Breast cancer microwave imaging method based on target response arrival time correction

Abstract

The invention discloses a breast cancer microwave imaging method based on target response arrival time correction, and the method specifically comprises the following steps: 1, carrying out the edge extraction of a prior breast structure knowledge image, and obtaining breast tissue edge information; step 2, carrying out secondary large connected domain identification on the extracted breast tissue edge to obtain an internal high dielectric tissue area; 3, calculating the average dielectric constant of the internal high-dielectric and low-dielectric areas; 4, calculating the lengths of a high dielectric region and a low dielectric region in a path between the imaging point and the antenna unit; step 5, calculating target response arrival time of each imaging point in the imaging area; and step 6, performing delay summation imaging on the whole mammary gland area, and calculating the image intensity of the whole mammary gland area so as to accurately position the abnormal position in the breast. According to the invention, clutters caused by gland tissues in a mammary gland microwave imaging method can be effectively inhibited, and tumor focusing imaging is realized.

Description

Technical Field

[0001] The invention belongs to the technical field of microwave medical imaging and relates to a breast cancer microwave imaging method based on target response arrival time correction. Background Art

[0002] Ultra-wideband microwave technology is a very promising means of early screening for breast cancer because of its outstanding advantages such as low electromagnetic radiation, high information load, and green safety. Confocal imaging is the most commonly used microwave imaging method for breast tumors. It usually estimates the wave velocity based on the assumption that the inside of the breast is a homogeneous medium, realizes the coherent superposition of abnormal signals inside the breast, and thus performs focused imaging of the breast tumor. How to achieve good tumor focused microwave images has attracted great attention from researchers at home and abroad, and has become a research hotspot and difficulty in the field of microwave breast tumor detection. In recent years, researchers have proposed a variety of methods for estimating equivalent dielectric properties inside the breast, such as dielectric property estimation methods based on image metrics (signal-to-noise ratio, focus quality index, etc.), thereby estimating an average wave velocity inside the breast to achieve confocal imaging. But in fact, the breast is a complex and heterogeneous tissue. Electromagnetic waves do not propagate at a uniform speed inside the breast at the wave speed corresponding to the equivalent dielectric properties. Traditional microwave confocal imaging based on a fixed equivalent dielectric constant will inevitably produce a large amount of clutter. When there are many glands, the positioning error of breast tumors is often large, and the image has a lot of irrelevant clutter, which will lead to tumor identification errors. Therefore, it is very important to study the accurate calculation of the target response arrival time correction method for breast microwave imaging. Summary of the invention

[0003] The purpose of the present invention is to provide a breast cancer microwave imaging method based on target response arrival time correction, which can effectively suppress the clutter caused by glandular tissue in the breast microwave imaging method and realize tumor focused imaging.

[0004] The technical solution adopted by the present invention is a breast cancer microwave imaging method based on target response arrival time correction, which specifically includes the following steps:

[0005] Step 1, extracting the edge of the prior breast structure knowledge image to obtain breast tissue edge information;

[0006] Step 2, identifying the second largest connected domain on the edge of the extracted breast tissue to obtain the internal high dielectric tissue area;

[0007] Step 3, based on the internal high dielectric tissue area and combined with the inverse scattered dielectric intensity image, the average dielectric constant of the internal high dielectric and low dielectric areas is calculated;

[0008] Step 4, calculating the lengths of the high dielectric region and the low dielectric region in the path between the imaging point and the antenna unit;

[0009] Step 5: Calculate the target response arrival time of each imaging point inside the imaging area;

[0010] Step 6: Perform delay and sum imaging on the entire breast area, calculate the image intensity of the entire breast area, and thus accurately locate the abnormal positions inside the breast.

[0011] The features of the present invention also lie in:

[0012] The specific process of Step 1 is as follows:

[0013] Step 1.1: Use the canny operator to perform preliminary edge recognition to obtain the breast tissue edge information;

[0014] Step 1.2: Set a rectangular 3×3 structural element, and perform dilation operation on the preliminary breast tissue edge obtained in Step 1.1 to ensure that the broken points are connected to form an overall edge.

[0015] The specific process of Step 2 is as follows:

[0016] Step 2.1: Extract the connected regions of the overall edge and retain the connected region with the second largest area. The connected region with the second largest area is the internal glandular tissue with a high dielectric constant, also known as the high dielectric region;

[0017] Step 2.2: Extract the set Fg of the edge coordinates of the high dielectric region.

[0018] The specific process of Step 3 is as follows:

[0019] Step 3.1: Perform internal filling on the high dielectric edge obtained in Step 2.2 to obtain the high dielectric region mask mask_h;

[0020] Step 3.2: Multiply the high dielectric region mask mask_h by the inverse scattering image I_b to calculate the average dielectric constant ε of the high dielectric region h , and the calculation formula is as follows:

[0021]

[0022] where m and n are the number of pixels in the width and height directions of the image, and mask_h i1,j1 , I_b i1,j1 are the values of the image mask_h and I_b at the position (i1, j1) respectively;

[0023] Step 3.3: Set the pixels with a value of 1 in the high dielectric region mask mask_h in the I_b image to 0 to obtain the dielectric distribution mask_l of the low dielectric region, and calculate the average dielectric constant ε of the low dielectric region l , and the calculation formula is as follows:

[0024]

[0025] Among them, mask_l i1,j1 is the value of the image mask_l at the position (i1, j1).

