of other remaining probes according to the distance between the two probes in the linear array ultrasonic detector, and reconstructing a total LUT lookup table according to the rules among all the LUT lookup tables;

step 5, when the photoacoustic image is reconstructed, according to the data on the probe to be reconstructed, the LUT lookup table of the first probe obtained in the step 3 and the LUT lookup table of the other probes obtained in the step 4 are used for deducing the lookup table corresponding to the probe;

and 6, finding the coordinates of the pixel points of the corresponding reconstructed image according to the lookup table corresponding to each probe, and then accumulating the amplitude of the ultrasonic signal of the point on the corresponding probe to the pixel value of the point of the reconstructed image to reconstruct the required photoacoustic image.

The present invention is also characterized in that,

in the step 1, the number n of the acquisition points on the side of the linear array ultrasonic detector is calculated according to the following formula:

n＝t/T

wherein, T is a sampling period of the linear array ultrasonic detector, T is 1/f, f is a sampling frequency of a signal on the linear array ultrasonic detector, and T is a total effective signal time of the probe.

The step 3 specifically comprises the following steps:

step 3.1, establishing an m multiplied by n 0 matrix, wherein the matrix is named as A, m is the number of probes, and n is the total number of the collection time points of a single probe;

step 3.2, firstly defining a variable y as a probe serial number, wherein the value range is 1-m, x is the number of probe acquisition time points, the value range is 1-n, a rectangular coordinate system is established by taking the position of the detector No. 1 as an original point, the minimum unit of abscissa is the acquisition depth h (1) of the probe in one period, the minimum unit of ordinate is the adjacent distance of two probes, the position point of the probe No. 1 is taken as the center of a circle, the depth h (x) is xvT and is taken as the radius to draw n circular arcs, the circular arcs intersect with

y

1 and 2 … m to generate an intersection point, and v is the propagation speed of the ultrasound in the biological tissue;

and 3.4, taking x in the radius for drawing the circular arc as a first bit address of the matrix, taking the ordinate of the intersection point as a second bit address of the matrix, and taking the abscissa of the intersection point as an amplitude value to write in the matrix to obtain an LUT lookup table of the No. 1 sensor.

And 3.2, in the process of drawing the arc, when the coordinates of the intersection points are decimal, performing approximate rounding operation on the intersection points, so that the coordinates of the intersection points are all integral multiples of 1.

h(1)＝vT。

The step 4 specifically comprises the following steps:

the time step length detected by each probe is the same as the detection period, so that the distance between every two detection points is the same, namely the radius of each point drawing arc is the same, the coordinates of the intersection points corresponding to the arc drawing is carried out on every point of the No. 1 probe, the coordinates of the corresponding intersection points are obtained by drawing the arc on every point of the No. 2 probe in the same process, the coordinates of the No. 2 probe are downwards translated by one grid point compared with the No. 1 probe, the abscissa of the intersection point of the first row of the arc is the same as the abscissa of the intersection point of the third row, the LUT lookup table is the coordinate information of the intersection point of the arc, and the LUT lookup table of the No. 2 probe is the first row which integrally translates the No. 1 probe downwards by one row and integrally copies the line 2 in the LUT lookup table of the No. 1 probe to the first row of the LUT lookup table of the No. 2 probe; in analogy, the LUT lookup table for the probe No. 3 is that the LUT lookup table for the probe No. 1 is translated downwards by two lines, the two

lines

2 and 3 of the LUT lookup table for the probe No. 1 are taken as the two

lines

2 and 1 of the LUT lookup table for the probe No. 3, a total LUT lookup table is reconstructed according to the rule, the total line is 2m-1, the line m is the first line of the LUT lookup table for the probe No. 1, the line m-1 and the line m +1 are both the second line of the LUT lookup table for the probe No. 1, the line m-2 and the line m +2 are both the third line of the LUT lookup table for the probe No. 1, and in analogy, the LUT lookup tables for all the rest probes can be deduced according to the LUT lookup table for the probe No. 1.

