Method and system for determining protein expression abundance based on gastrointestinal tissue

Abstract

The application provides a kind of based on digestive tract tissue protein expression abundance determination method and system, applied to digestive tract tissue image processing technical field. Including: based on the PVDF membrane image after color development of digestive tract tissue acquisition in vitro, contain at least one protein standard band and the protein band of multiple target proteins;When determining the presence of electrophoresis distortion according to the band of target protein standard, construct pixel space coordinate system;The target image is obtained by angle correction and linear repair to PVDF membrane image, and the protein band corresponding to multiple target proteins in target image is in horizontal alignment state under pixel space coordinate system;Determine the development area of each target protein in target image and the irregular region of interest wrapped the protein band corresponding to multiple target proteins, and for each target protein, the expression abundance of target protein is obtained by analyzing the development area and the irregular region of interest, and the accuracy of the protein expression abundance is higher.

Description

Technical Field

[0001] This application relates to the field of digestive tract tissue image processing technology, and in particular to a method and system for determining protein expression abundance based on digestive tract tissue. Background Technology

[0002] Currently, Western blotting is the core technique for quantitatively detecting the expression levels (i.e., abundance) of key proteins such as Ano1 and E-cadherin. However, when using Western blotting to perform quantitative protein analysis on digestive tract tissues, the complexity of these tissues and fluctuations in experimental conditions often result in protein image quality that is difficult to achieve ideal results.

[0003] On the one hand, polyvinylidene fluoride (PVDF) membranes are prone to generating unevenly distributed, nonspecific background signals and other interference signals during the color development stage. These interference signals are mixed with the characteristics of the target protein, making it difficult for conventional detection methods to accurately extract pure protein expression abundance information. On the other hand, the unstable physical environment during electrophoresis often causes nonlinear shifts in protein migration paths, resulting in irregular geometric shapes of the generated bands. Traditional regularized region extraction methods are prone to introducing interfering pixels from irrelevant regions when processing such deformed bands, leading to data bias. Furthermore, due to the lack of automated boundary recognition mechanisms, the determination of protein bands in digestive tract tissues still heavily relies on the experience of experimenters. This human intervention not only makes it difficult to unify quantitative standards but also limits the efficiency and reproducibility of large-scale digestive tract tissue sample analysis.

[0004] Therefore, how to achieve automated, high-precision, standardized quantitative analysis of protein images of digestive tract tissues under conditions of complex background interference and band morphology distortion, and generate protein expression abundance with high accuracy, has become an urgent problem to be solved. Summary of the Invention

[0005] This application provides a method and system for determining protein expression abundance based on digestive tract tissue. By processing the initial image corresponding to the isolated digestive tract tissue, and assuming electrophoretic distortion exists in the protein standard, a pixel spatial coordinate system is constructed, and angle correction and linear repair are performed. This achieves automated horizontal alignment of distorted bands, eliminating the interference of physical deformation on quantification. Combined with intensity-based irregular region of interest extraction and dynamic compensation of the neighborhood background, the complex background noise of the digestive tract tissue is effectively removed, eliminating irrelevant pixel biases introduced by traditional rectangular bounding boxes. The fully automated protein image processing not only avoids the subjective randomness of manual mapping but also significantly improves the purity and repeatability of signal extraction, providing a high-precision, standardized quantification method for digestive tract tissue protein expression analysis to obtain highly accurate protein expression abundance.

[0006] This application provides a method for determining protein expression abundance based on digestive tract tissue, including: Based on isolated digestive tract tissue, images of polyvinylidene fluoride (PVDF) membranes after colorimetric treatment were obtained. The PVDF membrane images contained bands of at least one protein standard and protein bands corresponding to multiple target proteins to be detected. If electrophoretic distortion is determined based on the bands of the target protein standard, a pixel space coordinate system covering the PVDF membrane image is constructed, wherein the target protein standard is any one of the at least one protein standards. Angle correction and linear restoration are performed on the PVDF membrane image to obtain a target image, wherein the protein bands corresponding to the plurality of target proteins in the target image are horizontally aligned in the pixel space coordinate system. The imaging regions of each of the multiple target proteins in the target image are determined, as well as the irregular regions of interest (ROIs) that enclose the protein bands corresponding to the multiple target proteins. For each target protein, feature analysis and background subtraction are performed on the imaging regions of the target protein and the ROIs of the corresponding protein bands to obtain the expression abundance of the target protein.

[0007] According to an embodiment of this application, a method for determining protein expression abundance based on digestive tract tissue includes constructing a pixel space coordinate system covering the PVDF membrane image when electrophoretic distortion is determined to exist in the target protein standard based on the bands. This includes: for each known molecular weight band in the target protein standard, calculating the geometric center or gray-level centroid of the known molecular weight band based on the edge pixels of the known molecular weight band; using the geometric center or gray-level centroid as the reference feature point of the known molecular weight band; fitting the tilt angle or curvature of the band of the target protein standard based on the pixel coordinates of the reference feature points of each known molecular weight band; determining that electrophoretic distortion exists in the target protein standard when the tilt angle or curvature meets a preset distortion condition; and constructing a pixel space coordinate system covering the PVDF membrane image based on the reference feature points of each known molecular weight band.

[0008] According to an embodiment of this application, a method for determining protein expression abundance based on digestive tract tissue is provided. The method for determining the developmental regions of each of the plurality of target proteins in the target image, and the irregular regions of interest (ROIs) enclosing the protein bands corresponding to the plurality of target proteins, includes: scanning the pixel intensity gradient of the region where each target protein is located in the target image along the lane direction of the pixel spatial coordinate system; identifying pixels where the pixel intensity gradient changes abruptly as edge candidate points for the pixel intensity gradient of the region where each target protein is located, and connecting the edge candidate points to form a closed irregular contour line; determining the internal region of the irregular contour line as the developmental region of the target protein, and determining the smallest set of pixels enclosed by the irregular contour line as the ROI of the protein band corresponding to the target protein.

[0009] According to an embodiment of this application, a method for determining protein expression abundance based on digestive tract tissue includes performing feature analysis and background subtraction on the developmental region of the target protein and the irregular region of interest corresponding to the protein band to obtain the expression abundance of the target protein. This includes: extending a predetermined pixel distance outward from the outer edge of the irregular region of interest along a direction perpendicular to the band extension to determine a neighboring background region with no signal distribution; calculating the average pixel gray value of the neighboring background region; subtracting the product of the average pixel gray value and the total number of pixels in the developmental region from the total gray value integral of the developmental region of the target protein to obtain the protein signal feature value of the target protein; and determining the expression abundance of the target protein based on the protein signal feature value.

