Accurate positioning method for prostate tumor cell therapy

Abstract

The invention relates to the field of auxiliary detection, in particular to a precise positioning method for prostate tumor cell therapy, and the method comprises the steps: obtaining an organ image collected by a collection terminal; dividing the organ image into a plurality of image areas, identifying nucleoplasm features of cell nucleuses in each image area, and analyzing feature obvious characterization values of each image area according to the nucleoplasm features so as to divide feature obvious degree categories of each image area; according to the method, the tumor cells in each image area are adaptively identified and positioned according to the feature obvious degree category, the prostate tumor cells can be accurately positioned, and the accuracy of tumor cell detection is improved.

Description

Technical Field

[0001] The present invention relates to the field of auxiliary detection, and in particular to a precise positioning method for prostate tumor cell therapy. Background Art

[0002] SEA-MAP (Single Epitope Multigen Antigen Peptide) is a single epitope multimeric antigen peptide. It is based on molecular modeling of bioinformatics to predict antigen epitope sequences, screen out effective peptide fragments, and then use combinatorial chemistry to synthesize dominant epitope peptides. By adding a hydrophobic branched oligolysine core skeleton, the orientation of the connected epitope peptides is made consistent, exponentially increasing the specific binding ability and reaction sensitivity of the epitope peptides within a unit space, and can be used for the development of a series of targeted substances for different tumors and. However, when this technology is applied to the treatment of prostate tumors, SEA-MAP (single surface multimeric antigen) needs to be accurately injected into prostate tumor cells to ensure that SEA-MAP can accurately target tumor cells for treatment, that is, this technology is highly dependent on the precise positioning of the prostate tumor.

[0003] Currently, many achievements in the field of physics have been introduced into the localization of prostate tumor cells. For example, the positioning equipment developed using the new electromagnetic induction principle assists in positioning by detecting the differences in electromagnetic properties between prostate tumor cells and surrounding normal tissues. In optical imaging technology, based on the characteristics of near-infrared light and other light that interact with biological tissues, an optical imaging system suitable for prostate tumor positioning is developed to determine the location of tumor cells by detecting changes in light absorption and scattering. Another example is the use of image analysis technology to analyze organ images to determine the location of lesions. These interdisciplinary combinations provide more new ideas and technical means for precise positioning.

[0004] For example, Chinese patent application publication number: CN114874989A, discloses a method for capturing circulating tumor cells, comprising the following steps: after the red blood cells of the cell fluid to be tested are lysed, they are added to a culture container provided with a cell chip, the cell chip includes a matrix and blind holes arrayed on the surface of the matrix, after the cells enter the blind holes, culture medium is added for culture, during the culture process, the circulating tumor cells in the cell fluid to be tested can proliferate into cell spheres, fluorescent antibodies that specifically react with circulating tumor cells are added, and the circulating tumor cells are located according to the fluorescence and cell proliferation conditions, and the CTC can be accurately identified and located by using the regularly arranged micropores, fluorescent antibodies, cell morphology and proliferation conditions.

[0005] However, there are still the following problems in the prior art:

[0006] During the screening and positioning of prostate tumor cells, the image screening method used did not take into account the prominence of features in the image, and did not adaptively change the image analysis method, resulting in poor accuracy in prostate tumor cell screening and positioning. Summary of the invention

[0007] To this end, the present invention provides a precise positioning method for prostate tumor cell treatment, so as to overcome the problem in the prior art that the image screening method adopted does not take into account the prominence of features in the image and does not adaptively change the image analysis method, resulting in poor accuracy in prostate tumor cell screening and positioning.

[0008] To achieve the above objectives, the present invention provides a method for accurate positioning in prostate tumor cell therapy, comprising:

[0009] Acquire organ images acquired by an acquisition terminal;

[0010] Dividing the organ image into a plurality of image regions, identifying nuclear-cytoplasmic features of the cell nucleus in each of the image regions, and analyzing the feature obviousness characterization value of each of the image regions according to the nuclear-cytoplasmic features to classify the feature obviousness degree category of each of the image regions, wherein the nuclear-cytoplasmic features include the nuclear-cytoplasmic ratio of the cell nucleus and the uniformity of the cell size;

[0011] According to the feature prominence category, tumor cells in each image region are identified and located, including:

[0012] Identify suspected tumor cell clusters and determine suspected abnormal regions according to the adhesion benchmark conditions between cells in the cell clusters in the image area, and determine tissue abnormality characterization coefficients according to the compression deformation of tissue in the suspected abnormal region and the arrangement neatness of cells in the suspected tumor cell clusters, so as to mark the suspected abnormal region;

[0013] Alternatively, the cell mutation characterization coefficient is determined according to the roughness of the cell surface texture and the clarity of the cell contour boundary in each of the image regions, so as to determine and mark the suspected mutation region.