The step 4 specifically comprises the following steps:

the linear array ultrasonic detector is set to meet the following two requirements:

the distance between every two probes is a fixed value;

acquiring frequency of each probe is the same as total time of effective signals;

deducing the LUT lookup tables of the other m-1 probes by using the two constraint conditions, and reconstructing a total LUT lookup table according to rules among all LUT lookup tables, wherein the deduction step specifically comprises the following steps:

step 4.1, the data of the Y line of the LUT lookup table of the probe number Y is consistent with the data of the first line of the probe number 1, and the first line of the LUT lookup table of the probe number 1 is filled;

step 4.2, when the row number distance between the rest rows in the LUT lookup table of the probe number Y and the Y row distance between the rest rows in the LUT lookup table of the probe number 1 and the row number distance between the rest rows in the LUT lookup table of the probe number 1 are consistent, the data in the rows are also consistent, the data with the consistent row number distances are filled in, and the LUT lookup table of the probe number Y can be reconstructed;

step 4.3, reconstructing an LUT lookup table of the probe number m from the LUT lookup table of the probe number two according to the step 4.2 and the step 4.3 to obtain LUT lookup tables of all the probes;

and 4.4, reconstructing a total LUT lookup table according to rules among all LUT lookup tables, wherein the total LUT lookup table is obtained by symmetrically removing all data of a first row from the LUT lookup table of the probe No. 1 at the position of the first row, reconstructing a total LUT lookup table of 2m-1 rows, namely, data of the first row of the LUT lookup table of the probe No. 1 in the mth row in the total LUT lookup table, data of the second row of the LUT lookup table of the probe No. 1 in the mth-1 row and data of the second row of the LUT lookup table of the probe No. 1 in the mth row in the LUT lookup table of the probe No. 1 in the mth row, and repeating the steps until the data of the mth row in the LUT lookup table of the probe No. 1 in the 1 row and the mth-1 in the 2m-1 row are reconstructed to obtain the total LUT lookup table.

The invention has the beneficial effects that:

the invention considers that when the sampling frequency of the ultrasonic detector is a fixed value, the time of each sampling point on the ultrasonic detector is determined, the distance from the point to the biological tissue which sends the ultrasonic signal is determined, and the point which the value is projected to the photoacoustic image with determined size is also determined, namely, the arc radius is also determined by back projecting the points on the same column. The coordinate value of the photoacoustic reconstruction image corresponding to each point on each detector is found and is made into an LUT lookup table, and then on photoacoustic image reconstruction, only the point of data on the detector is counted to obtain the second point, the preprocessed LUT table can be searched to quickly reconstruct the image without traversing and calculating the distance between the point on the whole image and the detector, so that the process of solving the square sum of the distances is simplified into the process of table lookup, and the calculation amount of the whole calculation process is greatly reduced;

meanwhile, the LUT lookup tables made by each detector are analyzed, and the distance between the detector probes is determined for the linear array detectors with determined models, so that a certain rule exists between each LUT lookup table, the positions of pixel points on other detectors can be quickly calculated by one LUT lookup table without storing the LUT tables on all the detectors, and the capacity requirement of a memory is greatly reduced.

The invention greatly simplifies the whole hardware system design of the classic back projection algorithm and can reconstruct the needed photoacoustic image by using less hardware resources.

Drawings

FIG. 1 is a diagram of a prior art back-projection photoacoustic imaging process as set forth in the background of the invention;

FIG. 2 is a schematic diagram of back projection of each point on a probe in the photoacoustic image back projection reconstruction method based on the single address LUT table according to the present invention;

FIG. 3 is a schematic diagram of the LUT table derivation of probe detection points No. 1 to No. 2 in the photoacoustic image back projection reconstruction method based on the single address LUT table according to the present invention;

FIG. 4 is a diagram of an initial biological tissue model in the photoacoustic image back-projection reconstruction method based on a single address LUT table according to the present invention;

FIG. 5 is a diagram of the propagation process of the simulated photoacoustic wave in the biological tissue in the photoacoustic image back-projection reconstruction method based on the single address LUT table according to the present invention;

FIG. 6 is a partial waveform diagram of photoacoustic waves collected by probe No. 1 in the photoacoustic image back-projection reconstruction method based on the single address LUT table according to the present invention;

FIG. 7 is a diagram of photoacoustic wave simulation data collected from 216 probes in an embodiment of the photoacoustic image back-projection reconstruction method based on the single-address LUT table according to the present invention;

FIG. 8 is LUT table partial data of probe No. 1 in an embodiment of the photoacoustic image back-projection reconstruction method based on single address LUT table according to the present invention;

fig. 9 is partial data of a total LUT table fabricated in an embodiment of the photoacoustic image back-projection reconstruction method based on a single address LUT table according to the present invention;

fig. 10 is an image reconstructed by LabVIEW in the embodiment of the photoacoustic image back projection reconstruction method based on the single address LUT table.