[0026] The specific process of step 4 is as follows:

[0027] Step 4.1, set up the microwave antenna test model for the breast. All antenna units A(x a , y a ) perform signal transmission and reception in a cyclic single - transmit and multi - receive mode. Denote the transmitting antenna as E(x E , y E ), the receiving antenna as D(x D , y D ), and the imaging point in the model as B(x b , y b ). Calculate the straight - line equation between the antenna unit A(x a , y a ) on the breast surface and the imaging point B(x b , y b ). The formula is as follows:

[0028] y = k AB ·x + c AB (4)

[0029] Among them, is the slope of the straight - line equation between the antenna unit and the imaging point, and c AB = y a - k AB ·x a is the intercept of the straight - line equation between the antenna unit and the imaging point. x and y are the abscissa and ordinate respectively;

[0030] Step 4.2, according to the coordinates of the transmitting antenna E(x E , y E ) and the imaging point B(x b , y b ), set the x value range as [x E , x b . Calculate the corresponding y coordinates according to formula (4) to obtain the set of line - segment coordinates E_B_cor between the position of the transmitting - antenna unit and the imaging point;

[0031] Step 4.3, according to the coordinates of the receiving antenna D(x D , y D ) and the imaging point B(x b , y b ), set the x range as [x D , x b, calculate the corresponding y coordinate according to formula (4), so as to obtain the set of line segment coordinates D_B_cor between the receiving antenna unit position and the imaging point;

[0032] Step 4.4, calculate the intersection point coordinate sets cross_E and cross_D between the edge coordinate set Fg in Step 2.2 and the line segments E_B_cor and D_B_cor;

[0033] The point Fg(i min1 ) with the minimum distance from Fg to the line segment E_B_cor is the intersection point cross_E, and the index coordinate i of the intersection point coordinate in the Fg set minE is calculated by the following formula:

[0034] i minE = min{iE|(iE, jE) ∈ argminD iE,jE}} (5)

[0035]

[0036] where D iE,jE represents the distance between the iE-th point Fg iE in the Fg edge and the jE-th point EB jE in the E_B_cor, then Fg(i minE ) is the cross_E set;

[0037] The point Fg(i minD ) with the minimum distance from Fg to the line segment D_B_cor is the intersection point cross_D, and the index coordinate i of the intersection point coordinate in the Fg set minD is calculated by the following formula:

[0038] i minD = min{iD|(iD, jD) ∈ argminD iD,jD}} (7)

[0039]

[0040] where D iD,jD represents the distance between the iD-th point Fg iD in the Fg edge and the jD-th point denoted as DB jD in the D_B_cor, then Fg(i minD ) is the cross_D set;

[0041] Step 4.5, calculate the path lengths of the high dielectric region and the low dielectric region in the path between the transmitting antenna and the imaging point B(x b , y b );

[0042] Step 4.6, calculate the path lengths of the high-dielectric regions and the low-dielectric regions in the path between the receiving antenna and the imaging point B(x b , y b ).

[0043] The specific process of Step 4.5 is as follows:

[0044] Step 4.5.1, determine the parity of the number of coordinates N cross_E in the set of intersection coordinates cross_E of the emission path and the high-dielectric region. According to the ray method principle, if the number of intersections is odd, the imaging point is inside the high-dielectric region curve; if it is even, the imaging point is outside.

[0045] Step 4.5.2, when the number of intersections N cross_E is odd, calculate the distance between the imaging point B(x b , y b ) and the intersection vector cross_E:

[0046]

[0047] where, represents the i2-th group of data in the intersection set cross_E. The intersection with the smallest distance in D 2,i2 is denoted as N0(x N0 , y N0 ), and the remaining intersections are arranged in descending order of distance as N1(x N1 , y N1 ), N2(x N2 , y N2 ), …, Nn(x Nn , y Nn ). Calculate the distances according to Eqs. (10), (11), and (12)

[0048]

[0049] ……

[0050]

[0051] Step 4.5.3, when the number of intersections N cross_E is even, calculate the distances between the imaging point B(x b , y b ) and each intersection. All intersections are arranged in descending order of distance and denoted as N1(x N1 , y N1 ), N2(x N2 , y N2 ), …, Nn(x Nn , y Nn), calculate the distance according to formulas (11) and (12).