[0014] Further, the characteristic significant characterization value of each of the image regions is analyzed according to the nuclear-cytoplasmic characteristics, including:

[0015] The ratio of the nuclear-cytoplasmic ratio of the cell nucleus to the nuclear-cytoplasmic ratio threshold is taken as the first obvious feature;

[0016] The ratio of the uniformity of cell size to the uniformity threshold is used as the second obvious feature;

[0017] A weighted sum of the first obvious feature and the second obvious feature is determined as a feature obvious characterization value.

[0018] Furthermore, the feature prominence level of each of the image regions is classified into categories, including:

[0019] If the feature obvious characterization value is greater than or equal to the feature obvious characterization threshold, then the feature of the image area is determined to be a high obviousness category;

[0020] If the feature conspicuousness value is less than the feature conspicuousness threshold, it is determined that the feature of the image region is of a low conspicuousness category.

[0021] Furthermore, according to the feature prominence category, tumor cells in each of the image regions are identified and located, including:

[0022] If the feature of the image area is a high-visibility category, then the suspected tumor cell cluster is identified and the suspected abnormal region is determined according to the adhesion benchmark condition between the cells in the cell cluster of the image area, and the tissue abnormality characterization coefficient is determined according to the compression deformation amount of the tissue in the suspected abnormal region and the arrangement neatness of the cells in the suspected tumor cell cluster, so as to mark the suspected abnormal region;

[0023] If the feature of the image area is of the low-visibility category, the cell mutation characterization coefficient is determined according to the cell surface texture roughness and the cell contour boundary clarity of each image area to determine the suspected mutation area and mark it.

[0024] Further, identifying suspected tumor cell clusters and determining suspected mutation areas according to the adhesion benchmark conditions between cells in the cell clusters in the image area includes:

[0025] Determining each closed contour curve within the cell cluster;

[0026] Determining an adhesion reference condition according to the spacing between the closed contour curves and the contact area;

[0027] If there is any cell cluster that meets the adhesion benchmark condition, the cell cluster is determined as a suspected tumor cell cluster, and the area where the suspected tumor cell cluster is located is determined as a suspected mutation area;

[0028] The adhesion benchmark condition is that the spacing between any closed contour curves in the cell cluster is greater than a spacing threshold and / or the contact area between any closed contour curves is less than or equal to a contact area threshold.

[0029] Furthermore, the tissue mutation characterization coefficient is determined according to the amount of compression deformation of the tissue in the suspected mutation area and the arrangement neatness of the cells in the suspected tumor cell cluster, including:

[0030] The ratio of the tissue's compression deformation to the deformation threshold is used as the first abnormal feature;

[0031] The ratio of the neatness threshold to the arrangement neatness is used as the second mutation feature;

[0032] Determine the sum of the first abnormality feature and the second abnormality feature as a tissue abnormality characterization coefficient;

[0033] The cell arrangement image of the suspected abnormal region is obtained, and the cell arrangement image is matched with the cell arrangement sample image to determine the regularity of cell arrangement in the suspected tumor cell cluster.

[0034] Further, the suspected mutation region is marked, including:

[0035] If the tissue mutation characterization coefficient is greater than or equal to the tissue mutation characterization coefficient threshold, the suspected mutation area is marked.

[0036] Furthermore, the cell mutation characterization coefficient is determined according to the cell surface texture roughness of the suspected mutation area and the cell outline boundary clarity, including:

[0037] The relative variance of the surface texture roughness of each cell in the suspected mutation region is used as the first cell feature;

[0038] The relative variance of the outline boundary clarity of each cell in the suspected mutation region is used as the second cell feature;

[0039] The sum of the first cell feature and the second cell feature is determined as a cell mutation characterization coefficient.