[0010] According to an embodiment of this application, a method for determining protein expression abundance based on digestive tract tissue includes the following steps: performing angle correction and linear repair on the PVDF membrane image to obtain a target image. This includes: calculating the offset of each pixel in the PVDF membrane image relative to the standard horizontal position of the pixel space coordinate system based on the tilt angle or curvature, and generating a corresponding affine transformation matrix; performing rotation and translation mapping on the PVDF membrane image using the affine transformation matrix to obtain a transformed PVDF membrane image; and performing linear stretching repair on the local coordinate system for arc-shaped regions in the transformed PVDF membrane image exhibiting smile or frown effects, based on the curvature, so that the protein bands corresponding to the multiple target proteins are horizontally aligned in the pixel space coordinate system, thereby obtaining the target image.

[0011] According to an embodiment of this application, a method for determining protein expression abundance based on digestive tract tissue is provided. The method for determining the expression abundance of the target protein based on the protein signal feature values ​​includes: locating internal reference protein bands in different lanes of the target image according to the pixel spatial coordinate system and a preset internal reference molecular weight range; identifying the target internal reference protein band located in the same lane as the target protein from all internal reference protein bands; performing feature analysis and background subtraction on the imaging region and the irregular region of interest of the target internal reference protein band to obtain the internal reference signal feature value of the target internal reference protein band; and using the ratio of the protein signal feature value to the internal reference signal feature value as the expression abundance of the target protein.

[0012] According to an embodiment of this application, a method for determining protein expression abundance based on digestive tract tissue is provided. After obtaining the internal reference signal feature value of the target internal reference protein band, the method further includes: using the ratio of the internal reference signal feature value to a preset standard internal reference value as a normalization coefficient; and using the ratio of the protein signal feature value to the normalization coefficient as the expression abundance of the target protein.

[0013] This application also provides a system for determining protein expression abundance based on digestive tract tissue, including: The acquisition module is used to acquire images of polyvinylidene fluoride (PVDF) membranes after color development based on isolated digestive tract tissues. The PVDF membrane images include bands of at least one protein standard and protein bands corresponding to multiple target proteins to be detected. The processing module is used to construct a pixel space coordinate system covering the PVDF membrane image when electrophoretic distortion is determined based on the bands of the target protein standard, wherein the target protein standard is any one of the at least one protein standard; perform angle correction and linear restoration on the PVDF membrane image to obtain a target image, wherein the protein bands corresponding to the plurality of target proteins in the target image are horizontally aligned in the pixel space coordinate system; determine the development region of each of the plurality of target proteins in the target image, and the irregular region of interest surrounding the protein bands corresponding to the plurality of target proteins, and perform feature analysis and background subtraction on the development region of the target protein and the irregular region of interest of the corresponding protein band for each target protein to obtain the expression abundance of the target protein.

[0014] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the protein expression abundance determination method based on digestive tract tissue as described above.

[0015] This application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the protein expression abundance determination method based on digestive tract tissue as described above.

[0016] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the protein expression abundance determination method based on digestive tract tissue as described above.

[0017] The method and system for determining protein expression abundance based on digestive tract tissue provided in this application acquire a polyvinylidene fluoride (PVDF) membrane image after colorimetric processing based on isolated digestive tract tissue. The PVDF membrane image includes bands of at least one protein standard and protein bands corresponding to multiple target proteins to be detected. If electrophoretic distortion of the target protein standard is determined based on its bands, a pixel space coordinate system covering the PVDF membrane image is constructed. The target protein standard is any one of the at least one protein standard. Angle correction and linear restoration are performed on the PVDF membrane image to obtain a target image. The protein bands corresponding to the multiple target proteins in the target image are horizontally aligned in the pixel space coordinate system. The development regions of each of the multiple target proteins in the target image and the irregular regions of interest (ROIs) surrounding the corresponding protein bands are determined. For each target protein, feature analysis and background subtraction are performed on the development regions and corresponding ROIs of the target protein to obtain the expression abundance of the target protein. This method processes the initial image corresponding to isolated digestive tract tissue. Given the presence of electrophoretic distortion in protein standards, it constructs a pixel spatial coordinate system and performs angle correction and linear repair, achieving automated horizontal alignment of distorted bands and eliminating the interference of physical deformation on quantification. Combined with intensity-based irregular region of interest extraction and dynamic compensation of the neighborhood background, it effectively removes the complex background noise of digestive tract tissue and eliminates irrelevant pixel biases introduced by traditional rectangular bounding boxes. This fully automated protein image processing not only avoids the subjective randomness of manual mapping but also significantly improves the purity and repeatability of signal extraction, providing a high-precision, standardized quantification method for digestive tract tissue protein expression analysis to obtain highly accurate protein expression abundance. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic flowchart of the method for determining protein expression abundance based on digestive tract tissue provided in the embodiments of this application; Figure 2 This is a schematic diagram showing the expression abundance of multiple target proteins provided in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of the protein expression abundance determination system based on digestive tract tissue provided in the embodiments of this application; Figure 4 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] It should be noted that the execution entity involved in the embodiments of this application can be a protein expression abundance determination system based on digestive tract tissue, or it can be an electronic device. Optionally, the electronic device may include: a computer / laptop, a mobile terminal, a server, a cloud processor, an embedded image processing module, and an automated detection terminal, etc.

[0022] The following uses an electronic device as an example to illustrate in detail the method for determining protein expression abundance based on digestive tract tissue provided in the embodiments of this application: Figure 1 This is a schematic flowchart of a method for determining protein expression abundance based on digestive tract tissue provided in an embodiment of this application. Figure 1 As shown, the method includes the following steps 101-104.

[0023] Step 101: Based on the isolated digestive tract tissue, obtain an image of a polyvinylidene fluoride (PVDF) membrane after colorimetric treatment. The PVDF membrane image contains the bands of at least one protein standard and the protein bands corresponding to multiple target proteins to be detected.

[0024] Among them, isolated digestive tract tissues refer to isolated specimens from tissues such as the esophagus, stomach, small intestine, or large intestine.

