Targeted molecule drug loading efficiency detection method and system based on fluorescence imaging

Abstract

The invention relates to the technical field of fluorescence image processing, in particular to a targeted molecule drug loading efficiency detection method and system based on fluorescence imaging, and the method comprises the following steps: constructing a targeted nano-carrier, incubating the targeted nano-carrier and a to-be-detected cell culture dish to obtain an incubation culture dish, and sequentially performing fluorescence excitation and image shooting on the incubation culture dish to obtain a targeted molecule drug loading efficiency detection result. The method comprises the following steps: obtaining an original fluorescence image, carrying out cluster division on the original fluorescence image to obtain an initial cluster group, carrying out cluster segmentation by using the initial cluster group to obtain a fluorescence segmentation image, and carrying out drug loading efficiency detection according to the fluorescence segmentation image to obtain a drug loading cell rate. According to the invention, the accuracy of fluorescent drug-loading image segmentation can be improved, and the overall reliability of drug-loading efficiency detection is enhanced.

Description

Technical Field

[0001] This invention relates to the field of fluorescence image processing technology, and in particular to a method and system for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging. Background Technology

[0002] Targeted molecule drug delivery efficiency testing is a core step in modern targeted drug development and evaluation. It determines whether a drug delivery system can accurately and efficiently deliver therapeutic drugs to the lesion site. By precisely quantifying this efficiency, researchers can objectively evaluate the design quality of targeted molecules, avoid "accidentally damaging" healthy tissues during drug delivery, and thus maximize efficacy and reduce toxic side effects.

[0003] Traditional methods typically employ simple thresholding or random initialization of cluster centers to segment fluorescence images, and directly count drug-loaded cells based on the number of connected components. These methods have several drawbacks: First, random initialization of cluster centers can easily cause the segmentation algorithm to get stuck in local optima, resulting in unstable and inaccurate results. Second, when using Euclidean distance for pixel classification, the correlation between image color channels is ignored, leading to decreased segmentation accuracy when the contrast between fluorescence and background is low or when there is color interference. Finally, directly counting connected components ignores cell overlap or aggregation, causing errors in the estimation of the number of drug-loaded cells. Summary of the Invention

[0004] This invention provides a method for detecting the drug loading efficiency of targeted molecules based on fluorescence imaging and a computer-readable storage medium. Its main purpose is to improve the accuracy of fluorescence drug loading image segmentation and enhance the overall reliability of drug loading efficiency detection.

[0005] To achieve the above objectives, the present invention provides a method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging, comprising: Receive drug loading efficiency detection information, and determine the cells to be tested and the drug to be tested based on the drug loading efficiency detection information; Based on the cells to be tested, culture the cells in the culture dish and construct the targeted nanocarrier according to the drug to be tested and the pre-set fluorescent molecular label; The targeted nanocarrier and the cell culture dish to be tested were incubated to obtain an incubation culture dish. The incubation culture dish was then subjected to fluorescence excitation and image capture to obtain the original fluorescence image. The original fluorescence image is filtered to obtain a filtered fluorescence image. Histogram extraction is performed on the filtered fluorescence image to obtain a grayscale histogram. Clustering was performed based on grayscale histograms and filtered fluorescence images to obtain initial cluster groups; The filtered fluorescence image is clustered and segmented using the initial cluster group to obtain a fluorescence segmentation image; Drug loading efficiency is detected based on fluorescence segmentation images to obtain drug-loaded cell rate, and the drug loading efficiency of targeted molecules is detected based on the drug-loaded cell rate using fluorescence imaging.

[0006] Optionally, the step of dividing the data into clusters based on grayscale histograms and filtered fluorescence images to obtain initial cluster groups includes: Obtain the original pixel value set in the filtered fluorescence image, wherein the original pixel value set includes multiple original pixel values; The distribution statistics of each original pixel value in the original pixel value set are obtained by using grayscale histograms; Curve fitting is performed based on the original pixel value set and the pixel distribution quantity set to obtain the grayscale distribution curve, where the horizontal axis of the grayscale distribution curve represents the original pixel value and the vertical axis of the grayscale distribution curve represents the pixel distribution quantity. Cluster center search is performed using grayscale distribution curves to obtain an initial cluster center vector set, which includes two initial cluster center vectors. Initial cluster groups are generated based on the initial cluster center vector group.

[0007] Optionally, the step of using the grayscale distribution curve to search for cluster centers and obtain an initial set of cluster center vectors includes: Identify groups of gray-level maxima and minima in a gray-level distribution curve; Perform the following operation on each grayscale maximum in the grayscale maximum group: In the gray-level minimum point group, gray-level minimum points are extracted sequentially. Based on the gray-level maximum points and the extracted gray-level minimum points, the gray-level contrast is calculated. The gray-level contrast corresponding to each gray-level minimum point is summarized to obtain the gray-level contrast group. Merge the gray-level contrast groups corresponding to each gray-level maximum point to obtain the gray-level contrast set; Brightness is identified in the filtered fluorescence image to obtain the current image brightness value; The optimal contrast is obtained by performing a high contrast query in the grayscale contrast set based on the current image brightness value; Determine the optimal maximum and minimum points corresponding to the optimal contrast. The initial cluster center vector set is generated based on the optimal maxima and minimum points.

[0008] Optionally, the step of performing a high contrast query in the grayscale contrast set based on the current image brightness value to obtain the optimal contrast includes: Set multiple test image brightness values; Contrast testing was performed based on the brightness values ​​of multiple test images to obtain multiple contrast structure data. Each contrast structure data includes: high-quality contrast and the brightness value of the test image. Based on the current image brightness value, similar contrast structure data is obtained by performing a similarity index on multiple contrast structure data. Identify the optimal contrast in a grayscale contrast set based on high-quality contrast in similar contrast structure data.

[0009] Optionally, the contrast test is performed based on the brightness values ​​of multiple test images to obtain multiple contrast structure data, including: The brightness values ​​of the test images are extracted sequentially from multiple test image brightness values, and the test fluorescence image is obtained based on the extracted test image brightness values; Identify multiple test grayscale contrasts in the test fluorescence image, where each test grayscale contrast corresponds to a test maximum and a test minimum. The test fluorescence image is segmented based on multiple test gray-level contrasts to obtain multiple test segmentation images, where each test segmentation image corresponds one-to-one with a test gray-level contrast. Segmentation quality analysis was performed on multiple test segmentation images to obtain multiple test segmentation quality results; Identify the highest segmentation quality among multiple test segmentation qualities, and record the test grayscale contrast corresponding to the highest segmentation quality as the high-quality contrast. By merging the high-quality contrast ratio and the extracted brightness values ​​of the test image, contrast structure data is obtained. By summing the contrast structure data corresponding to the brightness value of each test image, multiple contrast structure data are obtained.

[0010] Optionally, generating the initial cluster center vector set based on the optimal maxima and minimum points includes: In the grayscale distribution curve, determine the optimal grayscale value corresponding to the optimal maximum point, where the optimal grayscale value is the x-coordinate of the optimal maximum point; Identify a set of pixels with the same value based on the optimal grayscale value, wherein the set of pixels with the same value includes multiple pixels with the same value; In the filtered fluorescence image, obtain the pixel RGB channel vector of each pixel in the set of pixels with the same value to obtain the pixel RGB channel vector set; The pixel RGB channel vector set is vector-averaged to obtain the maximal RGB channel vector; Generate the minimum RGB channel vector based on the optimal minimum point; Merge the maximal RGB channel vectors and the minimum RGB channel vectors to obtain the initial cluster center vector set.