[0052]

[0053] Step 4.5.4, calculate the sum of the distances between all pairs of intersection points, which is the path length between the imaging point B and the high-dielectric region in the transmitting antenna path. The calculation formula is as follows:

[0054]

[0055] Step 4.5.5, calculate the distance d between the imaging point B(x b , y b ) and the transmitting antenna E(x E , y E ), and the calculation formula is as follows: B,E The calculation formula is as follows:

[0056]

[0057] Step 4.5.6, subtract the result obtained in Step 4.5.4 from the d obtained in Step 4.5.5 B,E to obtain the path length of the low-dielectric region. The calculation formula is as follows:

[0058]

[0059] The specific process of Step 4.6 is as follows:

[0060] Step 4.6.1, determine the parity of the number of coordinates N of the set cross_D of the intersection points of the receiving path and the high-dielectric region. According to the ray method principle, if the number of intersection points is odd, the imaging point is inside the high-dielectric region curve; if it is even, the imaging point is outside. cross_D When the number of intersection points is odd, calculate the distance between the imaging point B(x

[0061] Step 4.6.2, when the number of intersection points N cross_D is odd, calculate the distance between the imaging point B(x b , y b ) and the intersection point vector cross_D:

[0062]

[0063] Among them, represents the i3-th group of data in the intersection point set cross_D, and the intersection point with the smallest distance in D 2,i3 is denoted as N0'(x N0 ', y N0 '), and the remaining intersection points are arranged in descending order of distance and denoted as N1'(x N1 ', yN1 '), N2'(x N2 ', y N2 '), …, Nn'(x Nn ', y Nn '), calculate the distance according to formulas (17), (18), and (19).

[0064]

[0065] ……

[0066]

[0067] Step 4.6.3, the number of intersection points N cross_D When it is an even number, calculate the distance between the imaging point B(x b , y b ) and each intersection point according to formula (16). Arrange all the intersection points in descending order of distance as N1'(x N1 ', y N1 '), N2'(x N2 ', y N2 '), …, Nn'(x Nn ', y Nn '), calculate the distance according to formulas (18) and (19).

[0068] Step 4.6.4, calculate the sum of the distances between all intersection point pairs, which is the path length between the imaging point B and the high-dielectric region in the receiving antenna path. The calculation formula is as follows:

[0069]

[0070] Step 4.6.5, calculate the distance d b , y b ) of the imaging point B(x D , y D ) and the receiving antenna D(x B,D , and the calculation formula is as follows:

[0071]

[0072] Step 4.6.6, subtract the obtained in Step 4.6.4 and the d B,D obtained in Step 4.6.5 to get the path length of the low-dielectric region. The calculation formula is as follows:

[0073]

[0074] The specific process of Step 5 is as follows:

[0075] Step 5.1: Calculate the dielectric constants ε h and ε l of the high-dielectric region and the low-dielectric region obtained from Step 3, and calculate the wave velocities v h and v l of the high and low dielectric regions. The formula is as follows:

[0076]

[0077] where c is the speed of light;

[0078] Step 5.2: According to the path lengths and of the high and low dielectric regions of the transmitting antenna and the receiving antenna obtained from Steps 4.5 and 4.6, and the wave velocities v h and v l of the high and low dielectric regions obtained from Step 5.1, calculate the arrival time t all of the target response. The formula is as follows:

[0079] tall = th + tl (25)

[0080]

[0081] In Step 6, calculate the image intensity I of the entire breast region through the following formula (28):

[0082]

[0083] where S ij is the signal amplitude of the i-th transmitting antenna to the j-th receiving antenna channel, I and J are the numbers of transmitting and receiving antenna channels respectively, and T c is the system calibration time.

[0084] The beneficial effect of the present invention is that the present invention determines the correction relationship of electromagnetic propagation in the breast through the prior knowledge of breast tissue structure, extracts the edges of the internal high-dielectric regions of the prior image, estimates the average dielectric constants of the high-dielectric region and the low-dielectric region, accurately calculates the electromagnetic wave propagation path between each imaging point and each pair of antenna transceiver pairs, and thus calculates the arrival time of the target response. By identifying different dielectric regions of the internal tissue, accurately estimating the propagation time of electromagnetic waves in the breast under different paths, and finally realizing confocal imaging of tumor tissues, the present invention has more accurate tumor focusing and positioning results and higher image quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 is a flowchart of the microwave imaging method for breast cancer based on the correction of the arrival time of the target response according to the present invention;

[0086] Figure 2 It is a schematic diagram of a breast model and related symbol markings;

[0087] Figure 3 This is a comparison diagram of the microwave imaging effect of breast cancer based on the correction of the target response arrival time in the present invention. Specific implementation manners

[0088] The present invention will be described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0089] Example 1

[0090] The microwave imaging method for breast cancer based on the correction of the target response arrival time in the present invention has the overall process as Figure 1 shown.

