Biochip image analysis method and device, computer equipment and storage medium

Abstract

A method for analyzing a biochip image includes: (S1) acquiring a biochip image and preprocessing it to obtain a preprocessed image; (S2) performing angle deflection correction on the preprocessed image to obtain a deflection-corrected image; and (S3) performing enhancement processing on the deflection-corrected image and identifying the positive or negative nature of a region of interest in the preprocessed image based on the enhanced image. Furthermore, a biochip image analysis device (100), an image analysis method, a computer device (200), and a storage medium are also disclosed.

Description

Technical Field

[0001] The present application relates to the field of biological detection technology, and more specifically, to a method and device for analyzing biochip images, an image analysis method, a computer device, and a storage medium. Background Art

[0002] In the biomedical field, array-based biochip technology is a crucial tool for genetic analysis and accurate disease diagnosis. It overcomes the drawbacks of traditional methods, which require repeated experiments, effectively saves manpower and sample size, and improves detection accuracy. It is a key method for biological sample detection in modern biomedical fields. Automated detection of biochips typically involves extracting image features, autonomously determining the number of chip rows and columns, and positional information, to detect and analyze the positive and negative properties of sample points. Existing solutions generally analyze and process ideal scene images with a high signal-to-noise ratio. However, there is no effective method for processing high-throughput, low-signal-to-noise ratio images. Summary of the Invention

[0003] Embodiments of the present application provide a biochip image analysis method, a biochip image analysis device, an image analysis method, a computer device, and a storage medium.

[0004] A method for analyzing a biochip image according to an embodiment of the present application includes: acquiring a biochip image and performing preprocessing to obtain a preprocessed image; performing angle deflection correction on the preprocessed image to obtain a deflection-corrected image; performing enhancement processing on the deflection-corrected image and identifying the positivity of a region of interest in the preprocessed image based on the enhanced image.

[0005] In some embodiments, acquiring a biochip image to obtain a preprocessed image includes: acquiring an original image, a camera intrinsic parameter matrix and a distortion coefficient; and correcting the original image according to the camera intrinsic parameter matrix and the distortion coefficient to obtain the biochip image.

[0006] In certain embodiments, the method for analyzing the biochip image includes: using a calibration plate and a traditional calibration method to calibrate a camera used for shooting to obtain the camera intrinsic parameter matrix and the distortion coefficient.

[0007] In certain embodiments, the original image is a fluorescent image of a biochip where a biochemical reaction has occurred.

[0008] In some embodiments, the preprocessed image includes a high-frequency component image, and the acquiring of the biochip image to process the preprocessed image includes: performing Gaussian filtering on the biochip image to obtain a low-frequency component image; and subtracting the low-frequency component image from the biochip image to obtain the high-frequency component image.

[0009] In some embodiments, performing angle deflection correction on the preprocessed image to obtain a deflection-corrected image includes: selecting a preset number of detection areas in the preprocessed image; detecting the center and radius of the region of interest within the detection area using Hough circle transform; and drawing a circle based on the center and radius of the region of interest to determine the region of interest and segment the region of interest.

[0010] In some embodiments, selecting a preset number of detection areas in the pre-processed image includes: selecting corresponding detection areas within a predetermined area of ​​the pre-processed image.

[0011] In some embodiments, the detection area is a rectangular area, and the detection area includes at least two rows or at least two columns of parts of the regions of interest.

[0012] In some embodiments, performing angle deflection correction on the preprocessed image to obtain a deflection-corrected image includes: dilating the segmented image to connect adjacent regions of interest in a preset direction; performing principal component analysis on the maximum contour of the cavity detection area after the dilation process to obtain a contour direction; and determining the image deflection angle according to the contour direction to correct the preprocessed image to obtain the deflection-corrected image.

[0013] In some embodiments, performing angle deflection correction on the preprocessed image to obtain a deflection-corrected image includes: increasing the selected area by a preset ratio and reselecting a preset number of detection areas in the preprocessed image; and repeatedly iteratively detecting the image deflection angle until the image deflection angle is less than a preset angle threshold to obtain the deflection-corrected image.