[0011] Optionally, the step of clustering and segmenting the filtered fluorescence image using the initial cluster group to obtain a fluorescence segmentation image includes: The current cluster center vector group is set according to the initial cluster group, wherein the current cluster center vector in the current cluster center vector group corresponds one-to-one with the initial cluster; The set of filtered pixels in the filtered fluorescence image is identified, wherein the set of filtered pixels includes multiple filtered pixels; Perform the following operation on each filtered pixel in the set of filtered pixels: Obtain the filtered color vector of the filtered pixel; The distance to the filtered color vector is calculated using each current cluster center vector in the current cluster center vector group to obtain the pixel center distance group; The filtered pixels are assigned based on the pixel center distance group and the initial cluster group to obtain the updated cluster group; The updated cluster group is used as the initial cluster group, and the step of setting the current cluster center vector group based on the initial cluster group is returned until every filter pixel in the filter pixel set has been assigned. The updated cluster group when all filtered pixels in the set of filtered pixels have been assigned is denoted as the target cluster group; The current clustering error value is obtained by calculating the error based on the target cluster group. If the current clustering error value is greater than the preset standard error value, then the target cluster group is used as the initial cluster group, and the step of setting the current cluster center vector group based on the initial cluster group is returned until the current clustering error value is not greater than the standard error value. If the current clustering error value is not greater than the standard error value, then the target cluster group is recorded as the classification pixel cluster group; Identify and classify bright pixel clusters within a pixel cluster group, wherein a bright pixel cluster comprises multiple bright pixels; Fluorescence segmentation images are generated based on bright pixel clusters.

[0012] Optionally, the step of calculating the distance between the filtered color vector and each current cluster center vector in the current cluster center vector group to obtain a pixel center distance group includes: The filtered fluorescence image is converted to a color space to obtain a luminance space image, which is a Lab color image. A luminance covariance matrix is ​​generated based on the luminance spatial image, wherein the luminance covariance matrix is ​​a 3×3 matrix; Perform the following operation on each current cluster center vector in the current cluster center vector group: The pixel center distance is calculated based on the luminance covariance matrix, the current cluster center vector, and the filtered color vector, where the pixel center distance is expressed as:

[0013] in, Indicates the distance between pixel centers. Represents the filtered color vector. Represents the current cluster center vector. Let represent the luminance covariance matrix. Indicates the transpose symbol; Sum the pixel center distances corresponding to each current cluster center vector to obtain a pixel center distance group.

[0014] Optionally, the step of detecting drug loading efficiency based on the fluorescence segmentation image to obtain the drug-loaded cell rate includes: Connectivity analysis was performed on the fluorescence segmentation image to obtain a set of bright spot connected regions, which includes multiple bright spot connected regions. Obtain the size features of each bright spot connected region in the bright spot connected region set to obtain the size feature set; Bright spot statistics are performed on the bright spot connected region set using preset unit size features and size feature set to obtain the number of drug-loaded cells; The drug-loaded cell rate is calculated based on the number of drug-loaded cells and the preset total number of cells.

[0015] To achieve the above objectives, the present invention also provides a targeted molecular drug delivery efficiency detection system based on fluorescence imaging, comprising: The nanocarrier preparation module is used to receive drug loading efficiency detection information, determine the cells to be tested and the drugs to be tested based on the drug loading efficiency detection information, culture the cells to be tested in the cell culture dish based on the cells to be tested, and construct targeted nanocarriers according to the drugs to be tested and preset fluorescent molecular markers. The fluorescence image acquisition module is used to incubate the targeted nanocarrier and the cell culture dish to be detected to obtain the incubation culture dish. The incubation culture dish is sequentially subjected to fluorescence excitation and image capture to obtain the original fluorescence image. The original fluorescence image is filtered to obtain the filtered fluorescence image. Histogram extraction is performed based on the filtered fluorescence image to obtain the grayscale histogram. The fluorescence image segmentation module is used to perform clustering based on grayscale histograms and filtered fluorescence images to obtain initial cluster groups. The filtered fluorescence image is then segmented using the initial cluster groups to obtain a fluorescence segmentation image. The drug loading efficiency calculation module is used to detect drug loading efficiency based on fluorescence segmentation images and obtain the drug-loaded cell rate.

[0016] To address the above problems, the present invention also provides an electronic device, the electronic device comprising: Memory, storing at least one instruction; The processor executes the instructions stored in the memory to implement the above-described method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging.

[0017] To address the aforementioned problems, the present invention also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor in an electronic device to implement the aforementioned method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging.

[0018] To address the problems described in the background art, this invention involves culturing cells in a culture dish and constructing a targeted nanocarrier based on the target drug and a pre-defined fluorescent molecular marker. The surface of this targeted nanocarrier is modified with a targeting molecule that specifically recognizes the target cells. Compared to traditional non-targeted or passively targeted carriers, this improves the efficiency of targeted drug delivery to the target cells, thereby enhancing the specificity of subsequent fluorescence imaging signals. Next, clustering is performed based on the gray-level histogram and filtered fluorescence images to obtain initial cluster groups. This step analyzes the peaks and troughs of the gray-level histogram curve and dynamically queries the optimal contrast based on the current image brightness to determine the two initial cluster centers that best represent the background and fluorescence regions. Compared to the method of randomly initializing cluster centers in existing technologies, this method allows subsequent clustering algorithms to iterate from a starting point closer to the true classification, effectively avoiding getting trapped in local optima and accelerating the process. To improve convergence speed and enhance the accuracy and stability of final image segmentation, this invention further utilizes initial cluster groups to perform clustering segmentation on filtered fluorescence images, resulting in fluorescence segmentation images. In this step, when calculating the distance between pixels and cluster centers during cluster iteration, Mahalanobis distance based on the brightness space covariance matrix is ​​used instead of the traditional Euclidean distance. Compared to existing clustering segmentation methods, this improvement enhances segmentation accuracy in cases where the contrast between fluorescence and background colors is not significant or where color interference exists, making the extraction of fluorescence regions more accurate. Finally, during cell counting, instead of simply counting each fluorescent bright spot connected region as a cell, the size characteristics of each connected region are analyzed and compared with known unit cell sizes for approximate estimation. This allows for handling cases of cell overlap or aggregation. Compared to existing counting methods that simply map connected regions to cells one-to-one, this method can more realistically and accurately estimate the actual number of labeled drug-loaded cells. Therefore, this invention can improve the accuracy of fluorescent drug-loaded image segmentation and enhance the overall reliability of drug loading efficiency detection. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating a method for detecting the drug delivery efficiency of targeted molecules based on fluorescence imaging, provided in an embodiment of the present invention. Figure 2This is a functional block diagram of a targeted molecular drug delivery efficiency detection system based on fluorescence imaging provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of an electronic device for implementing the fluorescence imaging-based targeted molecule drug delivery efficiency detection method according to an embodiment of the present invention.

[0020] Explanation of reference numerals in the attached figures: 10. Electronic device; 11. Processor; 12. Memory; 13. Bus.