[0040] Furthermore, the process of determining the surface texture roughness of each cell and the outline boundary clarity of each cell includes:

[0041] Performing texture analysis on the surface of each cell to determine an energy parameter, and determining the energy parameter as the texture roughness of the cell surface;

[0042] According to matching the cell contour boundary lines in the suspected abnormal region with the normal cell contour boundary lines, the boundary line fitting degree is determined, and the boundary line fitting degree is determined as the contour boundary clarity;

[0043] The texture analysis of the cell surface includes determining a gray level co-occurrence matrix of the image within the cell contour boundary, and calculating energy parameters based on the gray level co-occurrence matrix.

[0044] Furthermore, the suspected abnormal region is marked, including, if there is any image region whose cell abnormality characterization coefficient is greater than or equal to a cell abnormality characterization coefficient threshold, determining the image region as a suspected abnormal region and marking it.

[0045] Compared with the prior art, the present invention obtains organ images collected by an acquisition terminal; divides the organ images into several image areas, identifies the nuclear and cytoplasmic characteristics of the cell nuclei in each image area, analyzes the feature obviousness characterization value of each image area based on the nuclear and cytoplasmic characteristics, and divides the feature obviousness category of each image area; according to the feature obviousness category, adaptively identifies and locates tumor cells in each image area. The present invention can accurately locate prostate tumor cells and improve the accuracy of tumor cell detection.

[0046] In particular, the present invention first processes the collected organ images through image processing technology to quickly identify dominant features, divide the feature significance of each image area, consumes less computing power, and for determining the feature significance characterization value of each image area, the present invention selects nuclear-cytoplasmic features for identification, including the nuclear-cytoplasmic ratio of the cell nucleus and the uniformity of cell size. These two features can be intuitively observed in the cross-section image, which saves computing power resources while improving the efficiency of locating tumor cells.

[0047] In particular, the present invention identifies and locates tumor cells in different image regions in a targeted manner according to the different categories of feature prominence of each image region, thereby improving the efficiency of locating tumor cells.

[0048] In particular, the present invention identifies suspected tumor cell clusters by observing the adhesion baseline conditions between cells in the cell clusters in the image area. Cells in normal tissues tend to have relatively stable and regular adhesion patterns, while tumor cells have different degrees of abnormal adhesion due to changes in their biological characteristics. Through multiple characteristic conditions that can be observed from the image and can characterize the cell adhesion, it is determined whether the cells in the area are suspected tumor cells, and the suspected tumor cell clusters are determined as suspected abnormal areas. Since the cells are pathologically damaged, the degree of compression and deformation of the tissue and the neat arrangement of the cells are affected to a certain extent. Therefore, the present invention determines the tissue mutation characterization coefficient according to the compression deformation of the tissue in the suspected abnormal area and the neat arrangement of the cells in the suspected tumor cell cluster to characterize the degree of tendency of the cells to mutate, thereby accurately locating prostate tumor cells and improving the accuracy of tumor cell detection.

[0049] In particular, the present invention considers the situation where the characteristics of the image area are of low prominence, that is, in the early stage of the appearance of tumor cells, the characteristics presented by the cells in the organ image are not obvious enough, but there may be a potential risk of cell mutation, and it is represented in the image data. Therefore, the present invention further analyzes the image area, and determines the cell mutation characterization coefficient based on the two cellular-level characteristics of the cell surface texture roughness and the contour boundary clarity that can be quickly and easily obtained to represent the degree of tendency of the cell to mutate, thereby accurately locating the prostate tumor cells and improving the accuracy of tumor cell detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A schematic diagram of the steps of a method for precise positioning in the treatment of prostate tumor cells according to an embodiment of the invention;

[0051] Figure 2 A logical decision diagram for classifying the feature prominence categories of each image region in an embodiment of the invention;

[0052] Figure 3 A logical decision diagram for identifying and locating tumor cells in each image region in an embodiment of the invention;

[0053] Figure 4 A logical decision diagram for identifying suspected tumor cell clusters according to an embodiment of the invention. DETAILED DESCRIPTION

[0054] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0055] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0056] It should be noted that, in the description of the present invention, the terms such as “inside” and the like indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0057] See also Figure 1 As shown, 1 is a schematic diagram of the steps of the method for accurate positioning in the treatment of prostate tumor cells according to an embodiment of the present invention. The method for accurate positioning in the treatment of prostate tumor cells according to an embodiment of the present invention includes:

[0058] Step S1, acquiring an organ image acquired by an acquisition terminal;