[0025] Polyvinylidene fluoride (PVDF) membrane images refer to digital photosensitive images captured by a cold charge-coupled device (CCD) imaging system, laser scanner, or optical camera after transferring proteins from digestive tract tissue gel to a PVDF membrane via Western blotting, followed by immunoreaction and enhanced chemiluminescence (ECL) or substrate color development, using Western blotting technology.

[0026] A protein marker is a standardized mixture of proteins containing multiple known molecular weights used to indicate the molecular weight position of a target protein. Optionally, the multiple known molecular weights may include at least 10 kilodaltons (kDa), 15 kDa, 25 kDa, 35 kDa, 45 kDa, 55 kDa, 70 kDa, 100 kDa, 130 kDa, 170 kDa, and 250 kDa, forming a continuous ladder covering the low to high molecular weight range.

[0027] A protein marker band refers to a pixel-rich region with a specific shape (usually near-rectangular or arc-shaped) formed on a PVDF membrane after electrophoresis and staining of a standardized protein mixture. The pixel intensity or gray value of the band is positively correlated with the amount of protein accumulated at that band.

[0028] Target proteins refer to specific proteins in the digestive tract tissues that are to be analyzed qualitatively or quantitatively. Optionally, multiple target proteins may include at least Ano1, E-cadherin, Vimentin, Fibronectin, and α-SMA.

[0029] The protein band corresponding to the target protein refers to the pixel-rich region with a specific shape formed on the PVDF membrane after electrophoresis and staining. The pixel intensity or gray value of the protein band is positively correlated with the protein content accumulated at the protein band.

[0030] In step 101, the electronic device acquires an initial image and a PVDF membrane image after color development based on ex vivo digestive tract tissue. The PVDF membrane image can retain the complete sixteen-bit grayscale dynamic range. In addition, the PVDF membrane image contains at least two parallel signal distributions: one is a marker lane containing multiple known molecular weight ladders, which is used to provide a spatial mapping benchmark to determine the geometric distortion parameters of the PVDF membrane image; the other is a target lane containing the sample to be tested (i.e., multiple target proteins to be detected).

[0031] The initial image is a raw image or a microscopic image of a tissue section used to macroscopically record the aforementioned ex vivo specimen. It records the origin, pathological features, and original morphology of the specimen and serves as background reference data for subsequent protein expression abundance analysis. It should be noted that the PVDF membrane image and the initial image have a pre-defined correlation. Specifically, the pixel coordinate regions in the PVDF membrane image have a logical mapping relationship with the sample labels in the initial image, ensuring that protein abundance data can be traced back to a specific tissue origin.

[0032] Optionally, the initial image is a single-channel grayscale image to preserve the original protein feature signal intensity; or, the initial image is a multi-channel color image, in which case the multi-channel color image can be converted into a grayscale image to eliminate the interference of color information on grayscale centroid calculation in the future.

[0033] Step 102: If the target protein standard is found to have electrophoretic distortion based on the bands of the target protein standard, construct a pixel space coordinate system covering the PVDF membrane image. The target protein standard is any one of at least one protein standard.

[0034] Electrophoretic distortion refers to the phenomenon where, during polyacrylamide gel electrophoresis, unstable voltage, uneven gel polymerization, or edge effects cause the protein molecules to deviate from the ideal vertical direction, which manifests as overall tilting, local bending (such as the "smiley face effect" or "crying face effect"), or nonlinear stretching of the bands in the PVDF membrane image.

[0035] The pixel space coordinate system is a two-dimensional rectangular coordinate system established with the key pixel point (such as the top left corner vertex) of the PVDF film image as the origin, and the pixel row direction as the X-axis and the pixel column direction as the Y-axis. It is used to assign a unique spatial index to each pixel point in the PVDF film image.

[0036] In step 102, after acquiring at least one protein standard, the electronic device can determine whether the target protein standard exhibits electrophoretic distortion based on the bands of any one of the protein standards, i.e., the target protein standard. If no distortion is found, it indicates that the protein bands in the current PVDF membrane image conform to an ideal vertical linear distribution, and the PVDF membrane image has not undergone significant physical deformation. In this case, the complex image reconstruction and affine transformation steps can be skipped, and the conventional segmentation and quantization process for the protein bands corresponding to the target protein can be directly initiated. If distortion is found, it indicates that the protein has been affected by an inhomogeneous electric field or gel stress during electrophoretic migration, resulting in tilting or bending. In this case, an adaptive correction mechanism can be triggered to construct a pixel spatial coordinate system covering the PVDF membrane image. This step, by introducing an automatic distortion determination mechanism, achieves separate processing of high-quality images and distorted images, ensuring accurate correction of complex distorted samples while optimizing the overall processing efficiency of the electronic device.

[0037] In some embodiments, when the electronic device determines that the target protein standard has electrophoretic distortion based on the bands of the target protein standard, constructing a pixel space coordinate system covering the PVDF membrane image may include: for each known molecular weight band in the target protein standard, the electronic device calculates the geometric center or gray-level centroid of the known molecular weight band based on the edge pixels of the known molecular weight band; and uses the geometric center or gray-level centroid as the reference feature point of the known molecular weight band; the electronic device fits the tilt angle or curvature of the band of the target protein standard based on the pixel coordinates of the reference feature point of each known molecular weight band; when the tilt angle or curvature meets the preset distortion conditions, the electronic device determines that the target protein standard has electrophoretic distortion, and constructs a pixel space coordinate system covering the PVDF membrane image based on the reference feature point of each known molecular weight band.

[0038] Among them, known molecular weight bands refer to the imaging patches formed on the membrane by a single protein component with a known molecular weight.

[0039] The geometric center refers to the centroid of the closed region containing a known molecular weight band, calculated solely based on the coordinates of the edge pixels of the known molecular weight band.

[0040] Gray-scale centroid refers to the weighted center point calculated using the gray-scale values ​​of each pixel within a known molecular weight band as weights.

[0041] The tilt angle of a strip refers to the angle between the line connecting the reference feature points of multiple known molecular weight strips and the vertical axis (or horizontal axis) of the pixel space coordinate system, and is used to measure the overall deflection of the PVDF film image.

[0042] The curvature of a strip refers to the degree to which the fitted curve of multiple reference feature points deviates from a straight line, and is used to characterize the nonlinear arc deformation generated during electrophoresis.