[0021] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0023] This application provides a method for detecting the drug delivery efficiency of targeted molecules based on fluorescence imaging. The execution entity of this method includes, but is not limited to, at least one electronic device that can be configured to execute the method provided in this application, such as a server or a terminal. In other words, the method can be executed by software or hardware installed on a terminal device or a server device, and the software may be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0024] Reference Figure 1 The diagram shown is a flowchart illustrating a method for detecting the drug loading efficiency of targeted molecules based on fluorescence imaging, according to an embodiment of the present invention. In this embodiment, the method for detecting the drug loading efficiency of targeted molecules based on fluorescence imaging includes: S1. Receive drug loading efficiency detection information, and determine the cells to be detected and the drugs to be detected based on the drug loading efficiency detection information.

[0025] Understandably, the drug loading efficiency detection information refers to relevant information that needs to be tested for drug loading efficiency. This drug loading efficiency detection information includes the cell to be tested and the drug to be tested. The cell to be tested refers to the target cell for which drug delivery and efficiency evaluation are required, and the drug to be tested refers to the drug molecule that needs to be encapsulated in a targeted nanocarrier and delivered into the cell. For example, the cell to be tested is a tumor cell line, and the drug to be tested is a certain chemotherapy drug.

[0026] S2. Based on the cell culture dish to be tested, a targeted nanocarrier is constructed according to the drug to be tested and the preset fluorescent molecular label.

[0027] It is understood that the cell culture dish to be tested refers to a special vessel used to hold and culture the cells to be tested for subsequent drug incubation and fluorescence imaging, such as a multi-well plate or culture dish. The specific method of culturing the cells to be tested in the culture dish is as follows: the cells to be tested are revived and passaged, then the revived and passaged cells are seeded into a sterile culture dish, and the culture dish is cultured in an incubator at 37°C and 5% CO2 until the cells grow to the required state (such as logarithmic growth phase or reaching a certain degree of confluence) for experimentation. At this point, the culture dish is the cell culture dish to be tested. The fluorescent molecular label refers to a chemical group or biomolecule that can emit fluorescence at a specific wavelength. This fluorescent molecular label is used to label the drug to be tested so that the drug can be imaged by fluorescence excitation in subsequent steps. Examples include fluorescent dyes (such as FITC, Cy3, Cy5) or fluorescent proteins (such as GFP). The targeted nanocarrier refers to a nanoscale particle system capable of loading a drug to be detected. The surface of the targeted nanocarrier is modified with targeting molecules that can specifically recognize and act on the cells to be detected, thereby achieving targeted delivery of the drug. The construction method of the targeted nanocarrier is as follows: First, a nanomaterial (such as liposomes, polymer micelles, inorganic nanoparticles, etc.) is selected as the carrier. Then, the drug to be detected is loaded into the interior of the nanomaterial. Fluorescent molecular labels are used to label the nanocarrier after loading the drug, that is, molecules (such as antibodies, peptides, aptamers, etc.) that can specifically recognize the surface markers of the cells to be detected are attached to the surface of the nanocarrier.

[0028] For example, human breast cancer cells MCF-7 were seeded in 35 mm confocal culture dishes and incubated at 37°C with 5% CO2 for 24 hours to allow the cells to adhere and enter the logarithmic growth phase, thus obtaining the cell culture dish to be tested. Liposomes loaded with the chemotherapeutic drug doxorubicin, i.e., targeted nanocarriers, were prepared using a thin-film hydration method. The fluorescent dye Cy5 was labeled (i.e., fluorescent molecular labeling) onto the phospholipids of these liposomes. Finally, an antibody targeting the HER2 receptor on the surface of MCF-7 cells was attached to the surface of the labeled liposomes, thereby constructing a Cy5-labeled HER2-targeting doxorubicin-loaded liposome nanocarrier.

[0029] S3. Incubate the targeted nanocarrier and the cell culture dish to be tested to obtain an incubation culture dish. Perform fluorescence excitation and image capture on the incubation culture dish in sequence to obtain the original fluorescence image.

[0030] It is clear that the incubation dish refers to the culture dish of the cells to be tested after incubation. Incubation of the targeted nanocarrier and the culture dish of the cells to be tested involves adding the targeted nanocarrier to the culture dish of the cells to be tested and co-culturing them for a period of time under conditions suitable for cell survival (e.g., 37°C, 5% CO2) to allow the nanocarrier to fully contact, bind, and be internalized by the cells. Fluorescence excitation refers to the process of using a targeted nanocarrier with a specific wavelength (matching the absorption spectrum of the fluorescent molecular label) to cause the labeled fluorescent molecules on the targeted nanocarrier to transition from the ground state to an excited state, and subsequently emit longer-wavelength fluorescence upon returning to the ground state. Image capture refers to microscopic imaging of the incubation dish after fluorescence excitation. The original fluorescence image refers to the fluorescence distribution image of the incubation dish obtained after image capture.

[0031] S4. Filter the original fluorescence image to obtain a filtered fluorescence image. Extract histograms from the filtered fluorescence image to obtain a grayscale histogram.

[0032] Understandably, the filtered fluorescence image refers to the original fluorescence image after filtering, where Gaussian filtering, median filtering, mean filtering, etc., can be used for this filtering operation. The grayscale histogram refers to a grayscale statistical graph, where the horizontal axis represents the grayscale value of a pixel in the filtered fluorescence image after grayscale conversion, and the vertical axis represents the number of pixels with that grayscale value. Histogram extraction is existing technology and will not be elaborated upon here.

[0033] S5. Based on the grayscale histogram and filtered fluorescence image, clustering is performed to obtain the initial cluster groups.

[0034] It is clear that the initial cluster group refers to the set of two initial clusters obtained after cluster division.

[0035] In detail, the clustering based on grayscale histograms and filtered fluorescence images to obtain initial cluster groups includes: Obtain the original pixel value set in the filtered fluorescence image, wherein the original pixel value set includes multiple original pixel values; The distribution statistics of each original pixel value in the original pixel value set are obtained by using grayscale histograms; Curve fitting is performed based on the original pixel value set and the pixel distribution quantity set to obtain the grayscale distribution curve, where the horizontal axis of the grayscale distribution curve represents the original pixel value and the vertical axis of the grayscale distribution curve represents the pixel distribution quantity. Cluster center search is performed using grayscale distribution curves to obtain an initial cluster center vector set, which includes two initial cluster center vectors. Initial cluster groups are generated based on the initial cluster center vector group.

[0036] It should be explained that the original pixel value set refers to a collection of multiple original pixel values, where each original pixel value refers to a pixel in the filtered fluorescence image. The pixel distribution data set refers to the set of the number of filtered pixels corresponding to each original pixel value in the original pixel value set. The grayscale distribution curve refers to the relationship curve between the original pixel values ​​and the number of pixel distributions, where Gaussian fitting, polynomial fitting, or other methods can be used for the curve fitting steps described above. The initial cluster center vector set refers to a set of two initial cluster center vectors, where each initial cluster center vector refers to the RGB channel value vector corresponding to the initial cluster center during subsequent clustering operations. The above-mentioned generation of initial cluster groups based on the initial cluster center vector set means that an empty set is allocated to each initial cluster center vector in the initial cluster center vector set, and this empty set is the initial cluster. All the initial clusters are combined into an initial cluster group.