Detailed Description

The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

The invention relates to a photoacoustic image back projection reconstruction method based on a single address LUT table, which is implemented by the following steps:

step 1, according to the number m of probes of the linear array ultrasonic detector and the number n of acquisition points on the side of the linear array ultrasonic detector, namely the total number n of acquisition time points of a single probe, wherein the number n of acquisition points on the side of the linear array ultrasonic detector is calculated according to the following formula:

n＝t/T

wherein, T is a sampling period of the linear array ultrasonic detector, T is 1/f, f is the sampling frequency of the signal on the linear array ultrasonic detector, and T is the total effective signal time of the probe;

step 2, determining the resolution of the reconstructed image as m x n according to the number m of the probes of the linear ultrasonic detector and the total number n of the acquisition time points of a single probe, namely determining how many probes correspond to how many lines of the reconstructed image;

step 3, pre-constructing an LUT lookup table of the first probe according to the adopted frequency of the first probe, the total effective signal time, the propagation speed of the ultrasound in the biological tissue and the size or resolution of a reconstructed image and a back projection algorithm; the size of the lookup table is consistent with the resolution of the reconstructed image and is m x n, the lookup table information is a data table which mutually corresponds the coordinates of the reconstructed image and the coordinates of the ultrasonic signals, the coordinates of the ultrasonic signals corresponding to the coordinates of the reconstructed image can be directly found by utilizing the information of the lookup table, and the corresponding amplitude values are superposed to the coordinates of the reconstructed image; the method specifically comprises the following steps:

step 3.2, firstly defining a variable y as a probe serial number, wherein the value range is 1-m, x is the number of probe acquisition time points, the value range is 1-n, a rectangular coordinate system is established by taking the position of the probe No. 1 as an original point, the minimum unit of the abscissa is the acquisition depth h (1) of the probe in one period is vT, in order to establish an LUT lookup table, vT is 1, namely the acquisition depth of the probe in one period is 1, the minimum unit of the ordinate is the adjacent distance of two probes, the distance between each probe is 1, namely the total length of the linear array probe is m-1, the position point of the probe No. 1 is taken as a center of a circle, the depth h (x) xvT is taken as a radius to draw n circular arcs, the circular arcs intersect with y 1,2 … m to generate an intersection point, and in the process of drawing the arc, when the coordinate of the intersection point is a decimal, approximate rounding operation is adopted for the intersection point, making the coordinates of the intersection points be integral multiples of 1, wherein v is the propagation speed of the ultrasound in the biological tissue; for example: when x is 5, the arc is drawn by taking the position of the detector No. 1 as the center, 5vT as the radius, the intersection point with y is 2 is (5,2), and the intersection point with y is 4 is (4,4), as shown in FIG. 2.

Step 3.4, taking x in the radius for drawing the circular arc as a first bit address of the matrix, taking the ordinate of the intersection point as a second bit address of the matrix, and taking the abscissa of the intersection point as an amplitude value to write in the matrix to obtain an LUT lookup table of the sensor No. 1; for example: the amplitude of A (7,4) is 6, the detection depth of the No. 1 sensor at the 7 th time point is drawn by a radius, and the intersection point of the detection depth and a straight line y is 4, namely the intersection point of the detection depth and the straight line passing through the No. 4 sensing line is (6, 4).

Step 4, reconstructing LUT lookup tables of other remaining probes according to the distance between the two probes in the linear array ultrasonic detector, and reconstructing a total LUT lookup table according to the rules among all the LUT lookup tables; the method specifically comprises the following steps:

because the time step length detected by each probe is the same as the detection period, the distance between each detection point is also the same, namely the radius of each point drawing arc is the same, the coordinates of the intersection point corresponding to the drawing arc of each point of the No. 1 probe in the graph 3 are obtained by drawing the arc of each point of the No. 2 probe in the same process, as can be seen from the graph 3, the coordinates of the No. 2 probe are downwards translated by one grid point compared with the No. 1 probe, the abscissa of the intersection point of the first row of the arc is the same as the abscissa of the intersection point of the third row, the LUT lookup table is the coordinate information of the intersection point of the arc, the LUT lookup table of the No. 2 probe is the first row of the LUT lookup table of the No. 1 probe which is wholly copied to the LUT lookup table of the No. 2 probe; in analogy, the LUT lookup table for the probe No. 3 is that the LUT lookup table for the probe No. 1 is translated downwards by two lines, the two lines 2 and 3 of the LUT lookup table for the probe No. 1 are taken as the two lines 2 and 1 of the LUT lookup table for the probe No. 3, a total LUT lookup table is reconstructed according to the rule, the total line is 2m-1, the line m is the first line of the LUT lookup table for the probe No. 1, the line m-1 and the line m +1 are both the second line of the LUT lookup table for the probe No. 1, the line m-2 and the line m +2 are both the third line of the LUT lookup table for the probe No. 1, and in analogy, the LUT lookup tables for all the rest probes can be deduced according to the LUT lookup table for the probe No. 1; the step 4 specifically comprises the following steps:

the distance between every two probes is a fixed value;

fourthly, the acquisition frequency of each probe is the same as the total time of the effective signals;

step 4.4, reconstructing a total LUT lookup table according to rules among all LUT lookup tables, wherein the total LUT lookup table is obtained by symmetrically removing all data of a first row from the LUT lookup table of the probe No. 1 at the position of the first row, reconstructing a total LUT lookup table of 2m-1 rows, namely reconstructing data of a first row of the LUT lookup table of the probe No. 1 in the mth row in the LUT lookup table, reconstructing data of a second row of the LUT lookup table of the probe No. 1 in the mth row, and repeating the steps until the data of the mth row in the LUT lookup table of the probe No. 1 in the mth row is the data of the LUT lookup table of the probe No. 1 in the mth row, and reconstructing a total LUT lookup table;

In the photoacoustic back projection imaging process based on the single-address LUT lookup table, the element which can replace the technical characteristic in the invention is that the reconstruction size of the single-address LUT lookup table is determined according to the biological tissue depth, the system sampling frequency and other hardware which can be detected by a detector, and the column number of the reconstructed image is also determined. The resolution of the imaging is related to the number of probes of the linear array detector, and if the linear array detector is a linear array with m probes, the number of rows of the image is m.

When the invention carries out back projection reconstruction on the acquisition point on the probe on the No. 1 probe during image reconstruction, an LUT lookup table corresponding to the pixel point coordinates of the acquisition point and the photoacoustic image is designed, and the calculation amount of photoacoustic image reconstruction can be greatly saved in the process. According to the rule obtained by the invention point 1, the LUT lookup tables of all the other probes can be deduced according to the LUT lookup table of the probe No. 1, and the LUT lookup tables corresponding to all the probes are combined according to the summarizing rule to obtain a single LUT lookup table, so that the LUT lookup tables of all the other probes are replaced, and the storage space of the LUT lookup tables is greatly saved; based on the proposed single LUT lookup table, the LUT lookup table reconstruction and photoacoustic image reconstruction algorithm of each probe during photoacoustic image reconstruction are designed.

According to the implementation process of the classical back projection algorithm, the feasibility and the beneficial effects of the method are easily analyzed. In the classical back-projection photoacoustic imaging, the process of calculating which pixel point falls on a reconstructed image for each point on a detector actually needs to be calculated in three steps:

(1) according to the coordinate value of said point on the waveform collected by said detector the distance r from said point to ultrasonic source can be calculated^*I.e. there are 1 multiplication.

(2) And calculating the horizontal and vertical distances from each pixel point of the reconstructed image area to the detector, namely, 2 times of power calculation exists.

(3) Calculating the linear distance r from each pixel point of the reconstructed image area to the detector; i.e. there are 1 addition and an opening operation.

(4) Judging whether the linear distance r is equal to r^*I.e. there are 1 equal judgments.

Through the above 4 steps, the classical back projection algorithm can determine whether the point on the image is a data drop point on the detector, i.e. 1 multiplication, 2 power, 1 addition, 1 power and 1 judgment are required to complete the determination. If each point is calculated in this way, the reconstruction process needs to calculate each point on the detector and the pixel point on the reconstructed image in the above four steps, so as to obtain the distribution of one point on the reconstructed image. Assuming that each detector can collect 778 points by using 216 detector lattices, the reconstruction of 216 × 778 points needs to be performed, each point needs to be calculated in the 4 steps on the reconstructed image, and the calculation amount is conceivably known, which seriously affects the speed of the photoacoustic imaging process.