[0043] Preset distortion conditions refer to the threshold standards used to trigger image correction actions.

[0044] Optionally, the preset distortion condition is that the tilt angle is greater than 0.5°, or the vertical displacement of the line connecting the midpoint and the endpoint of the arc corresponding to the curvature exceeds 2 pixels.

[0045] In this embodiment, the electronic device first acquires multiple known molecular weight bands in the target protein standard. For each known molecular weight band, the electronic device calculates the geometric center or gray-scale centroid of the known molecular weight band based on the edge pixels of the known molecular weight band, and uses the geometric center or gray-scale centroid as the reference feature point of the known molecular weight band. Based on this, the electronic device can acquire the reference feature points of each of the multiple known molecular weight bands. Then, the electronic device fits the tilt angle or curvature of the band of the target protein standard based on the pixel coordinates of the reference feature points of each of the multiple known molecular weight bands, and determines whether the tilt angle or curvature meets the preset distortion condition: if it does not meet the condition, it means that the target protein standard does not have electrophoretic distortion; if it does meet the condition, it means that the target protein standard has electrophoretic distortion. At this time, the reference feature points of each known molecular weight band are used as the reference, that is, any one of the reference feature points of each of the multiple known molecular weight bands is used as the key pixel point, and then as the origin, to construct a pixel space coordinate system covering the PVDF membrane image. The entire process establishes a pixel space coordinate system based on key pixels to map the originally non-linearly distributed protein signals into a standardized mathematical space. This method of determining coordinates based on feature points provides a physical positioning benchmark for the accurate comparison of trace protein expression in digestive tract tissues, avoiding quantitative errors caused by image skew.

[0046] Step 103: Perform angle correction and linear restoration on the PVDF membrane image to obtain the target image. The protein bands corresponding to multiple target proteins in the target image are horizontally aligned in the pixel space coordinate system.

[0047] The target image refers to a digital image that, after geometric deformation correction, eliminates global tilt and local arc curvature, so that the center lines of the bands of the same protein in different in vitro sample lanes are at the same pixel height.

[0048] In step 103, the electronic device performs angle correction and linear restoration on the PVDF membrane image using the pixel space coordinate system as a reference. This ensures that the protein bands corresponding to multiple target proteins in the PVDF membrane image are horizontally aligned in the pixel space coordinate system, resulting in a target image where all protein bands are horizontally aligned. This step is a fully automated restoration process that not only eliminates physical random deformations caused by experimental operations (such as membrane transfer and electrophoresis) but also unifies the search window for subsequent feature extraction. Understandably, because all protein bands are horizontally aligned, the electronic device can employ a consistent horizontal scanning strategy to accurately define the protein regions in each lane, significantly reducing subsequent computational complexity and providing an absolutely aligned reference benchmark for lateral comparison of protein expression abundance across in vitro samples.

[0049] In some embodiments, the electronic device performs angle correction and linear restoration on a PVDF membrane image to obtain a target image. This may include: the electronic device calculating the offset of each pixel in the PVDF membrane image relative to the standard horizontal position of the pixel space coordinate system based on the tilt angle or curvature, and generating a corresponding affine transformation matrix; the electronic device performing rotation and translation mapping on the PVDF membrane image using the affine transformation matrix to obtain a transformed PVDF membrane image; and the electronic device performing linear stretching restoration on the local coordinate system for arc-shaped regions in the transformed PVDF membrane image that exhibit smiley or sad face effects based on the curvature, so that the protein bands corresponding to multiple target proteins are horizontally aligned in the pixel space coordinate system, thereby obtaining the target image.

[0050] The standard horizontal position refers to the ideal pixel row coordinates in the pixel space coordinate system, which are preset and theoretically free from any physical distortion. It is usually represented as an absolutely horizontal straight line perpendicular to the direction of swimlane migration.

[0051] An affine transformation matrix is ​​a linear transformation matrix between two-dimensional coordinates. It is used to describe operations such as rotation, translation, scaling, and shearing of an image to correct the overall tilt angle of the transformed PVDF film image.

[0052] In this embodiment, the electronic device first calculates the offset of each pixel in the PVDF film image relative to the standard horizontal position of the pixel spatial coordinate system based on the tilt angle or curvature. Then, based on this offset, an affine transformation matrix is ​​generated, typically a 3×3 matrix. Next, the electronic device uses this affine transformation matrix to recalculate the position of each pixel in the PVDF film image, obtaining an image where the entire swimming lane is vertically downward, i.e., the transformed PVDF film image. Although the PVDF film image is straightened, uneven heat generation during electrophoresis often results in the middle stripe of the transformed PVDF film image having a different velocity than the two sides, leading to… The transformed PVDF membrane image exhibits a "smiley face" or "crying face" effect, resulting in residual arc-shaped nonlinear deformation after global correction. Therefore, local linear stretching is necessary. Specifically, for the arc-shaped regions exhibiting the smiley face or crying face effect, the vertical distance of each row of pixels deviating from the ideal horizontal line in this region is identified based on the curvature. Local coordinate system linear stretching is then performed to repair this effect, compensating for the curved pixels by longitudinal compression or stretching row by row. This achieves dynamic and local repair of the transformed PVDF membrane image, ensuring that the protein bands corresponding to multiple target proteins are horizontally aligned in the pixel space coordinate system, thus eliminating quantitative errors caused by spatial position and obtaining the target image. The entire process, through dynamic and local pixel resampling, eliminates nonlinear geometric distortions in the PVDF membrane image, ensuring strict horizontal alignment of the protein bands corresponding to multiple target proteins in the pixel space coordinate system. Furthermore, through a two-stage global and local transformation, the complex biochemical imaging results are converted into a standardized digital matrix, effectively eliminating quantitative errors caused by spatial position deviations and providing a normalized spatial benchmark for accurate protein abundance comparison across in vitro samples.

[0053] Step 104: Determine the developmental regions of multiple target proteins in the target image, as well as the irregular regions of interest (ROIs) that enclose the protein bands corresponding to the multiple target proteins. For each target protein, perform feature analysis and background subtraction on the developmental regions of the target protein and the ROIs of the corresponding protein bands to obtain the expression abundance of the target protein.

[0054] The development region of the target protein refers to the pixel cluster area in the target image that exhibits a higher gray intensity than the background after the protein binds to a specific antibody and undergoes a color development reaction. It physically characterizes the actual spatial distribution of the target protein on the PVDF film.