[0037] In detail, the step of using grayscale distribution curves to search for cluster centers and obtain an initial set of cluster center vectors includes: Identify groups of gray-level maxima and minima in a gray-level distribution curve; Perform the following operation on each grayscale maximum in the grayscale maximum group: In the gray-level minimum point group, gray-level minimum points are extracted sequentially. Based on the gray-level maximum points and the extracted gray-level minimum points, the gray-level contrast is calculated. The gray-level contrast corresponding to each gray-level minimum point is summarized to obtain the gray-level contrast group. Merge the gray-level contrast groups corresponding to each gray-level maximum point to obtain the gray-level contrast set; Brightness is identified in the filtered fluorescence image to obtain the current image brightness value; The optimal contrast is obtained by performing a high contrast query in the grayscale contrast set based on the current image brightness value; Determine the optimal maximum and minimum points corresponding to the optimal contrast. The initial cluster center vector set is generated based on the optimal maxima and minimum points.

[0038] It should be explained that the gray-level maxima group refers to the set of multiple coordinate points at the peaks of the gray-level distribution curve. The gray-level minima group refers to the set of multiple coordinate points at the troughs of the gray-level distribution curve. Here, the gray-level maxima correspond to gray-level regions in the filtered fluorescence image where the pixel distribution is relatively concentrated, such as the background region (the dark area with the most pixels) and the fluorescent region where the fluorescently labeled cells to be detected are located (the bright area with a large number of pixels). The gray-level minima correspond to the transition regions between different regions, such as the transition region between the background region and the fluorescent region. Therefore, these gray-level maxima and minima can represent the typical gray-level values ​​of different regions (such as the background region and the fluorescent region). These gray-level maxima and minima can be selected as the basis for dividing the initial cluster center vector group, so that the subsequent clustering algorithm can start iterating from a starting point that is closer to the true classification, thereby accelerating the convergence speed and improving the accuracy and stability of the final clustering segmentation.

[0039] Furthermore, grayscale contrast refers to the numerical value that quantifies the separability between the regions containing grayscale maxima and minima. A higher grayscale contrast indicates a more distinct grayscale feature between the regions containing grayscale maxima and minima, making them easier to segment accurately by clustering algorithms, thus resulting in higher efficiency and accuracy in subsequent clustering. The grayscale contrast is calculated as follows: Where α represents grayscale contrast. The x-coordinate of the gray-scale maximum point (i.e., the gray value of the gray-scale maximum point in the filtered fluorescence image after gray-scale conversion). The x-coordinate represents the minimum grayscale point. Since different image brightness values ​​affect the significance and location of peaks and troughs in the grayscale distribution curve, thus influencing the effectiveness of different grayscale contrasts in actual clustering and segmentation, it is necessary to determine the optimal contrast in a pre-established lookup table based on the current image brightness value of the filtered fluorescence image. The current image brightness value is a quantitative indicator of the overall brightness of the filtered fluorescence image, calculated as the average grayscale value of all pixels in the filtered fluorescence image. The optimal contrast ratio refers to the grayscale contrast ratio that produces the highest segmentation quality under the current brightness conditions, obtained by querying a pre-established "image brightness-optimal contrast" correspondence (i.e., multiple contrast structure data) based on the current image brightness value. The optimal maximum point refers to the grayscale maximum point corresponding to the optimal contrast ratio, and the optimal minimum point refers to the grayscale minimum point corresponding to the optimal contrast ratio.

[0040] It is clear that in cluster analysis, if the initial cluster centers are located within the same region or close to the region boundary, it is easy to cause slow iteration convergence or get stuck in local optima. The gray-level maxima and minima corresponding to the above-mentioned optimal contrast represent the gray-level feature combinations that have the greatest separation degree in different regions of the filtered fluorescence image under the current image brightness. Therefore, it is necessary to generate an initial cluster center vector group based on the two.

[0041] In detail, the step of performing a high contrast query in the grayscale contrast set based on the current image brightness value to obtain the optimal contrast includes: Set multiple test image brightness values; Contrast testing was performed based on the brightness values ​​of multiple test images to obtain multiple contrast structure data. Each contrast structure data includes: high-quality contrast and the brightness value of the test image. Based on the current image brightness value, similar contrast structure data is obtained by performing a similarity index on multiple contrast structure data. Identify the optimal contrast in a grayscale contrast set based on high-quality contrast in similar contrast structure data.

[0042] It should be explained that the test image brightness value refers to a manually set image brightness value, which can be set according to the current image brightness value that may be encountered during the actual experiment. The contrast structure data refers to the data set of high-quality contrast and test image brightness values. The similar contrast structure data refers to the contrast structure data obtained after similarity indexing. The similarity indexing method is as follows: identify the test image brightness value with the smallest distance from the current image brightness value among multiple contrast structure data, and record the contrast structure data corresponding to the test image brightness value with the smallest distance as similar contrast structure data. The above-mentioned optimal contrast is identified by: identifying the grayscale contrast closest to the high-quality contrast in the grayscale contrast set, and taking this grayscale contrast as the optimal contrast.

[0043] Specifically, the contrast test is performed based on the brightness values ​​of multiple test images to obtain multiple contrast structure data, including: The brightness values ​​of the test images are extracted sequentially from multiple test image brightness values, and the test fluorescence image is obtained based on the extracted test image brightness values; Identify multiple test grayscale contrasts in the test fluorescence image, where each test grayscale contrast corresponds to a test maximum and a test minimum. The test fluorescence image is segmented based on multiple test gray-level contrasts to obtain multiple test segmentation images, where each test segmentation image corresponds one-to-one with a test gray-level contrast. Segmentation quality analysis was performed on multiple test segmentation images to obtain multiple test segmentation quality results; Identify the highest segmentation quality among multiple test segmentation qualities, and record the test grayscale contrast corresponding to the highest segmentation quality as the high-quality contrast. By merging the high-quality contrast ratio and the extracted brightness values ​​of the test image, contrast structure data is obtained. By summing the contrast structure data corresponding to the brightness value of each test image, multiple contrast structure data are obtained.

[0044] It should be explained that the test fluorescence image refers to a fluorescence image whose image brightness value is the test image brightness value. The acquisition method of the test fluorescence image is the same as that of the original fluorescence image described above. The test grayscale contrast refers to the grayscale contrast exhibited by the test fluorescence image. The identification method of the test grayscale contrast is the same as that of the grayscale contrast described above. The test maxima and test minima refer to the grayscale maxima and grayscale minima corresponding to the test grayscale contrast, respectively. The test segmentation image refers to the fluorescence segmentation image obtained after clustering and segmenting the test fluorescence image according to a certain test grayscale contrast. The segmentation test described above is the same as the subsequent clustering and segmentation of the original fluorescence image to obtain the fluorescence segmentation image, and will not be repeated here. The test segmentation quality refers to the numerical value that quantifies the quality of the final clustering and segmentation result after setting the initial cluster centers according to a certain test grayscale contrast. The higher the test segmentation quality, the better the image segmentation effect obtained after using the test maxima and test minima corresponding to the test contrast for clustering and segmentation. For example, the separation of the fluorescence region and the background region is clearer and more accurate. Optionally, the test segmentation quality is calculated as follows: the ratio of the average gray level of the segmented fluorescent region (i.e., the test segmentation image) to the average gray level of the background region (i.e., the region of the test fluorescent image other than the test segmentation image). The maximum segmentation quality refers to the test segmentation quality with the largest value among multiple test segmentation qualities.