[0059] Step S2, dividing the organ image into a plurality of image regions, identifying the nuclear-cytoplasmic features of the cell nucleus in each of the image regions, and analyzing the feature obviousness characterization value of each of the image regions according to the nuclear-cytoplasmic features to classify the feature obviousness degree category of each of the image regions, wherein the nuclear-cytoplasmic features include the nuclear-cytoplasmic ratio of the cell nucleus and the uniformity of the cell size;

[0060] Step S3, identifying and locating tumor cells in each of the image regions according to the feature prominence category, including:

[0061] Identify suspected tumor cell clusters and determine suspected abnormal regions according to the adhesion benchmark conditions between cells in the cell clusters in the image area, and determine tissue abnormality characterization coefficients according to the compression deformation of tissue in the suspected abnormal region and the arrangement neatness of cells in the suspected tumor cell clusters, so as to mark the suspected abnormal region;

[0062] Alternatively, the cell mutation characterization coefficient is determined according to the roughness of the cell surface texture and the clarity of the cell contour boundary in each of the image regions, so as to determine and mark the suspected mutation region.

[0063] Specifically, there is no limitation on the specific structure of the acquisition terminal, and it only needs to have an instrument that can acquire organ images, such as a CT scanning device. Of course, the organ image at the target location can also be acquired through other existing technologies, which will not be repeated here.

[0064] Specifically, the characteristic significant characterization value of each of the image regions is analyzed according to the nuclear-cytoplasmic characteristics, including:

[0065] The ratio of the nuclear-cytoplasmic ratio of the cell nucleus to the nuclear-cytoplasmic ratio threshold is taken as the first obvious feature;

[0066] The ratio of the uniformity of cell size to the uniformity threshold is used as the second obvious feature;

[0067] A weighted sum of the first obvious feature and the second obvious feature is determined as a feature obvious characterization value.

[0068] In this implementation, when performing weighted summation, the weight of the first obvious feature is set to 0.55, and the weight of the second obvious feature is set to 0.45;

[0069] It can be understood that the nuclear-to-cytoplasmic ratio refers to the ratio of the volume of the cell nucleus to the volume of the cytoplasm. For normal prostate cells, the cell nucleus is of moderate size and the cytoplasm occupies a relatively appropriate proportion, maintaining the normal physiological functions and structural characteristics of the cells. In this embodiment, the threshold value of the nuclear-to-cytoplasmic ratio is selected within the interval [0.17, 0.25];

[0070] When the prostate develops abnormal conditions such as cancer, tumor cells often show changes such as enlarged nuclei and increased nuclear-cytoplasmic ratio. The nuclear-cytoplasmic ratio of some prostate cancer cells can reach 1 or even higher, accompanied by other cellular atypia such as irregular nuclear morphology and significantly enlarged nucleoli. Observation and comparison of the nuclear-cytoplasmic ratio and these related cell characteristics can help accurately determine whether the cells have diseased under an electron microscope. The preferred nuclear-cytoplasmic ratio threshold in this embodiment is 0.25. The degree of change in the nuclear-cytoplasmic ratio of cells is used to characterize the tendency of normal prostate cells to develop diseases and form tumor cells.

[0071] It can be understood that the uniformity of cell size refers to the image area variance formed by the image within the cell contour boundary, which will not be elaborated here.

[0072] The uniformity threshold of cell size is preset, and relevant data of a number of normal prostate cells are obtained, the cell size uniformity data are called, the uniformity mean is solved, and the cell size uniformity threshold is set to 1.1 times to 1.13 times the uniformity mean.

[0073] Specifically, the present invention first processes the collected organ images through image processing technology to quickly identify dominant features, divide the feature significance of each image area, consume less computing power, and for determining the feature significance characterization value of each image area, the present invention selects nuclear-cytoplasmic features for identification, including the nuclear-cytoplasmic ratio of the cell nucleus and the uniformity of cell size. These two features can be intuitively observed in the cross-section image, which saves computing power resources while improving the efficiency of locating tumor cells.

[0074] Specifically, see Figure 2 As shown, it is a logical decision diagram for classifying the feature conspicuity categories of each image region according to an embodiment of the present invention. The feature conspicuity categories of each image region are classified, including:

[0075] If the feature obvious characterization value is greater than or equal to the feature obvious characterization threshold, then the feature of the image area is determined to be a high obviousness category;

[0076] If the feature conspicuousness value is less than the feature conspicuousness threshold, it is determined that the feature of the image region is of a low conspicuousness category.