[0055] The irregular region of interest (ROI) for protein strips is a tightly closed region generated based on the actual geometric contour of the protein strip. Unlike traditional rectangular or circular ROIs, the boundary of this irregular ROI varies with the pixel gradient fluctuation of the corresponding protein strip, aiming to eliminate blank background pixels without protein signals to the greatest extent possible, thereby achieving precise wrapping of the target signal.

[0056] The expression abundance of a target protein, also known as relative expression abundance, is a measure of the signal intensity of the target protein.

[0057] In step 104, the electronic device identifies the target image, determines the developmental regions of multiple target proteins, and the irregular regions of interest (ROIs) surrounding the protein bands corresponding to the multiple target proteins. Then, for each target protein, the electronic device automatically generates an irregular mask that closely matches the band shape based on the pixel gradient changes at the edges of the developmental regions. This irregular mask defines the effective set of pixels participating in the integral calculation, excluding invalid background pixels outside the edges of the irregular bands. Simultaneously, background reference pixels are extracted from the outer periphery of the irregular ROI (e.g., within 5-10 pixels above and below the protein bands), and the average background noise per unit area of ​​this local region is calculated. Combined with the aforementioned effective pixel set, the expression abundance of the target protein is determined. This step eliminates the accumulation of background noise caused by blank areas at the band edges by using irregular ROIs instead of traditional ROIs. Combined with neighborhood background subtraction technology, this step can significantly reduce the background noise interference caused by the non-specific color development of the PVDF membrane, making the obtained expression abundance closer to the true physical content of the protein.

[0058] For example, Figure 2 This is a schematic diagram illustrating the expression abundance of multiple target proteins provided in the embodiments of this application. From Figure 2 It can be seen that, under the influence of TGF-β1 and control, the relative expression levels of Ano1, E-cadherin, Vimentin, Fibronectin and α-SMA are different.

[0059] In some embodiments, the electronic device determines the development regions of multiple target proteins in a target image, and the irregular regions of interest (ROIs) that enclose the protein bands corresponding to the multiple target proteins. This may include: the electronic device scanning the pixel intensity gradient of the region where each target protein is located in the target image along the lane direction of the pixel spatial coordinate system; for the pixel intensity gradient of the region where each target protein is located, the electronic device identifies pixels where the pixel intensity gradient changes abruptly as edge candidate points, and connects the edge candidate points to form a closed irregular contour line; the internal region of the irregular contour line is determined as the development region of the target protein, and the smallest set of pixels enclosed by the irregular contour line is determined as the ROI of the protein band corresponding to the target protein.

[0060] In this embodiment, the electronic device, during the process of determining the development region and irregular region of interest (ROI) of each target protein in the target image, can execute the following adaptive boundary recognition logic: First, the electronic device scans the region where the target protein is located in the target image row by row / column by column along the swimlane direction of the pixel spatial coordinate system (i.e., the longitudinal direction of protein migration); then, the electronic device identifies pixels where the pixel intensity gradient changes abruptly as edge candidate points by calculating the gray-level difference or Laplacian operator between adjacent pixels. These candidate points are used to characterize the physical boundary between the protein development signal and the background noise; subsequently, the electronic device uses a contour tracking algorithm (such as the eight-neighborhood search method) to smoothly connect these edge candidate points according to spatial topological relationships, forming a closed irregular contour line; finally, the electronic device determines the internal region of the irregular contour line as the development region of the target protein, i.e., the set of pixels that actually carry the protein signal, and determines the smallest set of pixels enclosed by the irregular contour line as the ROI of the corresponding protein band of the target protein. Based on this, the electronic device can obtain the development region and ROI of multiple target proteins. The entire process is based on gradient mutation identification, enabling irregular regions of interest to tightly encapsulate protein bands. Even when protein bands exhibit localized deformation or blurred edges, the electronic device dynamically adjusts the sampling range according to the true distribution of pixel intensity (i.e., pixel intensity gradient), effectively avoiding the signal dilution problem caused by traditional rectangular bounding boxes containing a large number of invalid background pixels. In other words, it achieves dynamic capture of denatured or blurred band signals, significantly improving signal purity and quantitative accuracy against complex backgrounds.

[0061] In some embodiments, the electronic device performs feature analysis and background subtraction on the development region of the target protein and the irregular region of interest corresponding to the protein band to obtain the expression abundance of the target protein. This may include: the electronic device extending a predetermined pixel distance outward from the outer edge of the irregular region of interest along a direction perpendicular to the band extension to determine a neighboring background region with no signal distribution; the electronic device calculating the average pixel gray value of the neighboring background region; the electronic device subtracting the product of the average pixel gray value and the total number of pixels in the development region from the total gray integral of the development region of the target protein to obtain the protein signal feature value of the target protein; and the electronic device determining the expression abundance of the target protein based on the protein signal feature value.

[0062] In this embodiment, after obtaining the development region of the target protein and the irregular region of interest of the corresponding protein band, the electronic device performs the following dynamic background compensation and abundance calculation logic: First, the electronic device extends outward by a preset pixel distance (e.g., 5-10 pixels upward and downward) along the outer edge of the irregular region of interest, perpendicular to the band extension direction (i.e., the longitudinal migration direction), to determine a neighboring background region without signal distribution (i.e., without target protein signal distribution); then, the electronic device calculates the average pixel gray value of the neighboring background region, which is used as the local noise floor reference for a specific band position; next, the electronic device performs differential integration: from the total gray value of the development region of the target protein, the product of the average pixel gray value and the total number of pixels in the development region is subtracted, thereby eliminating the interference of PVDF membrane background color development and optical noise, and obtaining a pure protein signal feature value; finally, based on the protein signal feature value, the electronic device determines the expression abundance of the target protein by performing a ratio calculation with the feature value of the internal reference protein (such as GAPDH or Actin) in the same lane. The entire process adopted a local sampling mode of "one protein, one background" instead of uniform background subtraction across the entire image. This effectively offset the spatial fluctuation interference caused by uneven color development of the PVDF film (such as dark edges and bright centers). At the same time, by using the logic of multiplying the average pixel gray value by the total number of pixels in the developed area, the deviation in the total background introduced by the different sizes of irregular developed areas was automatically compensated, ensuring the fairness of quantitative analysis between different shaped bands and improving the accuracy of expression abundance.

