CN101779956A - Early tumor detection method - Google Patents

Abstract

The present invention provides an early tumor detection method. The method combines the tissue structure partition information of the region of interest obtained by the existing medical detection equipment, and refers to the electromagnetic characteristic parameter distribution database of various tissues of the human body to assign electromagnetic characteristic parameter values to different tissue structure partitions. scope. Then, by solving the electromagnetic inverse problem, the parameter distribution of the electromagnetic characteristics of the tissue in the region of interest is obtained, which can effectively detect early tumors and tissue lesions without obvious boundaries. The present invention combines tissue partition information and tissue electromagnetic characteristic parameter distribution range information when solving the electromagnetic inverse problem, thereby greatly reducing the amount of calculation. The electromagnetic inverse problem solving method combined with the existing medical detection equipment can effectively detect early tumors and tissue lesions without obvious boundaries.

Description

A method for early tumor detection

technical field

The invention relates to a medical detection method for detecting the existence of tumors, in particular to a medical detection method capable of detecting early tumors and pathological changes of non-obvious border tissues.

Background technique

At present, the main tumor detection method is to use existing equipment such as CT, nuclear magnetic resonance, and B-ultrasound for detection. These medical detection devices emit a certain form of energy waves to the human body, then receive signals transmitted or reflected from the human body, and finally generate images through imaging algorithms. Although medical imaging technology has made considerable progress and development, there are still many problems in clinical use, mainly including: ① It is difficult to detect early tumors and tissue lesions without obvious boundaries; The accuracy rate of continuous area detection is not high.

In addition to the above-mentioned traditional tumor detection methods, an imaging method based on tissue electrical impedance parameters has been developed in the past two decades, which are mainly divided into two categories: microwave imaging and electrical impedance imaging. After years of research, these two imaging technologies based on electromagnetic waves have made great progress in both imaging systems and imaging algorithms. For imaging, for parts containing multiple layers of tissue and each layer is very thin, the current two imaging systems are not enough to provide sufficient resolution, and can only provide the equivalent conductivity of biological tissue in a large local area. While the organism is a complex heterogeneous structure, the equivalent conductivity cannot accurately reflect the change of the electrical properties of the specific tissue of the organism.

Contents of the invention

In order to overcome the deficiencies of the existing tumor detection methods, the present invention provides an early tumor detection method, which can combine the tissue structure partition information of the region of interest obtained by the existing medical detection equipment, and refer to the electromagnetic characteristic parameter distribution database of various tissues of the human body. Different organizational structure partitions assign value ranges of electromagnetic characteristic parameters. Then, by solving the electromagnetic inverse problem, the parameter distribution of the electromagnetic characteristics of the tissue in the region of interest is obtained, which can effectively detect early tumors and tissue lesions without obvious boundaries.

The technical solution adopted by the present invention to solve its technical problems is:

The first step is to obtain the constraints of the electromagnetic inverse problem

The acquisition of the constraint conditions of the electromagnetic inverse problem includes: obtaining the electrical signal boundary value of the region of interest through the electrical signal test, and using the existing medical detection equipment to obtain the organizational structure partition information of the region of interest.

The second step is to solve the electromagnetic inverse problem

Referring to the distribution of electromagnetic characteristic parameters of various tissues of the human body, the value range of the electromagnetic characteristic parameters of each tissue structure partition is set according to the tissue structure partitions of the region of interest. Based on the boundary value of the electrical signal obtained from the electrical signal test, the improved Monte Carlo inversion method is used to solve the electromagnetic inverse problem, and the tissue electromagnetic characteristic parameters of the region of interest are inversely deduced to obtain the tissue electromagnetic characteristic parameter distribution of the region of interest information.

The third step is to determine the lesion area

According to the electromagnetic characteristic parameter distribution of the region of interest obtained by the electromagnetic inverse problem, the region with abnormal electromagnetic characteristic parameters is found.

Since the electromagnetic characteristic parameters of normal tissue and tumor tissue differ greatly in the microwave radio frequency band (the difference in dielectric parameters reaches 1:5, and the difference in magnetic permeability parameters reaches 1:10), the distribution map of electromagnetic characteristic parameters of the measured tissue can be obtained. Access to rich pathological information. However, the electromagnetic inverse problem is a kind of ill-conditioned problem, which is very sensitive to the boundary disturbance, the calculation amount is huge, and the obtained solution is not unique. Therefore, we consider combining the organizational structure partition information of the region of interest collected by existing medical detection equipment such as CT, nuclear magnetic resonance and B-ultrasound when solving the electromagnetic inverse problem, improving the algorithm of the electromagnetic inverse problem, and using the organizational structure partition information and the region of interest The distribution range information of various tissue electromagnetic characteristic parameters contained in the method is used as a priori condition, so that the detection object is no longer a "black body" but a "gray body", which greatly improves the inversion efficiency and accuracy.