[0077] The feature-significant characterization threshold is selected in the interval [2.66, 2.78].

[0078] Specifically, see Figure 3 As shown, it is a logical decision diagram for identifying and locating tumor cells in each image area according to an embodiment of the present invention. Identifying and locating tumor cells in each image area includes:

[0079] If the feature of the image area is a high-visibility category, then the suspected tumor cell cluster is identified and the suspected abnormal region is determined according to the adhesion benchmark condition between the cells in the cell cluster of the image area, and the tissue abnormality characterization coefficient is determined according to the compression deformation amount of the tissue in the suspected abnormal region and the arrangement neatness of the cells in the suspected tumor cell cluster, so as to mark the suspected abnormal region;

[0080] If the feature of the image area is of the low-visibility category, the cell mutation characterization coefficient is determined according to the cell surface texture roughness and the cell contour boundary clarity of each image area to determine the suspected mutation area and mark it.

[0081] The present invention identifies and locates tumor cells in different image areas in a targeted manner according to the different categories of feature prominence of each image area, thereby improving the efficiency of locating tumor cells.

[0082] Specifically, for image areas of high-visibility categories, this embodiment identifies suspected tumor cell clusters by observing the adhesion baseline conditions between cells in the cell clusters in the image area. Cells in normal tissues tend to have relatively stable and regular adhesion patterns, while tumor cells have different degrees of abnormal adhesion between cells due to changes in their biological characteristics. For example, the adhesion between tumor cells is reduced, and they are likely to detach from the primary site and metastasize. Therefore, through multiple feature conditions that can be observed from the image and can characterize the cell adhesion, it is determined whether the cells in the area are suspected tumor cells, and the suspected tumor cell clusters are determined as suspected mutation areas.

[0083] Specifically, see Figure 4 As shown, it is a logical decision diagram for identifying suspected tumor cell clusters according to an embodiment of the present invention. According to the feature prominence category, the suspected tumor cell clusters are identified and the suspected mutation area is determined according to the adhesion benchmark condition between the cells in the cell clusters in the image area, including:

[0084] Determining each closed contour curve within the cell cluster;

[0085] Determining an adhesion reference condition according to the spacing between the closed contour curves and the contact area;

[0086] If there is any cell cluster that meets the adhesion benchmark condition, the cell cluster is determined as a suspected tumor cell cluster, and the area where the suspected tumor cell cluster is located is determined as a suspected mutation area;

[0087] The adhesion benchmark condition is that the spacing between any closed contour curves in the cell cluster is greater than a spacing threshold and / or the contact area between any closed contour curves is less than or equal to a contact area threshold.

[0088] In this embodiment, the spacing threshold between any closed contour curves and the contact area threshold between any closed contour curves are preset, and the spacing data and contact area data between a number of normal prostate cell clusters are called to solve the spacing mean and the contact area mean, and the spacing threshold between any closed contour curves is set to 1.05 times to 1.15 times the spacing mean, and the contact area threshold between any closed contour curves is set to 1.1 times to 1.3 times the contact area mean;

[0089] The spacing between any closed contour curves refers to the longest distance from the edge of any closed contour curve to the edge of the adjacent closed contour curve, which will not be described in detail.

[0090] Normal prostate cells form tumor cells due to cell mutation or other influences. At the same time, due to the proliferation characteristics of tumor cells, rapid proliferation will form accumulation and clustering. In this embodiment, the cell cluster that meets the adhesion benchmark condition is determined as a suspected tumor cell cluster, and the area where the suspected tumor cell cluster is located is determined as a suspected mutation area;

[0091] It can be understood that the state of the cells in the image is the curve shape of the closed contour. The suspected tumor cell clusters and the suspected mutation areas are determined according to the distribution of the closed contour curves in the cell clusters. Since the adhesion of tumor cells is lower than that of normal cells and they are easy to detach from the original position, this embodiment considers the spacing between the closed contour curves as one of the benchmark conditions for measuring the adhesion, and the contact situation in the spatial dimension can also reflect the degree of adhesion between cells. The state presented in the plane image is the situation where adhesion occurs between any closed contours. The increase in the contact area of ​​the closed contour after adhesion indicates that the adhesion is stronger. Therefore, this embodiment also uses the contact area between the closed contour curves as a benchmark condition for measuring the adhesion. Combining the two benchmark conditions can more comprehensively and accurately identify suspected tumor cell clusters.