[0045] Specifically, the generation of the initial cluster center vector set based on the optimal maxima and minimum points includes: In the grayscale distribution curve, determine the optimal grayscale value corresponding to the optimal maximum point, where the optimal grayscale value is the x-coordinate of the optimal maximum point; Identify a set of pixels with the same value based on the optimal grayscale value, wherein the set of pixels with the same value includes multiple pixels with the same value; In the filtered fluorescence image, obtain the pixel RGB channel vector of each pixel in the set of pixels with the same value to obtain the pixel RGB channel vector set; The pixel RGB channel vector set is vector-averaged to obtain the maximal RGB channel vector; Generate the minimum RGB channel vector based on the optimal minimum point; Merge the maximal RGB channel vectors and the minimum RGB channel vectors to obtain the initial cluster center vector set.

[0046] It should be explained that the set of pixels with the same value refers to the set of all pixels in the filtered fluorescence image after grayscale conversion whose grayscale value is the same as the optimal grayscale value (i.e., pixels with the same value). The pixel RGB channel vector refers to the numerical vector composed of the RGB three-channel values ​​at the location of a pixel with the same value in the filtered fluorescence image. The maximal RGB channel vector refers to the RGB channel vector obtained after vector averaging, where vector averaging refers to calculating the average value of the values ​​at the same vector positions in the pixel RGB channel vector set to obtain the average RGB channel value at each vector position. These average RGB channel values ​​constitute the maximal RGB channel vector. The minimum RGB channel value refers to the RGB channel vector obtained after performing the same processing as obtaining the maximal RGB channel vector.

[0047] S6. Use the initial cluster group to perform clustering and segmentation on the filtered fluorescence image to obtain a fluorescence segmentation image.

[0048] Understandably, the fluorescence segmentation image refers to the image representing the fluorescence region obtained after clustering segmentation.

[0049] In detail, the step of clustering and segmenting the filtered fluorescence image using an initial cluster group to obtain a fluorescence segmentation image includes: The current cluster center vector group is set according to the initial cluster group, wherein the current cluster center vector in the current cluster center vector group corresponds one-to-one with the initial cluster; The set of filtered pixels in the filtered fluorescence image is identified, wherein the set of filtered pixels includes multiple filtered pixels; Perform the following operation on each filtered pixel in the set of filtered pixels: Obtain the filtered color vector of the filtered pixel; The distance to the filtered color vector is calculated using each current cluster center vector in the current cluster center vector group to obtain the pixel center distance group; The filtered pixels are assigned based on the pixel center distance group and the initial cluster group to obtain the updated cluster group; The updated cluster group is used as the initial cluster group, and the step of setting the current cluster center vector group based on the initial cluster group is returned until every filter pixel in the filter pixel set has been assigned. The updated cluster group when all filtered pixels in the set of filtered pixels have been assigned is denoted as the target cluster group; The current clustering error value is obtained by calculating the error based on the target cluster group. If the current clustering error value is greater than the preset standard error value, then the target cluster group is used as the initial cluster group, and the step of setting the current cluster center vector group based on the initial cluster group is returned until the current clustering error value is not greater than the standard error value. If the current clustering error value is not greater than the standard error value, then the target cluster group is recorded as the classification pixel cluster group; Identify and classify bright pixel clusters within a pixel cluster group, wherein a bright pixel cluster comprises multiple bright pixels; Fluorescence segmentation images are generated based on bright pixel clusters.

[0050] It should be explained that the current cluster center vector set refers to the set of two current cluster center vectors (each corresponding to an initial cluster). The current cluster center vector refers to the center vector of the initial cluster. The current cluster center vector is obtained by identifying the RGB channel vectors of all pixels in the initial cluster and then averaging all the RGB channel vectors. The resulting vector is the current cluster center vector. In the initial state, this current cluster center vector is the initial cluster center vector. The filtered pixel set refers to a set of multiple filtered pixels, where filtered pixels refer to pixels in the filtered fluorescence image. The filtered color vector refers to the RGB channel vector of the filtered pixel. The pixel center distance set refers to a set of multiple pixel center distances, where the pixel center distance refers to the distance between a certain current cluster center vector and the filtered color vector. The smaller the pixel center distance, the more likely the filtered pixel corresponding to the filtered color vector is to be assigned to the initial cluster corresponding to the current cluster center vector. The updated cluster group refers to the initial cluster group after allocation. The specific method for allocating the filtered pixels based on the pixel center distance group and the initial cluster group is as follows: identify the pixel center distance with the smaller value in the pixel center distance group, and add the filtered pixel to the initial cluster corresponding to the pixel center distance with the smaller value. The initial cluster after the addition and the initial cluster without the addition form the updated cluster group. The current clustering error value refers to the numerical value that quantifies the degree of difference between the target cluster group obtained after the current iteration and the target cluster group obtained after the previous iteration (denoted as the historical cluster group). The current clustering error value is calculated as follows: calculate the target cluster center vector of each target cluster in the target cluster group to obtain the target cluster center vector group. The calculation method of the target cluster center vector is the same as the above steps of setting the current cluster center vector group based on the initial cluster group. Calculate the historical cluster center vector group of the historical cluster group. Calculate the vector Euclidean distance between each target cluster center vector in the target cluster center vector group and the corresponding historical cluster center vector in the historical cluster center vector group to obtain two vector Euclidean distances. Add the two vector Euclidean distances to obtain the current clustering error value.

[0051] Furthermore, the standard error value refers to a manually set threshold used to determine whether the clustering algorithm has converged. If the current clustering error value is greater than this standard error value, it indicates that the cluster centers are not yet stable and the clustering result has not reached the optimal level. The standard error value is set by using experience or experimentation to set a small positive number, such as 0.001 or 0.01, based on the accuracy requirements of the actual application. The bright pixel cluster refers to the cluster of classifying pixels with a smaller average gray value in the classification pixel cluster group. The bright pixel refers to the pixel in the bright pixel cluster. The above-mentioned generation of a fluorescence segmentation image based on bright pixel clusters refers to: creating a new image with the same size as the original fluorescence image but containing only binary information (including high brightness values ​​and low brightness values, where high brightness values ​​represent fluorescent areas and low brightness values ​​represent background areas); setting the corresponding positions of all bright pixels belonging to the bright pixel clusters in the new image to high brightness values, while setting all other pixels to low brightness values, thereby generating a binary segmentation image that only highlights the fluorescent areas.

[0052] In detail, the step of calculating the distance between the filtered color vector and each current cluster center vector in the current cluster center vector group to obtain a pixel center distance group includes: The filtered fluorescence image is converted to a color space to obtain a luminance space image, which is a Lab color image. A luminance covariance matrix is ​​generated based on the luminance spatial image, wherein the luminance covariance matrix is ​​a 3×3 matrix; Perform the following operation on each current cluster center vector in the current cluster center vector group: The pixel center distance is calculated based on the luminance covariance matrix, the current cluster center vector, and the filtered color vector, where the pixel center distance is expressed as:

[0053] in, Indicates the distance between pixel centers. Represents the filtered color vector. Represents the brightness covariance matrix. Indicates the transpose symbol; Sum the pixel center distances corresponding to each current cluster center vector to obtain a pixel center distance group.