The method for utilizing the LUT lookup table provided by the invention carries out preprocessing operation on pixel points of each point on each detector, which are located on a reconstructed image, each path of probe makes the LUT lookup table, and because the sampling frequency of each path of probe is a fixed value, the radius of a point-drawing arc in the same column is also a fixed value, so that a certain rule exists between each lookup table, the positions of the pixel points on other detectors can be quickly calculated by one lookup table, and the corresponding probe voltage amplitude signals can be inquired by inputting corresponding addresses, so that the amplitudes of each coordinate point are accumulated, and finally, an initial double-light-pressure point image is reconstructed and optimized. The method greatly simplifies the reconstruction process of the back projection photoacoustic image based on the linear array ultrasonic detector, converts all calculation processes into a table look-up process in the imaging process through preprocessing before imaging, and greatly reduces the calculation amount of the whole system. Meanwhile, by simplifying the LUT lookup table, all probe LUT lookup table data are required to be stored, and the LUT lookup table data are simplified to only one LUT lookup table, so that the system storage requirement is greatly reduced.

To illustrate that the present invention can simplify the calculation and reduce the system memory capacity, the following compares the classical backprojection algorithm with the single LUT look-up table based backprojection algorithm. Assuming that each probe can acquire 778 points by using 216 detector lattices, the reconstruction of 216 × 778 points needs to be performed, and the reconstruction process of each point is compared as follows:

(1) for the classical algorithm, it needs to judge whether all points on the reconstructed image are consistent with the distance of the point, that is, 216 × 778 times of judgment (the workload of each judgment is 1 multiplication +2 power +1 addition +1 evolution +1 judgment) is completed.

(2) For the single LUT look-up table algorithm, the LUT look-up table needs to be accessed only once, and all the point coordinates with equal distance are read out.

Therefore, the data points on the whole detector are 216 × 778 data convolution projections, and for the classical algorithm, the reconstruction needs the calculated amount of (216 × 778) × (1 multiplication +2 power +1 addition +1 evolution +1 judgment); for the single LUT look-up table algorithm, only 216 × 778 look-up tables are needed.

The invention also provides a method for using a single LUT lookup table, which combines and simplifies the characteristics of the LUT lookup tables of all probes on the linear array and can be realized by using one LUT lookup table. If the number of array probes is 216, the storage capacity is reduced to 1/216.

In summary, through the above analysis, the method provided by the present invention not only can greatly simplify the calculation, but also can greatly reduce the system storage capacity.

In order to verify the correctness and the effectiveness of the algorithm, the inventor simulates the whole system by adopting a simulation model. The simulation system adopts a 216-path ultrasonic probe array, the distance between the two probes is 0.1mm, and the width of the linear array probe can be obtained by calculation to be 2.15 cm; the propagation speed of the ultrasonic wave in the biological tissue is 1500m/s, the acquisition frequency is 50mHZ, T is 20ns according to the reciprocal relation of the period and the frequency, and the acquisition depth h (1) of one period is 0.03 mm; the total effective signal time of the probe is 15.56us, the total number of the collection time points is 778, and the vertical depth of the biological tissue detected by the detector is 2.334 cm. Therefore, the resolution of the photoacoustic image was 216 × 778, and the size of the image of the imaged biological tissue was 2.334cm × 2.15 cm. Corresponding to the photoacoustic image resolution, the simulated biological tissue model was assumed to be a grid of 216 × 778, with two points of optical absorption anomalies at the (110,82), (174,224) coordinate locations, as shown in fig. 4. The simulation software adopts Matlab-based green photoacoustic imaging software K-wave to simulate the photoacoustic signal propagation process, the propagation process of the photoacoustic wave in the biological tissue is shown in figure 5, and the ultrasonic signals on 216-path detectors are obtained. Fig. 6 shows the ultrasonic waveforms acquired by probe No. 1, and fig. 7 shows simulation data corresponding to the ultrasonic waveforms acquired by 216 probes.