[0091] Step 1: Based on the prior breast structure knowledge image I_m, perform edge extraction to obtain the breast tissue edge information;

[0092] Step 2: Identify the second largest connected region by area for the extracted breast tissue edge to obtain the internal high dielectric (relative dielectric constant 20 - 50) (fibroglandular) tissue region;

[0093] Step 3: According to the separated internal high dielectric (fibroglandular) region, combined with the inverse scattering dielectric strength image, calculate the average dielectric constants of the internal high dielectric and low dielectric (relative dielectric constant 5 - 20) (fat) regions;

[0094] Step 4: Calculate the lengths of the high dielectric region and the low dielectric region in the path between the imaging point and the antenna unit;

[0095] Step 5: Calculate the specific wave velocities of the high and low dielectric regions from the average dielectric constants of the high and low dielectric regions, and then, based on the path lengths of the high and low dielectric regions in the total path from the imaging point to the transmitting and receiving antennas, calculate the target response arrival time t of each imaging point inside the imaging region according to the wave velocity equation all ;

[0096] Step 6: Perform delay summation imaging on the entire breast region, calculate the image intensity I of the entire breast region, so as to accurately locate the abnormal position inside the breast. The calculation formula is as follows:

[0097]

[0098] where S ij is the signal amplitude of the i-th transmitting antenna to the j-th receiving antenna channel, I and J are the numbers of the transmitting and receiving antenna channels respectively, and T c is the system calibration time.

[0099] Example 2

[0100] The specific steps for extracting the edge information of the prior image (denoted as I_m, with image size [m, n]) in Step 1 may include:

[0101] Step 1.1, perform preliminary edge recognition using a canny operator with a threshold of 0.15 to obtain breast tissue edge information;

[0102] Step 1.2, set a rectangular 3×3 structural element, and perform a dilation operation on the preliminary edge to ensure that the broken points are connected to form an overall edge.

[0103] Example 3

[0104] For Step 2, the dielectric constant of breast fibroglandular tissue is relatively high, corresponding to the high-dielectric region in the internal central area, and the dielectric parameter of adipose tissue is relatively low, corresponding to the low-dielectric region surrounding the high-dielectric region. Therefore, the high-dielectric tissue region is separated, and the remaining region is the low-dielectric region. Step 2 may include the following steps:

[0105] Step 2.1, perform connected component extraction on the overall edge and retain the connected region with the second largest area. Since the largest connected component extracted is the breast skin edge and the second largest is the high-dielectric edge of the internal glandular tissue, the connected region with the second largest area is called the high-dielectric region;

[0106] Step 2.2, extract the set of edge coordinates Fg of the high-dielectric region;

[0107] Example 4

[0108] Based on the above Example 3, in Step 3, the average dielectric constants of the internal high-dielectric and low-dielectric regions are calculated by combining the inverse scattering image (denoted as I_b) of the verified breast structure knowledge image I_m. Step 3 may specifically include:

[0109] Step 3.1, perform internal filling on the high-dielectric edge obtained in Step 2.2 to obtain a high-dielectric region mask mask_h (pixels in the high-dielectric region are 1, and the remaining pixels are 0);

[0110] Step 3.2, multiply the high-dielectric region mask mask_h by the inverse scattering image I_b to calculate the average dielectric constant ε of the high-dielectric region h , and the calculation formula is as follows:

[0111]

[0112] where m and n are the number of pixels in the width and height directions of the image. mask_h i1,j1 ,I_b i1,j1 are the values of the image mask_h and I_b at the position (i1, j1) respectively.

[0113] Step 3.3, set the pixels with value 1 in the high-dielectric region mask mask_h in the I_b image to 0 to obtain the dielectric distribution mask_l of the low-dielectric region, and calculate the average dielectric constant ε of the low-dielectric region. The calculation formula is as follows: l , and the calculation formula is as follows:

[0114]

[0115] where mask_l i1,j1 is the value of the image mask_l at the position (i1, j1).

[0116] Example 5

[0117] Based on the above Example 4, calculate the lengths passing through the high-dielectric region and the low-dielectric (fat) region between the imaging point and the antenna unit path, so as to calculate the corrected target response arrival time estimation later. Step 4 includes:

[0118] Step 4.1, the microwave antenna test model of the breast is set up and relevant symbols are marked as Figure 2 shown. The black cross mark is the antenna unit. All antenna units (collectively referred to as A(x a , y a )) perform signal transmission and reception in a cyclic single-transmit multi-receive mode. During this process, the transmitting antenna is denoted as E(x E , y E ), the receiving antenna is denoted as D(x D , y D ), and the imaging point in the model is denoted as B(x b , y b ). The largest circle (pale yellow mark) is the breast model, and the internal blue circle is the fibrous gland inside the breast. Calculate the straight-line equation between the antenna unit A(x a , y a ) on the breast surface and the imaging point B(x b , y b ) to facilitate the subsequent calculation of the intersection point of the wave propagation process and the high-dielectric edge. The calculation formula is as follows:

[0119] y = k AB ·x + c AB (5)

[0120] where is the slope of the straight-line equation between the antenna unit and the imaging point, c AB = y a - k AB ·x a is the intercept of the straight-line equation between the antenna unit and the imaging point, and x and y are the horizontal and vertical coordinates respectively.