[0054] It should be explained that the luminance space image refers to a filtered fluorescence image converted to the Lab color space. The luminance covariance matrix is ​​a 3x3 matrix that describes the distribution characteristics of the dataset consisting of all pixels in the luminance space image (Lab image) across its three color channels (L channel, a channel, and b channel). The diagonal elements of this luminance covariance matrix are the variances of each color channel itself, and the off-diagonal elements are the covariances between each pair of different color channels. The construction method of this luminance covariance matrix is ​​existing technology and will not be described further here. This scheme uses Mahalanobis distance (the same method used to calculate pixel center distance) instead of the traditional Euclidean distance to calculate pixel center distance. The reason is that Mahalanobis distance takes into account the correlation and scale differences between data features. In a brightness space image, the color information of a pixel is composed of multiple channels (such as L, a, and b in Lab space), and the numerical distribution range of each channel may be different. By introducing the brightness covariance matrix, Mahalanobis distance makes the calculated pixel center distance equivalent to the Euclidean distance in the space after removing the correlation of each dimension and standardizing the variance. In this way, the pixel center distance can more accurately reflect the relative position of the pixel in the filtered fluorescence image, thereby improving the accuracy of pixel segmentation in complex color distribution backgrounds.

[0055] S6. Detect drug loading efficiency based on fluorescence segmentation images to obtain drug-loaded cell rate, and complete the detection of targeted molecule drug loading efficiency based on fluorescence imaging based on drug-loaded cell rate.

[0056] It is clear that the drug-loaded cell rate refers to a numerical value that quantifies drug loading efficiency. The higher the drug-loaded cell rate, the higher the drug loading efficiency.

[0057] Specifically, the step of detecting drug loading efficiency based on fluorescence segmentation images to obtain the drug-loaded cell rate includes: Connectivity analysis was performed on the fluorescence segmentation image to obtain a set of bright spot connected regions, which includes multiple bright spot connected regions. Obtain the size features of each bright spot connected region in the bright spot connected region set to obtain the size feature set; Bright spot statistics are performed on the bright spot connected region set using preset unit size features and size feature set to obtain the number of drug-loaded cells; The drug-loaded cell rate is calculated based on the number of drug-loaded cells and the preset total number of cells.

[0058] It should be explained that the set of bright spot connected regions refers to a collection of multiple bright spot connected regions. A bright spot connected region is an independent region in a fluorescence segmentation image formed by connecting adjacent bright pixels (in the sense of 8-connectivity or 4-connectivity). Since the labeled cells to be detected (i.e., drug-loaded cells) may partially overlap, one bright spot connected region corresponds to one or more labeled cells to be detected. The set of size features refers to a collection of multiple size features, where size features refer to numerical values ​​such as the radius, length, or width of the bounding matrix of the bright spot connected region. The unit size feature refers to the size feature of a single cell to be detected, which can be obtained by consulting the same literature. The number of drug-loaded cells refers to the total number of drug-loaded cells represented by the fluorescence segmentation image. The specific method for performing bright spot statistics on the bright spot connected region set using the preset unit size features and size feature set is as follows: Size features are extracted sequentially from the size feature set; the ratio of the size feature to the unit size feature is calculated; and this ratio is rounded to obtain an approximate ratio. If the approximate ratio equals an integer S, it indicates that the bright spot connected region corresponding to that size feature contains S drug-loaded cells. These approximate ratios are then summarized to obtain an approximate ratio set. The sum of these approximate ratio sets is then calculated, and the sum is the number of drug-loaded cells. The total number of cells refers to the total number of cells to be detected in the same culture dish, determined in advance by independent cell counting methods (such as trypan blue staining counting, nuclear fluorescence labeling counting, etc.). The drug-loaded cell rate can be expressed as the ratio of the number of drug-loaded cells to the total number of cells.

[0059] To address the problems described in the background art, this invention involves culturing cells in a culture dish and constructing a targeted nanocarrier based on the target drug and a pre-defined fluorescent molecular marker. The surface of this targeted nanocarrier is modified with a targeting molecule that specifically recognizes the target cells. Compared to traditional non-targeted or passively targeted carriers, this improves the efficiency of targeted drug delivery to the target cells, thereby enhancing the specificity of subsequent fluorescence imaging signals. Next, clustering is performed based on the gray-level histogram and filtered fluorescence images to obtain initial cluster groups. This step analyzes the peaks and troughs of the gray-level histogram curve and dynamically queries the optimal contrast based on the current image brightness to determine the two initial cluster centers that best represent the background and fluorescence regions. Compared to the method of randomly initializing cluster centers in existing technologies, this method allows subsequent clustering algorithms to iterate from a starting point closer to the true classification, effectively avoiding getting trapped in local optima and accelerating the process. To improve convergence speed and enhance the accuracy and stability of final image segmentation, this invention further utilizes initial cluster groups to perform clustering segmentation on filtered fluorescence images, resulting in fluorescence segmentation images. In this step, when calculating the distance between pixels and cluster centers during cluster iteration, Mahalanobis distance based on the brightness space covariance matrix is ​​used instead of the traditional Euclidean distance. Compared to existing clustering segmentation methods, this improvement enhances segmentation accuracy in cases where the contrast between fluorescence and background colors is not significant or where color interference exists, making the extraction of fluorescence regions more accurate. Finally, during cell counting, instead of simply counting each fluorescent bright spot connected region as a cell, the size characteristics of each connected region are analyzed and compared with known unit cell sizes for approximate estimation. This allows for handling cases of cell overlap or aggregation. Compared to existing counting methods that simply map connected regions to cells one-to-one, this method can more realistically and accurately estimate the actual number of labeled drug-loaded cells. Therefore, this invention can improve the accuracy of fluorescent drug-loaded image segmentation and enhance the overall reliability of drug loading efficiency detection.

[0060] like Figure 2 The diagram shown is a functional block diagram of a targeted molecular drug delivery efficiency detection system based on fluorescence imaging provided in an embodiment of the present invention.

[0061] The fluorescence imaging-based targeted molecule drug loading efficiency detection system 100 of this invention can be installed in an electronic device. Depending on the functions implemented, the fluorescence imaging-based targeted molecule drug loading efficiency detection system 100 may include a nanocarrier preparation module 101, a fluorescence image acquisition module 102, a fluorescence image segmentation module 103, and a drug loading efficiency calculation module 104. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device. The nanocarrier preparation module 101 is used to receive drug loading efficiency detection information, determine the cells to be tested and the drugs to be tested based on the drug loading efficiency detection information, culture the cells to be tested in a culture dish based on the cells to be tested, and construct a targeted nanocarrier according to the drugs to be tested and preset fluorescent molecular markers. The fluorescence image acquisition module 102 is used to incubate the targeted nanocarrier and the cell culture dish to be detected to obtain an incubation culture dish, to perform fluorescence excitation and image capture on the incubation culture dish in sequence to obtain a raw fluorescence image, to perform a filtering operation on the raw fluorescence image to obtain a filtered fluorescence image, and to extract a histogram based on the filtered fluorescence image to obtain a grayscale histogram. The fluorescence image segmentation module 103 is used to perform clustering based on the grayscale histogram and the filtered fluorescence image to obtain an initial cluster group, and then use the initial cluster group to perform clustering segmentation on the filtered fluorescence image to obtain a fluorescence segmentation image. The drug loading efficiency calculation module 104 is used to detect drug loading efficiency based on the fluorescence segmentation image to obtain the drug-loaded cell rate.