[0063] In some embodiments, the electronic device determines the expression abundance of a target protein based on protein signal characteristic values, which may include one of the following implementation methods: Implementation Method 1: The electronic device locates the internal reference protein band in different lanes of the target image based on the pixel spatial coordinate system and the preset internal reference molecular weight range; the electronic device identifies the target internal reference protein band that is in the same lane as the target protein from all internal reference protein bands; the electronic device performs feature analysis and background subtraction on the development area and irregular region of interest of the target internal reference protein band to obtain the internal reference signal feature value of the target internal reference protein band; the electronic device uses the ratio of the protein signal feature value to the internal reference signal feature value as the expression abundance of the target protein.

[0064] In implementation method 1, the electronic device first automatically retrieves and locates the pixel intervals of the internal reference protein bands in each lane of the target image based on the constructed pixel spatial coordinate system and the preset internal reference molecular weight range (e.g., GAPDH is approximately 36 kDa, β-actin is approximately 42 kDa). Then, the electronic device accurately locks the target internal reference protein band that is on the same migration path (i.e., in the same lane) as the target protein from all identified internal reference bands, ensuring that the target signal and the reference signal originate from the same sample and have the highest biological comparability. Next, the electronic device uses the same irregular region of interest extraction and neighborhood background subtraction algorithm as in step 104 to process the target internal reference protein band. Specifically, by performing grayscale integration on the internal reference imaging area and subtracting its local background noise, the internal reference signal feature value of the lane is calculated. Finally, the electronic device performs normalization calculation: the protein signal feature value of the target protein is compared with the corresponding internal reference signal feature value, and the resulting ratio is used as the expression abundance of the target protein in the ex vivo digestive tract tissue. The entire process, through this same-lane internal control calibration mechanism, effectively eliminates errors caused by non-experimental factors such as minor deviations in the amount of digestive tract tissue sampled, protein loss during electrophoresis, and non-uniformity of membrane transfer. Converting absolute gray values ​​into relative ratios based on internal controls ensures that the detection results accurately and objectively reflect the physiological expression levels of target proteins such as Ano1 and E-cadherin in the tissue.

[0065] Implementation Method 2: The electronic device locates the internal reference protein band in different lanes of the target image based on the pixel spatial coordinate system and the preset internal reference molecular weight range; the electronic device identifies the target internal reference protein band in the same lane as the target protein from all internal reference protein bands; the electronic device performs feature analysis and background subtraction on the development area and irregular region of interest of the target internal reference protein band to obtain the internal reference signal feature value of the target internal reference protein band; the electronic device uses the ratio of the internal reference signal feature value to the preset standard internal reference value as the normalization coefficient; the electronic device uses the ratio of the protein signal feature value to the normalization coefficient as the expression abundance of the target protein.

[0066] In implementation method 2, the process of acquiring the internal reference signal feature value by the electronic device is the same as that in implementation method 1, and will not be elaborated here. Based on this, after obtaining the internal reference signal feature value of the target internal reference protein band, the electronic device can perform a ratio calculation between this internal reference signal feature value and a preset standard internal reference value (which is usually a grayscale baseline or the statistical mean of multiple samples under a pre-set ideal loading amount) to obtain the normalization coefficient for that specific lane. This normalization coefficient reflects the degree of deviation of the actual loading amount in the current lane from the standard level. Finally, the electronic device divides the protein signal feature value by the normalization coefficient to obtain the ratio as the expression abundance of the target protein. The entire process establishes a standard reference system. Since the colorimetric intensity of different batches of PVDF membranes may have systematic differences in actual processes, by introducing a preset standard internal reference value, the electronic device can map the signal intensity of all samples to the same standard scale, obtaining a more accurate target protein expression abundance.

[0067] In this embodiment, the technical solution described in steps 101-104 above processes the initial image corresponding to the isolated digestive tract tissue. When electrophoretic distortion is confirmed in the protein standard, a pixel spatial coordinate system is constructed, and angle correction and linear repair are performed. This achieves automated horizontal alignment of distorted bands, eliminating the interference of physical deformation on quantification. Combined with intensity-based irregular region of interest extraction and dynamic compensation of the neighborhood background, the complex background noise of the digestive tract tissue is effectively removed, eliminating irrelevant pixel bias introduced by traditional rectangular bounding boxes. This fully automated protein image processing not only avoids the subjective randomness of manual mapping but also significantly improves the purity and repeatability of signal extraction, providing a high-precision, standardized quantification method for digestive tract tissue protein expression analysis to obtain highly accurate protein expression abundance.

[0068] The following describes the protein expression abundance determination system based on digestive tract tissue provided in the embodiments of this application. The protein expression abundance determination system based on digestive tract tissue described below can be referred to in correspondence with the protein expression abundance determination method based on digestive tract tissue described above.

[0069] Figure 3 This is a schematic diagram of the structure of the protein expression abundance determination system based on digestive tract tissue provided in an embodiment of this application. Figure 3 As shown, the system includes: an acquisition module 301 and a processing module 302.

[0070] The acquisition module 301 is used to acquire a polyvinylidene fluoride (PVDF) membrane image after color development based on isolated digestive tract tissue. The PVDF membrane image contains at least one protein standard band and protein bands corresponding to multiple target proteins to be detected. The processing module 302 is used to construct a pixel space coordinate system covering the PVDF membrane image when it is determined that the target protein standard has electrophoretic distortion based on the bands of the target protein standard. The target protein standard is any one of the at least one protein standard. The module performs angle correction and linear repair on the PVDF membrane image to obtain a target image. The protein bands corresponding to the multiple target proteins in the target image are horizontally aligned in the pixel space coordinate system. The module determines the development region of each of the multiple target proteins in the target image and the irregular region of interest that surrounds the protein bands corresponding to the multiple target proteins. For each target protein, the module performs feature analysis and background subtraction on the development region of the target protein and the irregular region of interest of the corresponding protein band to obtain the expression abundance of the target protein.