[0121] Step 4.2, according to the coordinates of the transmitting antenna E(x E ,y E ) and the imaging point B(x b ,y b ), set the x value range as [x E ,x b , calculate the corresponding y coordinates according to formula (5), so as to obtain the line segment coordinate set E_B_cor between the position of the transmitting antenna unit and the imaging point;

[0122] Step 4.3, according to the coordinates of the receiving antenna D(x D ,y D ) and the imaging point B(x b ,y b ), set the x range as [x D ,x b , calculate the corresponding y coordinates according to formula (5), so as to obtain the line segment coordinate set D_B_cor between the position of the receiving antenna unit and the imaging point;

[0123] Step 4.4, calculate the intersection point coordinate sets cross_E and cross_D of the edge coordinate set Fg in Step 2.2 with the line segments E_B_cor and D_B_cor. First, calculate the point Fg(i min1 ) with the minimum distance between Fg and the line segment E_B_cor, which is the intersection point cross_E. The index coordinate i minE of the intersection point coordinate in the Fg set can be calculated by the following formula:

[0124] i minE =min{iE|(iE,jE)∈argminD iE,jE} (6)

[0125]

[0126] Among them, D iE,jE represents the distance between the iE-th point (denoted as Fg iE ) in the Fg edge and the jE-th point (denoted as EB jE ) in the E_B_cor. Then Fg(i minE ) is the cross_E set. Then calculate the point Fg(i minD ) with the minimum distance between Fg and the line segment D_B_cor, which is the intersection point cross_D. The index coordinate i minD of the intersection point coordinate in the Fg set can be calculated by the following formula:

[0127] i minD =min{iD|(iD,jD)∈argminD iD,jD} (8)

[0128]

[0129] Among them, D iD,jD represents the distance between the iD-th point (denoted as Fg iD ) in the Fg edge and the jD-th point (denoted as DB jD ) in D_B_cor. Then, Fg(i minD ) is the cross_D set.

[0130] Based on the above embodiments, according to the intersection coordinates obtained in step 4.4, calculate the path lengths of the imaging point B(x b , y b ) and the paths passing through the high-dielectric region and the low-dielectric region in the transmitting and receiving antenna paths;

[0131] Step 4.5, first calculate the path lengths of the high-dielectric region and the low-dielectric region in the path between the transmitting antenna and the imaging point B(x b , y b ). Accordingly, step 4.5 may include:

[0132] Step 4.5.1, judge the parity of the number of coordinates N cross_E of the intersection coordinate set cross_E of the transmitting path and the high-dielectric region. According to the ray method principle, if the number of intersections is odd, the imaging point is inside the high-dielectric region curve; if it is even, the imaging point is outside.

[0133] Step 4.5.2, when the number of intersections N cross_E is odd, calculate the distance between the imaging point B(x b , y b ) and the intersection vector cross_E:

[0134]

[0135] Among them, represents the i2-th group of data of the intersection set cross_E. The intersection with the smallest distance in D 2,i2 is denoted as N0(x N0 , y N0 ), and the remaining intersections are arranged in descending order of distance as N1(x N1 , y N1 ), N2(x N2 , y N2 ), …, Nn(x Nn , y Nn ). Calculate the distances according to equations (11), (12), and (13)

[0136]

[0137] ……

[0138]

[0139] Step 4.5.3, the number of intersection points N cross_E When it is an even number, calculate the distances between the imaging point B(x b , y b ) and each intersection point according to formula (10). All the intersection points are arranged in descending order of distance and denoted as N1(x N1 , y N1 ), N2(x N2 , y N2 ), …, Nn(x Nn , y Nn ). Calculate the distances according to formulas (12) and (13)

[0140]

[0141] Step 4.5.4, calculate the sum of the distances between all pairs of intersection points, which is the path length between the imaging point B and the high-dielectric region in the transmitting antenna path The calculation formula is as follows:

[0142]

[0143] Step 4.5.5, calculate the distance d b , y b ) between the imaging point B(x E , y E ) and the transmitting antenna E(x B,E . The calculation formula is as follows:

[0144]

[0145] Step 4.5.6, use the obtained in Step 4.5.4 and the d B,E obtained in Step 4.5.5, and subtract them to get the path length of the low-dielectric region The calculation formula is as follows:

[0146]

[0147] Step 4.6, calculate the path lengths of the high-dielectric region and the low-dielectric region in the path between the receiving antenna and the imaging point B(x b , y b ). Correspondingly, Step 4.6 may include:

[0148] Step 4.6.1: Determine the number of coordinates N in the set cross_D of intersection points between the receiving path and the high-dielectric region. cross_D Parity: According to the ray method principle, if the number of intersection points is odd, the imaging point is inside the high-dielectric region curve; if it is even, the imaging point is outside.