In Matlab software, probe No. 1 look-up table is first constructed, and a matrix 0 of 216 x 778, named a, is first created. Defining a variable y as a probe serial number, wherein the value range is 1-216, x is the number of probe acquisition time points, the value range is 1-778, a rectangular coordinate system is established by taking the position of the probe No. 1 as an original point, the minimum unit of the abscissa is the detection depth of one period of 0.03mm, the minimum unit of the ordinate is the adjacent distance of two probes of 0.1mm, the position point of the probe No. 1 is taken as a circle center, the depth h (x) is 0.03xmm, n circular arcs are drawn by taking the radius of 1,2 … 778, and an intersection point is generated when any one of the straight lines of the arc and the straight lines of the probe No. 1 and the straight lines of the probe No. 2 … 216 intersect. All information of the process of generating one intersection point is recorded into a matrix to be used as a lookup table of the probe No. 1, and the specific operation is as follows: the method is used for drawing x in the arc radius h (x) as a first bit address of a matrix, the ordinate y of an intersection point is used as a second bit address of the matrix, a numerical value obtained after rounding operation on the abscissa of the intersection point is written as a storage value of the matrix, and finally a No. 1 sensor LUT lookup table is obtained, and the result is shown in figure 8 after the design and calculation of Matlab software. For example: the amplitude of the point of the matrix A (3,7) is 2, the meaning is that an arc is drawn by taking the acquisition depth of the 7 th time point of the No. 1 sensor as a radius (the radius length is 0.21mm), an intersection point is generated with a horizontal straight line y passing through the No. three sensor as 3, and the coordinate of the intersection point is (2, 3); the meaning in the back projection algorithm is that the projection coordinate of the seventh acquisition time point of the No. 1 probe is (2,3), and the position is the position of the second acquisition time point of the No. 3 probe. Attention is paid to: no. 1 probe 1-6 bit collection time point drawing arcs do not generate intersection points with y being 3, at the moment, the data of 1-6 bits in the third row of the lookup table are filled with 0, namely when n collection time points draw arcs, intersection points are generated, the abscissa x of the intersection points is written in, and if no intersection points are generated, 0 data is written in.

Since the time step detected by each probe is the same as the detection period, the distance between each acquisition time point is also the same, that is, the arc radius is the same for the same acquisition time point of different detectors. Obtaining other lookup tables according to the rule, wherein the LUT lookup table of the No. 2 probe is the first line of the LUT lookup table of the No. 1 probe which is wholly translated downwards by one line and is wholly copied to the LUT lookup table of the No. 2 probe; by analogy, the look-up table of the probe No. 3 is that the look-up table No. 1 is translated downwards by two lines, the two

lines

2 and 3 of the look-up table No. 1 are taken as the two

lines

2 and 1 of the look-up table No. 3, and a total look-up table of the LUT is reconstructed according to the rule.

Reconstruct a total LUT look-up table as shown in fig. 9, a single LUT look-up table was created using the EXCEL software. When indexing from 216 rows and looping up with this table, each row is decremented by 1 and 216 x 779 bit elements are simultaneously taken as new look-up table data, for example: the LUT lookup table for probe No. 3 is indexed from 214 rows, and 216 × 779 bits of elements are taken as the lookup table for probe No. 3. Thus, 216 lookup tables are obtained.

The image reconstruction part corresponds the sensor data and the projection points by utilizing a lookup table, and adopts LabVIEW to realize the photoacoustic back projection imaging program based on the single-address LUT lookup table. Firstly, deriving a two-dimensional array corresponding to 216 lookup tables from a single LUT table, and then establishing a for loop, wherein a loop variable i is defined as a longitudinal coordinate y value of a projection point; then, a for loop is established, and a loop variable i' is defined as the abscissa x value of the projection point. The purpose of this loop is to, based on the coordinates (x, y) of the projection points: searching X values on corresponding projection points and outputting amplitude abscissa X values corresponding to the X values; finally, after obtaining the X value of the abscissa of the amplitude, in the previous chapter, the Y value of the ordinate of the amplitude point has been obtained at the stage of preprocessing the data, and at this time, the amplitude table of the probe is input, so that the amplitude of the amplitude point corresponding to the target projection point can be found, and 216 lookup tables need to be circulated in total, and each lookup table needs to circulate each point.

And finally, image reconstruction is to superpose the amplitude values corresponding to the same coordinate for imaging. The 216 × 778 matrix is initialized first, a shift register is added in the for loop, amplitude values are superimposed by looping each point in the matrix, and after all loops, the final reconstructed image can be obtained as shown in fig. 10.