[0062] In detail, the modules in the fluorescence imaging-based targeted molecule drug delivery efficiency detection system 100 described in this embodiment of the invention employ the same methods as described above. Figure 1 The method described herein is the same as the fluorescence imaging-based targeted molecule drug delivery efficiency detection method and can produce the same technical effect, so it will not be repeated here.

[0063] like Figure 3 The diagram shown is a schematic representation of an electronic device for implementing a method for detecting the drug delivery efficiency of targeted molecules based on fluorescence imaging, according to an embodiment of the present invention.

[0064] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a program for detecting the efficiency of targeted molecular drug delivery based on fluorescence imaging.

[0065] The memory 11 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as a portable hard drive. In other embodiments, the memory 11 can be an external storage device of the electronic device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device 1. Furthermore, the memory 11 includes both internal storage units and external storage devices of the electronic device 1. The memory 11 can be used not only to store application software and various types of data installed on the electronic device 1, such as the code of a targeted molecular drug delivery efficiency detection method program based on fluorescence imaging, but also to temporarily store data that has been output or will be output.

[0066] In some embodiments, the processor 10 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 10 is the control unit of the electronic device, connecting various components of the entire electronic device through various interfaces and lines. It executes programs or modules stored in the memory 11 (e.g., a program for detecting the efficiency of targeted molecular drug delivery based on fluorescence imaging) and calls data stored in the memory 11 to perform various functions of the electronic device 1 and process data.

[0067] The bus 12 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 can be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to realize the connection and communication between the memory 11 and at least one processor 10, etc.

[0068] Figure 3 Only electronic devices with components are shown; it will be understood by those skilled in the art that... Figure 3The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0069] For example, although not shown, the electronic device 1 may also include a power supply (such as a battery) to power the various components. Preferably, the power supply can be logically connected to the at least one processor 10 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0070] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the electronic device 1 and other electronic devices.

[0071] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), and optionally, a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the electronic device 1 and to display a visual user interface.

[0072] The program for detecting the targeted molecule drug loading efficiency based on fluorescence imaging, stored in the memory 11 of the electronic device 1, is a combination of multiple instructions. When run in the processor 10, it can achieve the following: Receive drug loading efficiency detection information, and determine the cells to be tested and the drug to be tested based on the drug loading efficiency detection information; Based on the cells to be tested, culture the cells in the culture dish and construct the targeted nanocarrier according to the drug to be tested and the pre-set fluorescent molecular label; The targeted nanocarrier and the cell culture dish to be tested were incubated to obtain an incubation culture dish. The incubation culture dish was then subjected to fluorescence excitation and image capture to obtain the original fluorescence image. The original fluorescence image is filtered to obtain a filtered fluorescence image. Histogram extraction is performed on the filtered fluorescence image to obtain a grayscale histogram. Clustering was performed based on grayscale histograms and filtered fluorescence images to obtain initial cluster groups; The filtered fluorescence image is clustered and segmented using the initial cluster group to obtain a fluorescence segmentation image; Drug loading efficiency is detected based on fluorescence segmentation images to obtain drug-loaded cell rate, and the drug loading efficiency of targeted molecules is detected based on the drug-loaded cell rate using fluorescence imaging.

[0073] Specifically, the processor 10's implementation method for the above instructions can be found in [reference needed]. Figures 1 to 3 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0074] Furthermore, if the modules / units integrated in the electronic device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, or a read-only memory (ROM).

[0075] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor of an electronic device, can perform the following: Receive drug loading efficiency detection information, and determine the cells to be tested and the drug to be tested based on the drug loading efficiency detection information; Based on the cells to be tested, culture the cells in the culture dish and construct the targeted nanocarrier according to the drug to be tested and the pre-set fluorescent molecular label; The targeted nanocarrier and the cell culture dish to be tested were incubated to obtain an incubation culture dish. The incubation culture dish was then subjected to fluorescence excitation and image capture to obtain the original fluorescence image. The original fluorescence image is filtered to obtain a filtered fluorescence image. Histogram extraction is performed on the filtered fluorescence image to obtain a grayscale histogram. Clustering was performed based on grayscale histograms and filtered fluorescence images to obtain initial cluster groups; The filtered fluorescence image is clustered and segmented using the initial cluster group to obtain a fluorescence segmentation image; Drug loading efficiency is detected based on fluorescence segmentation images to obtain drug-loaded cell rate, and the drug loading efficiency of targeted molecules is detected based on the drug-loaded cell rate using fluorescence imaging.

[0076] In the embodiments provided by this invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and actual implementations may have other classification methods.

[0077] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0078] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0079] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging, characterized in that, The method includes: Receive drug loading efficiency detection information, and determine the cells to be tested and the drug to be tested based on the drug loading efficiency detection information; Based on the cells to be tested, culture the cells in the culture dish and construct the targeted nanocarrier according to the drug to be tested and the pre-set fluorescent molecular label; The targeted nanocarrier and the cell culture dish to be tested were incubated to obtain an incubation culture dish. The incubation culture dish was then subjected to fluorescence excitation and image capture to obtain the original fluorescence image. The original fluorescence image is filtered to obtain a filtered fluorescence image. Histogram extraction is performed on the filtered fluorescence image to obtain a grayscale histogram. Clustering was performed based on grayscale histograms and filtered fluorescence images to obtain initial cluster groups; The filtered fluorescence image is clustered and segmented using the initial cluster group to obtain a fluorescence segmentation image; Drug loading efficiency is detected based on fluorescence segmentation images to obtain drug-loaded cell rate, and the drug loading efficiency of targeted molecules is detected based on the drug-loaded cell rate using fluorescence imaging.

2. The method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging as described in claim 1, characterized in that, The clustering based on grayscale histograms and filtered fluorescence images yields initial cluster groups, including: Obtain the original pixel value set in the filtered fluorescence image, wherein the original pixel value set includes multiple original pixel values; The distribution statistics of each original pixel value in the original pixel value set are obtained by using grayscale histograms; Curve fitting is performed based on the original pixel value set and the pixel distribution quantity set to obtain the grayscale distribution curve, where the horizontal axis of the grayscale distribution curve represents the original pixel value and the vertical axis of the grayscale distribution curve represents the pixel distribution quantity. Cluster center search is performed using grayscale distribution curves to obtain an initial cluster center vector set, which includes two initial cluster center vectors. Initial cluster groups are generated based on the initial cluster center vector group.