[0092] Specifically, the tissue mutation characterization coefficient is determined according to the compression deformation of the tissue in the suspected mutation area and the arrangement neatness of the cells in the suspected tumor cell cluster, including:

[0093] The ratio of the tissue's compression deformation to the deformation threshold is used as the first abnormal feature;

[0094] The ratio of the neatness threshold to the arrangement neatness is used as the second mutation feature;

[0095] Determine the sum of the first abnormality feature and the second abnormality feature as a tissue abnormality characterization coefficient;

[0096] The cell arrangement image of the suspected abnormal region is obtained, and the cell arrangement image is matched with the cell arrangement sample image to determine the regularity of cell arrangement in the suspected tumor cell cluster.

[0097] It can be understood that the compressive deformation amount of the tissue refers to the degree of morphological change in the tissue structure due to compression. In this embodiment, the tissue structure image in the normal cell image area is called and compared with the tissue structure image in the suspected abnormal area. The bulge or depression of the tissue structure is determined as the compressive deformation of the tissue, and the maximum bulge length or the maximum depression length is determined as the compressive deformation amount of the tissue. This will not be repeated.

[0098] In this embodiment, the compression deformation threshold and the arrangement regularity threshold of the tissue are preset, and image data of a number of normal prostate tissues are obtained, and the compression deformation data and the arrangement regularity data of the tissue are called, and the compression deformation mean and the arrangement regularity mean are solved, and the compression deformation threshold of the tissue is set to 1.15 times to 1.3 times the compression deformation mean, and the arrangement regularity threshold is set to 1.15 times to 1.3 times the arrangement regularity mean;

[0099] It can be understood that the cell arrangement sample image is a pre-collected arrangement image of normal prostate cells. Of course, the cell arrangement sample image can also be obtained by other means. For the matching of the cell arrangement image with the cell arrangement sample image, the existing image fitting technology can be used for matching to determine the similarity between the two images and characterize the cell arrangement neatness. If the similarity is too low, it may indicate that the cell arrangement neatness is low and the abnormality of the formed cell cluster is high. There is no limitation on the algorithm for determining the image similarity. For example, the cosine similarity after image vectorization can be used to characterize the similarity of the two images. Other methods can also be used, which will not be repeated here.

[0100] Specifically, the tissue abnormality characterization coefficient is determined according to the amount of compression deformation of the tissue in the suspected abnormal area and the arrangement neatness of the cells in the suspected tumor cell cluster. In actual situations, due to cell pathology, the degree of compression deformation of the tissue and the arrangement neatness of the cells will be affected to a certain extent. For example, the unlimited proliferation characteristics of tumor cells will cause them to continuously divide and grow, and their number in the local area will increase rapidly, gradually occupying the space originally belonging to normal cells, forcing the normal tissue to change its original morphology and deform to varying degrees. At the same time, the density of tumor cells is often higher than that of normal cells. Due to the gradual increase in internal pressure of tumor cells, the tumor cells will expand toward the surrounding normal tissues. "Pressure" is applied to the tissue to cause a certain degree of deformation; since tumor cells capture the constraints of the normal body's regulatory mechanism and proliferate disorderly, these disorderly proliferating tumor cells will be interspersed among normal cells, disrupting the original cell arrangement order, making the cell arrangement order chaotic, and, due to the factor of reduced adhesion, tumor cells are more likely to detach from the cell accumulation group, which makes it difficult to maintain a neat arrangement of cells, and they are mostly loose and disordered distribution, destroying the neatness of cell arrangement and the normal and orderly cell distribution pattern. Therefore, this embodiment uses the tissue mutation characterization coefficient to characterize the degree of tendency of cells to mutate, so as to mark the suspected mutation area.

[0101] Specifically, marking the suspected mutation region includes:

[0102] If the tissue mutation characterization coefficient is greater than or equal to the tissue mutation characterization coefficient threshold, marking the suspected mutation area;

[0103] If the tissue mutation characterization coefficient is less than the tissue mutation characterization coefficient threshold, there is no need to mark the suspected mutation area.

[0104] The threshold of tissue variation characterization coefficient is selected in the interval [1.26,1.37].