Claims

1. The photoacoustic image back projection reconstruction method based on the single address LUT is characterized by comprising the following steps:

step 3, according to the adopted frequency of the first probe, the total time of effective signals, the propagation speed of ultrasound in biological tissues and the size or resolution of a reconstructed image, a LUT lookup table of the first probe is pre-constructed according to a back projection algorithm, the size of the lookup table is consistent with the resolution of the reconstructed image and is m x n, the lookup table is used as a data table for mutually corresponding the coordinates of the reconstructed image and the coordinates of the ultrasound signals, the coordinates of the ultrasound signals corresponding to the coordinates of the reconstructed image can be directly found by utilizing the information of the lookup table, and the corresponding amplitude values are superposed to the coordinates of the reconstructed image;

step 4, deducing the LUT lookup tables of the other m-1 probes from the LUT lookup table of the first probe, and reconstructing a total LUT lookup table according to rules among all the LUT lookup tables;

and 6, the lookup table is an information table of two relations of the coordinates of the pixel points of the reconstructed image and the ultrasonic signals, the coordinates of the corresponding pixel points of the reconstructed image are found according to the lookup table corresponding to each probe, and then the amplitude of the ultrasonic signals of the corresponding probe is added to the pixel value of the point of the reconstructed image, so that the required photoacoustic image can be reconstructed.

2. The photoacoustic image back-projection reconstruction method based on the single-address LUT table as claimed in claim 1, wherein the number of the acquisition points n on the side of the linear array ultrasound probe in step 1 is calculated according to the following formula:

n＝t/T

3. The photoacoustic image back-projection reconstruction method based on the single-address LUT table according to claim 2, wherein the step 3 is specifically as follows:

step 3.2, firstly defining a variable y as a probe serial number, wherein the value range is 1-m, x is the number of probe acquisition time points, the value range is 1-n, a rectangular coordinate system is established by taking the position of the detector No. 1 as an original point, the minimum unit of abscissa is the acquisition depth h (1) of the probe in one period, the minimum unit of ordinate is the adjacent distance of two probes, the position point of the probe No. 1 is taken as the center of a circle, the depth h (x) is xvT and is taken as the radius to draw n circular arcs, the circular arcs intersect with y 1 and 2 … m to generate an intersection point, and v is the propagation speed of the ultrasound in the biological tissue;

4. The photoacoustic image back-projection reconstruction method based on the single-address LUT table as claimed in claim 3, wherein in the step 3.2 of performing the arc drawing, when the coordinates of the intersection points are decimal, an approximate rounding operation is performed on the intersection points, so that the coordinates of the intersection points are all integer multiples of 1.

5. The single address LUT table based photoacoustic image backprojection reconstruction method of claim 4, wherein h (1) ═ vT.

6. The photoacoustic image back-projection reconstruction method based on the single-address LUT table according to claim 5, wherein the step 4 is specifically as follows:

the time step length detected by each probe is the same as the detection period, so that the distance between every two detection points is the same, namely the radius of each point drawing arc is the same, the coordinates of the intersection points corresponding to the arc drawing is carried out on every point of the No. 1 probe, the coordinates of the corresponding intersection points are obtained by drawing the arc on every point of the No. 2 probe in the same process, the coordinates of the No. 2 probe are downwards translated by one grid point compared with the No. 1 probe, the abscissa of the intersection point of the first row of the arc is the same as the abscissa of the intersection point of the third row, the LUT lookup table is the coordinate information of the intersection point of the arc, and the LUT lookup table of the No. 2 probe is the first row which integrally translates the No. 1 probe downwards by one row and integrally copies the line 2 in the LUT lookup table of the No. 1 probe to the first row of the LUT lookup table of the No. 2 probe; in the same way, the LUT lookup table for the No. 3 probe is that the LUT lookup table for the No. 1 probe is translated downwards by two lines, the two lines 2 and 3 of the LUT lookup table for the No. 1 probe are taken as the two lines 2 and 1 of the LUT lookup table for the No. 3 probe, a total LUT lookup table is reconstructed according to the rule, the total line is 2m-1, the line m-1 and the line m +1 are both the second line of the LUT lookup table for the No. 1 probe, the line m-2 and the line m +2 are both the third line of the LUT lookup table for the No. 1 probe, and by analogy, the total LUT lookup table can be deduced according to the LUT lookup table for the No. 1 probe, namely the total lookup table is a summarized result of all lookup tables, and the lookup table for any probe can be obtained through the total lookup table.

7. The photoacoustic image back-projection reconstruction method based on the single-address LUT table according to claim 6, wherein the step 4 is specifically as follows:

the distance between every two probes is a fixed value;