3. The method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging as described in claim 2, characterized in that, The method of using grayscale distribution curves to search for cluster centers to obtain an initial set of cluster center vectors includes: Identify groups of gray-level maxima and minima in a gray-level distribution curve; Perform the following operation on each grayscale maximum in the grayscale maximum group: In the gray-level minimum point group, gray-level minimum points are extracted sequentially. Based on the gray-level maximum points and the extracted gray-level minimum points, the gray-level contrast is calculated. The gray-level contrast corresponding to each gray-level minimum point is summarized to obtain the gray-level contrast group. Merge the gray-level contrast groups corresponding to each gray-level maximum point to obtain the gray-level contrast set; Brightness is identified in the filtered fluorescence image to obtain the current image brightness value; The optimal contrast is obtained by performing a high contrast query in the grayscale contrast set based on the current image brightness value; Determine the optimal maximum and minimum points corresponding to the optimal contrast. The initial cluster center vector set is generated based on the optimal maxima and minimum points.

4. The method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging as described in claim 3, characterized in that, The step of performing a high contrast query in the grayscale contrast set based on the current image brightness value to obtain the optimal contrast includes: Set multiple test image brightness values; Contrast testing was performed based on the brightness values ​​of multiple test images to obtain multiple contrast structure data. Each contrast structure data includes: high-quality contrast and the brightness value of the test image. Based on the current image brightness value, similar contrast structure data is obtained by performing a similarity index on multiple contrast structure data. Identify the optimal contrast in a grayscale contrast set based on high-quality contrast in similar contrast structure data.

5. The method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging as described in claim 4, characterized in that, The contrast test is performed based on the brightness values ​​of multiple test images to obtain multiple contrast structure data, including: The brightness values ​​of the test images are extracted sequentially from multiple test image brightness values, and the test fluorescence image is obtained based on the extracted test image brightness values; Identify multiple test grayscale contrasts in the test fluorescence image, where each test grayscale contrast corresponds to a test maximum and a test minimum. The test fluorescence image is segmented based on multiple test gray-level contrasts to obtain multiple test segmentation images, where each test segmentation image corresponds one-to-one with a test gray-level contrast. Segmentation quality analysis was performed on multiple test segmentation images to obtain multiple test segmentation quality results; Identify the highest segmentation quality among multiple test segmentation qualities, and record the test grayscale contrast corresponding to the highest segmentation quality as the high-quality contrast. By merging the high-quality contrast ratio and the extracted brightness values ​​of the test image, contrast structure data is obtained. By summing the contrast structure data corresponding to the brightness value of each test image, multiple contrast structure data are obtained.

6. The method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging as described in claim 5, characterized in that, The generation of the initial cluster center vector set based on the optimal maxima and minimum points includes: In the grayscale distribution curve, determine the optimal grayscale value corresponding to the optimal maximum point, where the optimal grayscale value is the x-coordinate of the optimal maximum point; Identify a set of pixels with the same value based on the optimal grayscale value, wherein the set of pixels with the same value includes multiple pixels with the same value; In the filtered fluorescence image, obtain the pixel RGB channel vector of each pixel in the set of pixels with the same value to obtain the pixel RGB channel vector set; The pixel RGB channel vector set is vector-averaged to obtain the maximal RGB channel vector; Generate the minimum RGB channel vector based on the optimal minimum point; Merge the maximal RGB channel vectors and the minimum RGB channel vectors to obtain the initial cluster center vector set.

7. The method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging as described in claim 6, characterized in that, The step of using initial cluster groups to cluster and segment the filtered fluorescence image to obtain a fluorescence segmentation image includes: The current cluster center vector group is set according to the initial cluster group, wherein the current cluster center vector in the current cluster center vector group corresponds one-to-one with the initial cluster; The set of filtered pixels in the filtered fluorescence image is identified, wherein the set of filtered pixels includes multiple filtered pixels; Perform the following operation on each filtered pixel in the set of filtered pixels: Obtain the filtered color vector of the filtered pixel; The distance to the filtered color vector is calculated using each current cluster center vector in the current cluster center vector group to obtain the pixel center distance group; The filtered pixels are assigned based on the pixel center distance group and the initial cluster group to obtain the updated cluster group; The updated cluster group is used as the initial cluster group, and the step of setting the current cluster center vector group based on the initial cluster group is returned until every filter pixel in the filter pixel set has been assigned. The updated cluster group when all filtered pixels in the set of filtered pixels have been assigned is denoted as the target cluster group; The current clustering error value is obtained by calculating the error based on the target cluster group. If the current clustering error value is greater than the preset standard error value, then the target cluster group is used as the initial cluster group, and the step of setting the current cluster center vector group based on the initial cluster group is returned until the current clustering error value is not greater than the standard error value. If the current clustering error value is not greater than the standard error value, then the target cluster group is recorded as the classification pixel cluster group; Identify and classify bright pixel clusters within a pixel cluster group, wherein a bright pixel cluster comprises multiple bright pixels; Fluorescence segmentation images are generated based on bright pixel clusters.

8. The method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging as described in claim 7, characterized in that, The step of calculating the distance between each current cluster center vector in the current cluster center vector group and the filtered color vector to obtain a pixel center distance group includes: The filtered fluorescence image is converted to a color space to obtain a luminance space image, which is a Lab color image. A luminance covariance matrix is ​​generated based on the luminance spatial image, wherein the luminance covariance matrix is ​​a 3×3 matrix; Perform the following operation on each current cluster center vector in the current cluster center vector group: The pixel center distance is calculated based on the luminance covariance matrix, the current cluster center vector, and the filtered color vector, where the pixel center distance is expressed as: in, Indicates the distance between pixel centers. Represents the filtered color vector. Represents the current cluster center vector. Let represent the luminance covariance matrix. Indicates the transpose symbol; Sum the pixel center distances corresponding to each current cluster center vector to obtain a pixel center distance group.

9. The method for detecting the targeted molecular drug delivery efficiency based on fluorescence imaging as described in claim 8, characterized in that, The step of detecting drug loading efficiency based on fluorescence segmentation images to obtain the drug-loaded cell rate includes: Connectivity analysis was performed on the fluorescence segmentation image to obtain a set of bright spot connected regions, which includes multiple bright spot connected regions. Obtain the size features of each bright spot connected region in the bright spot connected region set to obtain the size feature set; Bright spot statistics are performed on the bright spot connected region set using preset unit size features and size feature set to obtain the number of drug-loaded cells; The drug-loaded cell rate is calculated based on the number of drug-loaded cells and the preset total number of cells.

10. A targeted molecular drug delivery efficiency detection system based on fluorescence imaging, characterized in that, The system includes: The nanocarrier preparation module is used to receive drug loading efficiency detection information, determine the cells to be tested and the drugs to be tested based on the drug loading efficiency detection information, culture the cells to be tested in the cell culture dish based on the cells to be tested, and construct targeted nanocarriers according to the drugs to be tested and preset fluorescent molecular markers. The fluorescence image acquisition module is used to incubate the targeted nanocarrier and the cell culture dish to be detected to obtain the incubation culture dish. The incubation culture dish is sequentially subjected to fluorescence excitation and image capture to obtain the original fluorescence image. The original fluorescence image is filtered to obtain the filtered fluorescence image. Histogram extraction is performed based on the filtered fluorescence image to obtain the grayscale histogram. The fluorescence image segmentation module is used to perform clustering based on grayscale histograms and filtered fluorescence images to obtain initial cluster groups. The filtered fluorescence image is then segmented using the initial cluster groups to obtain a fluorescence segmentation image. The drug loading efficiency calculation module is used to detect drug loading efficiency based on fluorescence segmentation images and obtain the drug-loaded cell rate.

Images