[0105] Specifically, the cell mutation characterization coefficient is determined according to the cell surface texture roughness of the suspected mutation area and the cell outline boundary clarity, including:

[0106] The relative variance of the surface texture roughness of each cell in the suspected mutation region is used as the first cell feature;

[0107] The relative variance of the outline boundary clarity of each cell in the suspected mutation region is used as the second cell feature;

[0108] The sum of the first cell feature and the second cell feature is determined as a cell mutation characterization coefficient.

[0109] It is understandable that normal cells have a relatively smooth and regular surface texture, while in a region with lesions, the surface texture roughness and contour boundary clarity of tumor cells vary. There may be tiny protrusions and depressions on the surface of tumor cells. The protrusions and depressions can be determined in the image based on the number of wrinkles on the cell surface. At the same time, the contour boundary of the tumor cell will appear fuzzy due to the lesions, making it difficult to demarcate the boundaries. Therefore, this embodiment determines the boundary clarity of the cell by the completeness of the cell contour boundary. If the boundary line is intermittent, it means that the boundary clarity of the cell is low.

[0110] Specifically, the process of determining the surface texture roughness of each cell and the outline boundary clarity of each cell includes:

[0111] Performing texture analysis on the surface of each cell to determine an energy parameter, and determining the energy parameter as the texture roughness of the cell surface;

[0112] According to matching the cell contour boundary lines in the suspected abnormal region with the normal cell contour boundary lines, the boundary line fitting degree is determined, and the boundary line fitting degree is determined as the contour boundary clarity;

[0113] The texture analysis of the cell surface includes determining a gray level co-occurrence matrix of the image within the cell contour boundary, and calculating energy parameters based on the gray level co-occurrence matrix.

[0114] Specifically, the energy parameter is the sum of squares of the elements of the gray-level co-occurrence matrix, also known as the second-order moment, which will not be described in detail here.

[0115] Specifically, marking the suspected mutation region includes:

[0116] If there is any image region whose cell mutation characterization coefficient is greater than or equal to the cell mutation characterization coefficient threshold, the image region is determined as a suspected mutation region and marked.

[0117] Specifically, the present invention considers the situation where the characteristics of the image area are of low prominence, that is, in the early stage of the appearance of tumor cells, the characteristics presented by the cells in the organ image are not obvious enough, but there may be a potential risk of cell mutation, and it is represented in the image data. Therefore, the present invention further analyzes the image area, and determines the cell mutation characterization coefficient based on the two cellular-level characteristics of the cell surface texture roughness and the contour boundary clarity that can be quickly and easily obtained to characterize the degree of tendency of the cell to mutate, thereby accurately locating the prostate tumor cells and improving the accuracy of tumor cell detection.

[0118] If the precise positioning method for prostate tumor cell treatment of the present invention is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention, and the aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0119] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A method for accurate positioning in prostate tumor cell therapy, characterized in that: include: Acquire organ images acquired by an acquisition terminal; Dividing the organ image into a plurality of image regions, identifying nuclear-cytoplasmic features of the cell nucleus in each of the image regions, and analyzing the feature obviousness characterization value of each of the image regions according to the nuclear-cytoplasmic features to classify the feature obviousness degree category of each of the image regions, wherein the nuclear-cytoplasmic features include the nuclear-cytoplasmic ratio of the cell nucleus and the uniformity of the cell size; According to the feature prominence category, tumor cells in each image region are identified and located, including: Identify suspected tumor cell clusters and determine suspected abnormal regions according to the adhesion benchmark conditions between cells in the cell clusters in the image area, and determine tissue abnormality characterization coefficients according to the compression deformation of tissue in the suspected abnormal region and the arrangement neatness of cells in the suspected tumor cell clusters, so as to mark the suspected abnormal region; Alternatively, the cell mutation characterization coefficient is determined according to the roughness of the cell surface texture and the clarity of the cell contour boundary in each of the image regions, so as to determine and mark the suspected mutation region.

2. The precise positioning method for prostate tumor cell therapy according to claim 1, characterized in that: Analyzing the characteristic significant characterization value of each of the image regions according to the nuclear and cytoplasmic characteristics includes: The ratio of the nuclear-cytoplasmic ratio of the cell nucleus to the nuclear-cytoplasmic ratio threshold is taken as the first obvious feature; The ratio of the uniformity of cell size to the uniformity threshold is used as the second obvious feature; A weighted sum of the first obvious feature and the second obvious feature is determined as a feature obvious characterization value.