[0149] Step 4.6.2: When the number of intersection points N cross_D is odd, calculate the distance between the imaging point B(x b , y b ) and the intersection vector cross_D:

[0150]

[0151] Among them, represents the i3 -th group of data in the intersection point set cross_D, and the intersection point with the smallest distance in D 2,i3 is denoted as N0'(x N0 ', y N0 '), and the remaining intersection points are arranged in descending order of distance and denoted as N1'(x N1 ', y N1 '), N2'(x N2 ', y N2 '), …, Nn'(x Nn ', y Nn '). Calculate the distance according to equations (18), (19), and (20)

[0152]

[0153] ……

[0154]

[0155] Step 4.6.3: When the number of intersection points N cross_D is even, calculate the distance between the imaging point B(x b , y b ) and each intersection point. All intersection points are arranged in descending order of distance N1'(x N1 ', y N1 '), N2'(x N2 ', y N2 '), …, Nn'(x Nn ', y Nn '). Calculate the distance according to equations (19) and (20)

[0156] Step 4.6.4: Calculate the sum of the distances between all intersection point pairs, which is the path length between the imaging point B and the high-dielectric region in the receiving antenna path. The calculation formula is as follows:

[0157]

[0158] Step 4.6.5, calculate the distance d b , y b ) between the imaging point B(x D , y D ) and the receiving antenna D(x B,D ), and the calculation formula is as follows:

[0159]

[0160] Step 4.6.6, subtract the obtained in Step 4.6.4 from the d B,D obtained in Step 4.6.5 to get the path length of the low dielectric region. The calculation formula is as follows:

[0161]

[0162] In Step 5, combine the wave velocity equation to calculate the arrival time t all of the target response at each imaging point inside the imaging region. The specific steps may include:

[0163] Step 5.1, calculate the wave velocities v h and v l of the high and low dielectric regions from the dielectric constants ε h and ε l of the high and low dielectric regions calculated in Step 3. The formulas are as follows:

[0164]

[0165] where c is the speed of light (3×10 8 m / s).

[0166] Step 5.2, calculate the arrival time t and of the target response according to the path lengths h and v l of the high and low dielectric regions of the transmitting and receiving antennas obtained in Steps 4.5 and 4.6 all and the wave velocities v

[0167] t all = t h + t l . (26)

[0168]

[0169] Example 6

[0170] An example of the result of the microwave breast cancer imaging method for correcting the target response arrival time provided by the present invention is as follows Figure 3 as shown Figure 3 In it, (a) is breast model I, and the tumor T is located in adipose tissue; (d) is breast model II, and the tumor T is located in glandular tissue; (g) is breast model III, a breast model with a complex glandular shape; (b) is the conventional confocal imaging result diagram of model I; (e) is the conventional confocal imaging result diagram of model II; (h) is the conventional confocal imaging result diagram of model III; (c) is the time-corrected confocal imaging result diagram proposed by the present invention for model I; (f) is the time-corrected confocal imaging result diagram proposed by the present invention for model II; (i) is the time-corrected confocal imaging result diagram proposed by the present invention for model III. The mark "T" in figures (a), (d), and (g) represents the yellow tumor area, and the red circles in figures (b), (c), (e), (f), (h), and (i) are the areas of the actual tumor positions. It can be seen that when the tumor is located outside the high dielectric region, as Figure 3 (a) shows, the conventional microwave confocal imaging method cannot effectively focus on the tumor area, and due to the occlusion of the high dielectric glandular tissue in the middle, its imaging effect becomes very poor, as Figure 3 (b) shows. However, the time-corrected imaging method proposed by the present invention effectively focuses on the abnormal tumor position and can accurately locate and identify the tumor, as Figure 3 (c) shows. On the other hand, when the tumor is located inside the high dielectric region, as Figure 3 (d) shows, multiple high bright spots appear in the conventional microwave confocal imaging method, and it is impossible to confirm which one is the abnormal tumor area, as Figure 3 (e) shows. However, the time-corrected imaging method proposed by the present invention effectively focuses on the abnormal tumor position and can accurately locate and identify the tumor, as Figure 3 (f) shows. The present invention also verifies the imaging result in the case of a more complex internal morphology Figure 3 (g). Compared with that obtained by the traditional method Figure 3 (h), Figure 3 (i) the imaging result is still excellent, with less clutter and accurate focusing of the tumor position.

[0171] Table 1 gives the tumor localization error and imaging signal-to-clutter ratio (SCR) of the conventional confocal imaging method and the method proposed by the present invention. The smaller the localization error, the more accurate the localization, and the larger the SCR, the higher the image quality. It can be seen from the table that the tumor localization error of the method proposed by the present invention is generally smaller than that of the conventional confocal imaging method, and the SCR is higher.

[0172] Generally speaking, the microwave breast imaging method based on response arrival time correction proposed by the present invention has more accurate tumor localization results and higher imaging quality.