The present invention combines tissue partition information and tissue electromagnetic characteristic parameter distribution range information when solving the electromagnetic inverse problem, greatly reduces the calculation amount, and can form a high-efficiency electromagnetic inverse problem solving algorithm. The present invention can calculate large-scale problems under the same hardware conditions, for example, the grid can be refined, the resolution can be improved, and the fine organizational structure can be reflected in the result.

This electromagnetic inverse problem solving method combined with existing medical detection equipment can effectively detect early tumors and tissue lesions without obvious boundaries. Strive for valuable pre-treatment time for patients.

Description of drawings

Fig. 1 is a schematic flow chart of an early tumor detection method according to the present invention;

Fig. 2 is a schematic diagram of a method for obtaining boundary values of electrical signals;

Fig. 3 is a schematic diagram of the second acquisition method of the electric signal boundary value;

Fig. 4 is the schematic flow sheet of the second step electromagnetic inverse problem solving of the present invention;

Detailed ways

Embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings.

FIG. 1 is a schematic flow chart of the method for early tumor detection according to the present invention, which includes three steps: obtaining electromagnetic inverse problem constraint conditions, solving electromagnetic inverse problem, and judging lesion area. The first step, the acquisition of the constraints of the electromagnetic inverse problem, includes the acquisition of the boundary value of the electrical signal and the acquisition of organizational structure partition information. The second step, solving the electromagnetic inverse problem, is to use the improved Monte Carlo inversion method to inversely deduce the distribution of tissue electromagnetic characteristic parameters in the region of interest. The third step, judging the lesion area, is to use the existing medical knowledge to manually search for the abnormal area of the electromagnetic characteristic parameters based on the electromagnetic characteristic parameter distribution of the region of interest obtained by the electromagnetic inverse problem. In order to achieve the purpose of detecting early tumors and tissue lesions without obvious borders.

In the first step, use the electrical signal test to detect the region of interest, and obtain the electrical signal boundary value C=[c ₁ ,c ₂ ,...,c _n ] of the region of interest, as shown in Figure 2 or Figure 3. either way. The method for obtaining the boundary value of the electrical signal shown in Fig. 2 is to excite the biological tissue 3 in the region of interest by exciting the electrode pair 1, and obtain the boundary value of the electrical signal of the biological tissue from the electrode array 2 distributed on the boundary of the biological tissue. The method for obtaining the boundary value of the electrical signal shown in Figure 4 is to transmit a broadband pulse signal to the biological tissue 3 in the region of interest through the transmitting antenna 4, and the receiving antenna array 5 receives the scattered signal in the region 6, which can be regarded as the biological tissue in the region of interest Electrical signal boundary value.

In the first step, the acquisition of tissue structure partition information is to use traditional medical detection equipment such as CT, nuclear magnetic resonance, and B-ultrasound to detect and image the area of interest to obtain a grayscale image, and perform high-pass filtering on the obtained grayscale image. The high-frequency part in the figure is extracted, and the structural partition position information of various tissues contained in the region of interest is obtained, which is used as one of the constraints for solving the electromagnetic inverse problem.

Fig. 4 is a schematic flow chart of solving the electromagnetic inverse problem described in the second step of the present invention, specifically comprising the following steps:

Step 1. Randomly generate the initial value of the parameter distribution of electromagnetic characteristics

The region of interest is divided into several grids according to the actual situation.

Using the organizational structure partition information of the region of interest obtained in the first step, referring to the distribution database of electromagnetic characteristic parameters of various tissues of the human body, according to the characteristics of different value ranges of electromagnetic characteristic parameters of different tissue structures, assign electromagnetic characteristics to each grid The initial value of the parameter G ₀ =[g _0,1 ,,g _0,2 ,...,g _0,n ] ^T , where n is the number of divided grids in the region of interest. Let the current electromagnetic characteristic parameter distribution G _NOW =G ₀ .