3. The precise positioning method for prostate tumor cell therapy according to claim 1, characterized in that: Classifying the feature prominence levels of each of the image regions into categories, including: If the feature obvious characterization value is greater than or equal to the feature obvious characterization threshold, then the feature of the image area is determined to be a high obviousness category; If the feature conspicuousness value is less than the feature conspicuousness threshold, it is determined that the feature of the image region is of a low conspicuousness category.

4. The precise positioning method for prostate tumor cell therapy according to claim 1, characterized in that: According to the feature prominence category, tumor cells in each image region are identified and located, including: If the feature of the image area is a high-visibility category, then the suspected tumor cell cluster is identified and the suspected abnormal region is determined according to the adhesion benchmark condition between the cells in the cell cluster of the image area, and the tissue abnormality characterization coefficient is determined according to the compression deformation amount of the tissue in the suspected abnormal region and the arrangement neatness of the cells in the suspected tumor cell cluster, so as to mark the suspected abnormal region; If the feature of the image area is of the low-visibility category, the cell mutation characterization coefficient is determined according to the cell surface texture roughness and the cell contour boundary clarity of each image area to determine the suspected mutation area and mark it.

5. The precise positioning method for prostate tumor cell therapy according to claim 1, characterized in that: Identifying suspected tumor cell clusters and determining suspected abnormal regions according to the adhesion benchmark conditions between cells in the cell clusters in the image region includes: Determining each closed contour curve within the cell cluster; Determining an adhesion reference condition according to the spacing between the closed contour curves and the contact area; If there is any cell cluster that meets the adhesion benchmark condition, the cell cluster is determined as a suspected tumor cell cluster, and the area where the suspected tumor cell cluster is located is determined as a suspected mutation area; The adhesion benchmark condition is that the spacing between any closed contour curves in the cell cluster is greater than a spacing threshold and / or the contact area between any closed contour curves is less than or equal to a contact area threshold.

6. The precise positioning method for prostate tumor cell therapy according to claim 1, characterized in that: The tissue mutation characterization coefficient is determined according to the compression deformation of the tissue in the suspected mutation area and the arrangement neatness of the cells in the suspected tumor cell cluster, including: The ratio of the tissue's compression deformation to the deformation threshold is used as the first abnormal feature; The ratio of the neatness threshold to the arrangement neatness is used as the second mutation feature; Determine the sum of the first abnormality feature and the second abnormality feature as a tissue abnormality characterization coefficient; The cell arrangement image of the suspected mutation area is obtained, and the cell arrangement image is matched with the cell arrangement sample image to determine the regularity of cell arrangement in the suspected tumor cell cluster.

7. The precise positioning method for prostate tumor cell therapy according to claim 6, characterized in that: Marking the suspected mutation region includes: If the tissue mutation characterization coefficient is greater than or equal to the tissue mutation characterization coefficient threshold, the suspected mutation area is marked.

8. The precise positioning method for prostate tumor cell therapy according to claim 1, characterized in that: The cell mutation characterization coefficient is determined according to the cell surface texture roughness of the suspected mutation area and the cell outline boundary clarity, including: The relative variance of the surface texture roughness of each cell in the suspected mutation region is used as the first cell feature; The relative variance of the outline boundary clarity of each cell in the suspected mutation region is used as the second cell feature; The sum of the first cell feature and the second cell feature is determined as a cell mutation characterization coefficient.

9. The precise positioning method for prostate tumor cell therapy according to claim 8, characterized in that: The process of determining the roughness of the surface texture of each cell and the clarity of the outline boundary of each cell includes: Performing texture analysis on the surface of each cell to determine an energy parameter, and determining the energy parameter as the texture roughness of the cell surface; According to matching the cell contour boundary lines in the suspected abnormal region with the normal cell contour boundary lines, the boundary line fitting degree is determined, and the boundary line fitting degree is determined as the contour boundary clarity; The texture analysis of the cell surface includes determining a gray level co-occurrence matrix of the image within the cell contour boundary, and calculating energy parameters based on the gray level co-occurrence matrix.

10. The precise positioning method for prostate tumor cell therapy according to claim 8, characterized in that: Marking the suspected mutation region includes: If there is any image region whose cell mutation characterization coefficient is greater than or equal to the cell mutation characterization coefficient threshold, the image region is determined as a suspected mutation region and marked.

Images