[0173] Table 1. Comparison of breast tumor imaging performance

[0174]

Claims

1. A method for microwave imaging of breast cancer based on target response arrival time correction, characterized in that: The specific steps include: Step 1, extracting the edge of the prior breast structure knowledge image to obtain breast tissue edge information; Step 2, identifying the second largest connected domain on the edge of the extracted breast tissue to obtain the internal high dielectric tissue area; Step 3, based on the internal high dielectric tissue area and combined with the inverse scattered dielectric intensity image, the average dielectric constant of the internal high dielectric and low dielectric areas is calculated; Step 4, calculating the lengths of the high dielectric region and the low dielectric region in the path between the imaging point and the antenna unit; Step 5, calculating the target response arrival time of each imaging point inside the imaging area; Step 6: Perform time-delay summation imaging on the entire breast area and calculate the image intensity of the entire breast area, so as to accurately locate the abnormal position inside the breast.

2. The method for breast cancer microwave imaging based on target response arrival time correction according to claim 1, characterized in that: The specific process of step 1 is as follows: Step 1.1, using the Canny operator to perform preliminary edge recognition to obtain breast tissue edge information; Step 1.2, setting a rectangular 3×3 structure element, and performing a dilation operation on the preliminary edge of the breast tissue obtained in step 1.1 to ensure that the breakpoints are connected to form an overall edge.

3. The method for breast cancer microwave imaging based on target response arrival time correction according to claim 2, characterized in that: The specific process of step 2 is: Step 2.1, extracting the connected domain of the whole edge and retaining the connected region with the second largest area, the connected region with the second largest area is the high dielectric edge of the internal glandular tissue, also called the high dielectric region; Step 2.2, extract the edge coordinate set Fg of the high dielectric region.

4. The method for breast cancer microwave imaging based on target response arrival time correction according to claim 3, characterized in that: The specific process of step 3 is as follows: Step 3.1, filling the high dielectric edge obtained in step 2.2 to obtain a high dielectric region mask mask_h; Step 3.2, multiply the high dielectric region mask mask_h with the inverse scattering image I_b to calculate the average dielectric constant ε of the high dielectric region h , the calculation formula is as follows: Among them, m and n are the number of pixels in the width and height directions of the image, mask_h i1,j1 , I_b i1,j1 They are the values ​​of image mask_h and I_b at position (i1, j1) respectively; Step 3.3, set the pixels that are 1 in the high dielectric region mask mask_h in the I_b image to 0, obtain the dielectric distribution mask_l of the low dielectric region, and calculate the average dielectric constant ε of the low dielectric region l , the calculation formula is as follows: Among them, mask_l i1,j1 It is the value of the image mask_l at position (i1, j1).

5. The method for breast cancer microwave imaging based on target response arrival time correction according to claim 4, characterized in that: The specific process of step 4 is as follows: Step 4.1, set up the breast microwave antenna test model, all antenna units A(x a ,y a ) uses a cyclic single-transmit multiple-receive mode to transmit and receive signals, and the transmitting antenna is denoted by E(x E ,y E ), the receiving antenna is denoted as D(x D ,y D ), the imaging point in the model is denoted as B(x b ,y b ), calculate the breast surface antenna unit A(x a ,y a ) and the imaging point B(x b ,y b ) is as follows: y=k AB ·x+c AB (4) in, is the slope of the straight line equation between the antenna unit and the imaging point, c AB =y a -k AB ·x a is the intercept of the straight line equation between the antenna unit and the imaging point, and x and y are the horizontal and vertical coordinates respectively; Step 4.2, according to the transmitting antenna E(x E ,y E ) and imaging point B(x b ,y b ) coordinates, set the x value interval to [x E ,x b ], the corresponding y coordinate is calculated according to formula (4), thereby obtaining the line segment coordinate set E_B_cor between the transmitting antenna unit position and the imaging point; Step 4.3, according to the receiving antenna D(x D ,y D ) and imaging point B(x b ,y b ) coordinates, set the x interval to [x D ,x b ], the corresponding y coordinate is calculated according to formula (4), thereby obtaining the line segment coordinate set D_B_cor between the receiving antenna unit position and the imaging point; Step 4.4, calculate the edge coordinate set Fg in step 2.2 and the intersection coordinate sets cross_E and cross_D of the E_B_cor and D_B_cor line segments; The minimum distance between Fg and line segment E_B_cor is Fg(i min1 ) is the intersection point cross_E, and the index coordinates of the intersection point in the Fg set are i minE Calculated by the following formula: i minE =min{iE|(iE,jE)∈argminD iE,jE } (5) Among them, D iE,jE represents the iEth point Fg on the edge of Fg iE and the jEth point EB in E_B_cor jE The distance between them, then Fg(i minE ) is the cross_E set; The minimum distance between Fg and line segment D_B_cor is Fg(i minD ) is the intersection point cross_D, and the index coordinates of the intersection point in the Fg set are i minD Calculated by the following formula: i minD =min{iD|(iD,jD)∈argminD iD,jD } (7) Among them, D iD,jD represents the iDth point Fg on the edge of Fg iD The jDth point in D_B_cor is denoted as DB jD The distance between them, then Fg(i minD ) is the cross_D set; Step 4.5, calculate the distance between the transmitting antenna and the imaging point B(x b ,y b ) the path length of the high dielectric region and the low dielectric region in the path; Step 4.6, calculate the receiving antenna and imaging point B(x b ,y b )The path length of the high dielectric region and the low dielectric region in the path.