Step 2, calculate the fitness value under the current electromagnetic characteristic parameter distribution

The fitness value of the current electromagnetic characteristic parameter distribution is calculated by numerical method. The fitness function is proposed according to the requirements and constraints of the parameter distribution of the electromagnetic characteristics of organisms. These requirements and constraints include prior knowledge, organizational structure partition information, and excitation signal information using electrical signal testing. Suppose there are m requirements and constraints for solving this problem, then there are

$\left\{\begin{array}{c}{\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="g"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">g</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; 12</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="g"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">g</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; <figure-callout\; id="1"\; label="s"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">s</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="c"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">c</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\\ {\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; one</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="g"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">g</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; two</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="g"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">g</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; <figure-callout\; id="2"\; label="s"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">s</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="c"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">c</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\\ <span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&Center\; Dot;</span>\\ {\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="1"\; label="g"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">g</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="g"\; filenames="HSA00000020231200012.png"\; state="\{\{state\}\}">g</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; mn</span>}}{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}\end{array}\right.$

为已知的约束矩阵，G＝[g₁，g₂，…，g_n]为感兴趣区域的真实电磁特性参数向量，C＝[c₁，c₂，…，c_m]^T为得到的感兴趣区域的电信号边界值，即图2所示的电极阵2获得生物组织电信号边界值或图3所示的接收天线阵5得到的散射信号。针对本发明所涉及的问题，往往有m＜n。对此适应度函数方程组写成矩阵形式，表示为in

is the known constraint matrix, G=[g ₁ , g ₂ ,…,g _n ] is the real electromagnetic characteristic parameter vector of the region of interest, C=[c ₁ ,c ₂ ,…,c _m ] ^T is the obtained The boundary value of the electrical signal of the region of interest, that is, the boundary value of the electrical signal of the biological tissue obtained by the electrode array 2 shown in FIG. 2 or the scattered signal obtained by the receiving antenna array 5 shown in FIG. 3 . For the problems involved in the present invention, it is often m<n. The fitness function equations are written in matrix form, expressed as

SG=C

Calculate the fitness value ∥ΔC‖=‖CC _NOW ‖ under the current electromagnetic characteristic parameter distribution, where C _NOW =SG _NOW .

Step 3, judging whether the current electromagnetic characteristic parameter distribution meets the requirements or the number of iterations is full. If the fitness value of the current electromagnetic characteristic parameter distribution meets the requirements or the number of iterations is full, exit the loop and output the electromagnetic characteristic parameter distribution G _BEST corresponding to the minimum fitness value as the calculation result; otherwise, perform step 4.

Step 4: Adjust the electromagnetic characteristic parameters of each part of the region of interest within the assigned value range of the electromagnetic characteristic parameters to obtain the electromagnetic characteristic parameter distribution G _NOW for the next round of iterative calculation, and return to step 2.

Claims (2)

1. A method for early tumor detection, specifically comprising the steps of: The first step is to obtain the constraints of the electromagnetic inverse problem Obtain the electrical signal boundary value of the region of interest through the electrical signal test, and use the existing medical detection equipment to obtain the organizational structure partition information of the region of interest; The second step is to solve the electromagnetic inverse problem Referring to the distribution of electromagnetic characteristic parameters of various tissues of the human body, set the value range of electromagnetic characteristic parameters of each tissue structure partition according to the tissue structure partition of the region of interest; use the improved Monte The Carlo inversion method solves the electromagnetic inverse problem, and obtains the tissue electromagnetic characteristic parameter distribution information of the region of interest; The third step is to determine the lesion area According to the electromagnetic characteristic parameter distribution of the region of interest obtained by the electromagnetic inverse problem, the region with abnormal electromagnetic characteristic parameters is found. 2. The early tumor detection method according to claim 1, wherein the solving of the electromagnetic inverse problem described in the second step specifically comprises the following steps: Step 1. Randomly generate the initial value of the parameter distribution of electromagnetic characteristics Divide the region of interest into several grids according to the actual situation; assign the initial value G ₀ of electromagnetic characteristic parameters to each grid =[g _0,1, ,g _0,2 ,...,g _0,n ] ^T ; let Current electromagnetic characteristic parameter distribution G _NOW =G ₀ ; Step 2, calculate the fitness value under the current electromagnetic characteristic parameter distribution Calculate the fitness value under the current electromagnetic characteristic parameter distribution ||ΔC||=||CC _NOW ||, where C _NOW =SG _NOW , S is the known constraint matrix, C=[c ₁ ,c ₂ ,…, c _m ] ^T is the electrical signal boundary value of the region of interest obtained; Step 3, judge whether the current electromagnetic characteristic parameter distribution meets the requirements or the number of iterations is full If the fitness value of the current electromagnetic characteristic parameter distribution meets the requirements or the number of iterations is full, exit the loop, and output the electromagnetic characteristic parameter distribution G _BEST corresponding to the minimum fitness value as the calculation result; otherwise, perform step 4; Step 4: Adjust the electromagnetic characteristic parameters of each part of the region of interest within the assigned value range of the electromagnetic characteristic parameters to obtain the electromagnetic characteristic parameter distribution G _NOW for the next round of iterative calculation, and return to step 2.

Images