[0071] Optionally, the processing module 302 is specifically used to calculate the geometric center or gray-scale centroid of each known molecular weight band in the target protein standard based on the edge pixels of the known molecular weight band; and use the geometric center or gray-scale centroid as the reference feature point of the known molecular weight band; and fit the tilt angle or curvature of the band of the target protein standard based on the pixel coordinates of the reference feature points of each known molecular weight band; if the tilt angle or curvature meets the preset distortion conditions, determine that the target protein standard has electrophoretic distortion, and construct a pixel space coordinate system covering the PVDF membrane image based on the reference feature points of each known molecular weight band.

[0072] Optionally, the processing module 302 is specifically used to scan the pixel intensity gradient of each target protein region in the target image along the lane direction of the pixel spatial coordinate system; for each target protein region, identify pixels where the pixel intensity gradient changes abruptly as edge candidate points, and connect the edge candidate points to form a closed irregular contour line; determine the internal region of the irregular contour line as the development region of the target protein, and determine the smallest set of pixels enclosed by the irregular contour line as the irregular region of interest of the protein band corresponding to the target protein.

[0073] Optionally, the processing module 302 is specifically configured to extend a preset pixel distance outward along the outer edge of the irregular region of interest in a direction perpendicular to the strip extension direction to determine a neighboring background area with no signal distribution; calculate the average pixel gray value of the neighboring background area; subtract the product of the average pixel gray value and the total number of pixels in the developing area from the total gray value integral of the developing area of ​​the target protein to obtain the protein signal feature value of the target protein; and determine the expression abundance of the target protein based on the protein signal feature value.

[0074] Optionally, the processing module 302 is specifically used to calculate the offset of each pixel in the PVDF membrane image relative to the standard horizontal position of the pixel space coordinate system according to the tilt angle or the curvature, and generate the corresponding affine transformation matrix; perform rotation and translation mapping on the PVDF membrane image using the affine transformation matrix to obtain the transformed PVDF membrane image; for the arc-shaped regions in the transformed PVDF membrane image that have smiley or sad face effects, perform linear stretching repair of the local coordinate system according to the curvature, so that the protein bands corresponding to the multiple target proteins are horizontally aligned in the pixel space coordinate system, thereby obtaining the target image.

[0075] Optionally, the processing module 302 is specifically used to locate the internal reference protein band in different lanes of the target image according to the pixel spatial coordinate system and the preset internal reference molecular weight range; determine the target internal reference protein band in the same lane as the target protein from all internal reference protein bands; perform feature analysis and background subtraction on the development area and the irregular region of interest of the target internal reference protein band to obtain the internal reference signal feature value of the target internal reference protein band; and use the ratio of the protein signal feature value to the internal reference signal feature value as the expression abundance of the target protein.

[0076] Optionally, the processing module 302 is further configured to use the ratio of the feature value of the internal reference signal to the preset standard internal reference value as a normalization coefficient; and to use the ratio of the protein signal feature value to the normalization coefficient as the expression abundance of the target protein.

[0077] Figure 4 This is a schematic diagram of the structure of the electronic device provided in an embodiment of this application. For example... Figure 4As shown, the electronic device may include: a processor 410, a communications interface 420, a memory 430, and a communications bus 440, wherein the processor 410, the communications interface 420, and the memory 430 communicate with each other through the communications bus 440. The processor 410 can call logic instructions in the memory 430 to execute a method for determining protein expression abundance based on digestive tract tissue. This method includes: acquiring a polyvinylidene fluoride (PVDF) membrane image after colorimetric processing based on isolated digestive tract tissue, wherein the PVDF membrane image contains bands of at least one protein standard and protein bands corresponding to multiple target proteins to be detected; constructing a pixel space coordinate system covering the PVDF membrane image, wherein the target protein standard is any one of the at least one protein standard, based on the bands of the target protein standard; performing angle correction and linear restoration on the PVDF membrane image to obtain a target image, wherein the protein bands corresponding to the multiple target proteins in the target image are horizontally aligned in the pixel space coordinate system; determining the development regions of each of the multiple target proteins in the target image, and the irregular regions of interest (ROIs) surrounding the protein bands corresponding to the multiple target proteins, and performing feature analysis and background subtraction on the development regions of the target proteins and the ROIs of the corresponding protein bands for each target protein to obtain the expression abundance of the target protein.

[0078] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0079] On the other hand, embodiments of this application also provide a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the protein expression abundance determination method based on digestive tract tissue provided by the above methods. This method includes: acquiring a polyvinylidene fluoride (PVDF) membrane image after colorimetric treatment based on isolated digestive tract tissue, wherein the PVDF membrane image includes bands of at least one protein standard and protein bands corresponding to multiple target proteins to be detected; and constructing a membrane covering the target protein standard if electrophoretic distortion is determined based on the bands of the target protein standard. The pixel space coordinate system of the PVDF membrane image is defined, and the target protein standard is any one of the at least one protein standard. Angle correction and linear restoration are performed on the PVDF membrane image to obtain a target image. The protein bands corresponding to the plurality of target proteins in the target image are horizontally aligned in the pixel space coordinate system. The development regions of each of the plurality of target proteins in the target image, and the irregular regions of interest (ROIs) surrounding the corresponding protein bands of the plurality of target proteins are determined. For each target protein, feature analysis and background subtraction are performed on the development region of the target protein and the corresponding ROI of the protein band to obtain the expression abundance of the target protein.

[0080] Furthermore, embodiments of this application also provide a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, this computer program performs the protein expression abundance determination method based on digestive tract tissue provided by the methods described above. This method includes: acquiring a polyvinylidene fluoride (PVDF) membrane image after colorimetric treatment based on isolated digestive tract tissue, wherein the PVDF membrane image contains bands corresponding to at least one protein standard and multiple target proteins to be detected; and, if electrophoretic distortion is determined based on the bands of the target protein standards, constructing a pixel space covering the PVDF membrane image. A coordinate system is used, wherein the target protein standard is any one of the at least one protein standard; the PVDF membrane image is subjected to angle correction and linear restoration to obtain a target image, wherein the protein bands corresponding to the plurality of target proteins in the target image are horizontally aligned in the pixel space coordinate system; the development regions of each of the plurality of target proteins in the target image and the irregular regions of interest surrounding the protein bands corresponding to the plurality of target proteins are determined, and for each target protein, feature analysis and background subtraction are performed on the development region of the target protein and the irregular regions of interest of the corresponding protein bands to obtain the expression abundance of the target protein.

[0081] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units 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. Those skilled in the art can understand and implement this without any creative effort.