6. The method for breast cancer microwave imaging based on target response arrival time correction according to claim 5, characterized in that: The specific process of step 4.5 is as follows: Step 4.5.1, determine the number of coordinates N of the coordinate set cross_E of the intersection of the emission path and the high dielectric region cross_E Parity: According to the principle of the ray method, if the number of intersection points is odd, the imaging point is inside the high dielectric region curve; if it is even, the imaging point is outside; Step 4.5.2, when the number of intersection points N cross_E When it is an odd number, calculate the imaging point B(x b ,y b ) and the distance between the intersection vector cross_E: in, Represents the i2th group of data of the intersection set cross_E; D 2,i2 The intersection point with the smallest distance is denoted as N0(x N0 ,y N0 ), the remaining intersection points are arranged in descending order of distance N1(x N1 ,y N1 ), N2(x N2 ,y N2 ),…,Nn(x Nn ,y Nn ), calculate the distance according to equations (10), (11), and (12) Step 4.5.3, number of intersection points N cross_E When it is an even number, the imaging point B(x b ,y b ) and the distance between each intersection point, all intersection points are arranged in descending order of distance and recorded as N1(x N1 ,y N1 ), N2(x N2 ,y N2 ),…,Nn(x Nn ,y Nn ), calculate the distance according to equations (11) and (12): Step 4.5.4, calculate the sum of the distances of all intersection points to calculate the path length between imaging point B and the high dielectric region in the transmitting antenna path The calculation formula is as follows: Step 4.5.5, calculate the imaging point B(x b ,y b ) and the transmitting antenna E(x E ,y E ) B,E , the calculation formula is as follows: Step 4.5.6, use the value obtained in step 4.5.4 and d obtained in step 4.5.5 B,E Subtracting the two gives the path length of the low dielectric region. The calculation formula is as follows:

7. The method for breast cancer microwave imaging based on target response arrival time correction according to claim 6, characterized in that: The specific process of step 4.6 is as follows: Step 4.6.1, determine the number of coordinates N of the intersection coordinate set cross_D of the receiving path and the high dielectric region cross_D Parity: According to the principle of the ray method, if the number of intersection points is odd, the imaging point is inside the high dielectric region curve; if it is even, the imaging point is outside; Step 4.6.2, when the number of intersection points N cross_D When it is an odd number, calculate the imaging point B(x b ,y b ) and the distance between the intersection vector cross_D: in, Represents the i3th group of data of the intersection set cross_D, D 2,i3 The intersection point with the smallest distance is denoted as N0'(x N0 ',y N0 '), the remaining intersection points are arranged in descending order of distance and are recorded as N1'(x N1 ',y N1 '), N2'(x N2 ',y N2 '), ..., Nn'(x Nn ',y Nn '), calculate the distance according to equations (17), (18), and (19) Step 4.6.3, number of intersection points N cross_D When it is an even number, the imaging point B(x b ,y b ) and the distance between each intersection point, all intersection points are arranged in descending order of distance N1'(x N1 ',y N1 '), N2'(x N2 ',y N2 '), ..., Nn'(x Nn ',y Nn '), calculate the distance according to equations (18) and (19) Step 4.6.4, calculate the sum of the distances of all intersection points to calculate the path length of the high dielectric region in the path between the imaging point B and the receiving antenna. The calculation formula is as follows: Step 4.6.5, calculate the imaging point B(x b ,y b ) and the receiving antenna D(x D ,y D ) B,D , the calculation formula is as follows: Step 4.6.6, use the value obtained in step 4.6.4 and d obtained in step 4.6.5 B,D Subtracting the two gives the path length of the low dielectric region. The calculation formula is as follows:

8. The method for breast cancer microwave imaging based on target response arrival time correction according to claim 7, characterized in that: The specific process of step 5 is as follows: Step 5.1, the dielectric constant ε of the high dielectric region and the low dielectric region calculated in step 3 h and ε l , calculate the wave speed v in the high and low dielectric regions h and v l , the formula is as follows: Where c is the speed of light; Step 5.2: Path lengths of the high and low dielectric regions of the transmitting and receiving antennas obtained in Steps 4.5 and 4.6 and And the wave speed v in the high and low dielectric regions obtained in step 5.1 h and v l , calculate the target response arrival time t all , the formula is as follows: tall=th+tl (25) 9. The method for breast cancer microwave imaging based on target response arrival time correction according to claim 8, characterized in that: In step 6, the image intensity I of the entire breast area is calculated by the following formula (28): Among them, S ij is the signal amplitude from the i-th transmitting antenna to the j-th receiving antenna channel, I and J are the number of transmitting and receiving antenna channels respectively, T c Calibrate the system time.

Images