[0082] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for determining protein expression abundance based on digestive tract tissue, characterized in that, include: Based on isolated digestive tract tissue, images of polyvinylidene fluoride (PVDF) membranes after colorimetric treatment were obtained. The PVDF membrane images contained bands of at least one protein standard and protein bands corresponding to multiple target proteins to be detected. If electrophoretic distortion is determined based on the bands of the target protein standard, a pixel space coordinate system covering the PVDF membrane image is constructed, wherein the target protein standard is any one of the at least one protein standards. Angle correction and linear restoration are performed on the PVDF membrane image to obtain a target image, wherein the protein bands corresponding to the plurality of target proteins in the target image are horizontally aligned in the pixel space coordinate system. The imaging regions of each of the multiple target proteins in the target image are determined, as well as the irregular regions of interest (ROIs) that enclose the protein bands corresponding to the multiple target proteins. For each target protein, feature analysis and background subtraction are performed on the imaging regions of the target protein and the ROIs of the corresponding protein bands to obtain the expression abundance of the target protein.

2. The method for determining protein expression abundance based on digestive tract tissue according to claim 1, characterized in that, The step of constructing a pixel space coordinate system covering the PVDF membrane image, in the case where electrophoretic distortion of the target protein standard is determined based on its bands, includes: For each known molecular weight band in the target protein standard, the geometric center or gray-scale centroid of the known molecular weight band is calculated based on the edge pixels of the known molecular weight band; and the geometric center or gray-scale centroid is used as the reference feature point of the known molecular weight band. Based on the pixel coordinates of the reference feature points of each known molecular weight band, the tilt angle or curvature of the target protein standard band is fitted to obtain the result. When the tilt angle or the curvature meets the preset distortion conditions, it is determined that the target protein standard has electrophoretic distortion, and a pixel space coordinate system covering the PVDF membrane image is constructed based on the reference feature points of each known molecular weight band.

3. The method for determining protein expression abundance based on digestive tract tissue according to claim 1, characterized in that, Determining the developmental regions of each of the plurality of target proteins in the target image, and the irregular regions of interest enclosing the protein bands corresponding to the plurality of target proteins, includes: Scan the pixel intensity gradient of the region where each target protein is located in the target image along the lane direction of the pixel space coordinate system; For the pixel intensity gradient of the region where each target protein is located, the pixels where the pixel intensity gradient changes abruptly are identified as edge candidate points, and closed irregular contour lines are formed by connecting the edge candidate points; the internal region of the irregular contour lines is determined as the development region of the target protein, and the smallest set of pixels enclosed by the irregular contour lines is determined as the irregular region of interest of the protein band corresponding to the target protein.

4. The method for determining protein expression abundance based on digestive tract tissue according to claim 1 or 3, characterized in that, The process of performing feature analysis and background subtraction on the visualized region of the target protein and the corresponding irregular region of interest of the protein band to obtain the expression abundance of the target protein includes: At the outer edge of the irregular region of interest, a predetermined pixel distance is extended outward along the direction perpendicular to the strip extension to determine the adjacent background area with no signal distribution; Calculate the average pixel grayscale value of the adjacent background area; The protein signal feature value of the target protein is obtained by subtracting the product of the average pixel gray value and the total number of pixels in the developed area from the total gray integral of the developed area of ​​the target protein. The expression abundance of the target protein is determined based on the protein signal feature values.

5. The method for determining protein expression abundance based on digestive tract tissue according to claim 2, characterized in that, The step of performing angle correction and linear restoration on the PVDF film image to obtain the target image includes: Based on the tilt angle or the curvature, calculate the offset of each pixel in the PVDF film image relative to the standard horizontal position of the pixel space coordinate system, and generate the corresponding affine transformation matrix. The PVDF film image is rotated and translated using the affine transformation matrix to obtain the transformed PVDF film image. For the arc-shaped regions in the transformed PVDF membrane image that exhibit smiley or sad face effects, linear stretching repair is performed on the local coordinate system based on the curvature, so that the protein bands corresponding to the multiple target proteins are horizontally aligned in the pixel space coordinate system, thus obtaining the target image.

6. The method for determining protein expression abundance based on digestive tract tissue according to claim 4, characterized in that, Determining the expression abundance of the target protein based on the protein signal feature values ​​includes: Based on the pixel spatial coordinate system and the preset internal reference molecular weight range, the internal reference protein bands are located in different lanes of the target image; From all the internal control protein bands, identify the target internal control protein band that is in the same lane as the target protein; Feature analysis and background subtraction are performed on the developed region of the target internal reference protein band and the irregular region of interest of the target internal reference protein band to obtain the internal reference signal feature value of the target internal reference protein band; The ratio of the protein signal feature value to the internal reference signal feature value is used as the expression abundance of the target protein.

7. The method for determining protein expression abundance based on digestive tract tissue according to claim 4, characterized in that, After obtaining the internal reference signal characteristic values ​​of the target internal reference protein band, the method further includes: The ratio of the intrinsic parameter signal feature value to the preset standard intrinsic parameter value is used as the normalization coefficient; The ratio of the protein signal feature value to the normalization coefficient is used as the expression abundance of the target protein.

8. A system for determining protein expression abundance based on digestive tract tissue, characterized in that, include: The acquisition module is used to acquire images of polyvinylidene fluoride (PVDF) membranes after color development based on isolated digestive tract tissues. The PVDF membrane images include bands of at least one protein standard and protein bands corresponding to multiple target proteins to be detected. The processing module is used to construct a pixel space coordinate system covering the PVDF membrane image when it is determined that the target protein standard has electrophoretic distortion based on the bands of the target protein standard, wherein the target protein standard is any one of the at least one protein standard. Angle correction and linear restoration are performed on the PVDF membrane image to obtain a target image, wherein the protein bands corresponding to the plurality of target proteins in the target image are horizontally aligned in the pixel space coordinate system. The imaging regions of each of the multiple target proteins in the target image are determined, as well as the irregular regions of interest (ROIs) that enclose the protein bands corresponding to the multiple target proteins. For each target protein, feature analysis and background subtraction are performed on the imaging regions of the target protein and the ROIs of the corresponding protein bands to obtain the expression abundance of the target protein.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for determining protein expression abundance based on digestive tract tissue as described in any one of claims 1-7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for determining protein expression abundance based on digestive tract tissue as described in any one of claims 1-7.

Images