[0014] In some embodiments, the preset angle threshold value range can be determined by the following conditional formula: Wherein, θ is the preset angle threshold, dist is the area spacing of the regions of interest, rad is the area radius of the regions of interest, m is the number of rows of the regions of interest in the detection area, and n is the number of columns of the regions of interest in the detection area.

[0015] In some embodiments, performing angle deflection correction on the pre-processed image to obtain a deflection-corrected image includes: increasing the selected area by a preset ratio to reselect a preset number of detection areas in the pre-processed image; and repeatedly iteratively detecting the image deflection angle for a preset number of times to obtain the deflection-corrected image.

[0016] In some embodiments, the step of enhancing the deflection-corrected image and identifying the positivity of the region of interest in the pre-processed image based on the enhanced image includes: constructing a notch filter; and filtering the deflection-corrected image using the notch filter to obtain a periodic pattern-enhanced image.

[0017] In certain embodiments, the enhancing processing of the deflection correction image and identifying the positivity of the region of interest in the pre-processed image based on the enhanced image includes: smoothing filtering the periodic pattern enhanced image using a box filter; integrating the pixel values ​​of the smoothed image in the horizontal and vertical directions to obtain a first integral curve in the horizontal direction and a second integral curve in the vertical direction, taking the minimum point set of the first integral curve and the second integral curve to determine the grid interval line; and dividing the grid area according to the grid interval line.

[0018] In some embodiments, the length and width of the box filter operator satisfy the following conditional formula: Wherein, b is the length and width of the box filter operator, dist is the area spacing of the region of interest, and rad is the area radius of the region of interest.

[0019] In certain embodiments, the enhancing the deflection-corrected image and identifying the positivity of the region of interest in the pre-processed image based on the enhanced image includes: traversing the grid area and calculating the mean square error of pixel values ​​for each grid area corresponding to the pre-processed image; when the mean square error is greater than a variance threshold, marking the corresponding region of interest sample as positive; and when the mean square error is not greater than a variance threshold, marking the corresponding region of interest sample as negative.

[0020] In certain embodiments, the biochip image analysis method includes: outputting a positive or negative identification result of the region of interest.

[0021] An embodiment of the present application further provides an image analysis method, comprising the steps of the biochip image analysis method described in any of the above embodiments.

[0022] An embodiment of the present application also provides a biochip image analysis device including an acquisition module, a correction module, a processing module and an identification module, wherein the acquisition module is used to acquire the biochip image and perform preprocessing to obtain a preprocessed image; the correction module is used to perform angle deflection correction on the preprocessed image to obtain a deflection-corrected image; the processing module is used to perform periodic pattern enhancement processing on the deflection-corrected image and divide the chamber grid; and enhance the deflection-corrected image and identify the positivity of the area of ​​interest in the preprocessed image based on the enhanced image.

[0023] An embodiment of the present application also provides a computer device, including a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the biochip image analysis method of any of the above embodiments, or the image analysis method described in the above embodiments.

[0024] An embodiment of the present application also provides a storage medium having a computer program stored thereon, which, when executed by one or more processors, implements the biochip image analysis method described in any of the above embodiments, or the image analysis method described in the above embodiments.

[0025] The biochip image analysis method and device, computer equipment and storage medium of the embodiments of the present application can achieve effective recognition of high-throughput, low signal-to-noise ratio matrix biochip fluorescence images. Specifically, the improved processing method solves the problem of uneven fluorescence illumination of the microscopic chip through filtering; and successfully realizes the automated analysis of chamber position positioning and sample positive and negative determination through grid division.

[0026] Additional aspects and advantages of the embodiments of the present application will be given in part in the description below, and in part will become obvious from the description below, or will be learned through practice of the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0028] Figure 1 It is a schematic flow chart of the method for analyzing biochip images according to an embodiment of the present application.

[0029] Figure 2 It is a module schematic diagram of the biochip image analysis device according to an embodiment of the present application.

[0030] Figure 3 4 is another flowchart of the method for analyzing biochip images according to an embodiment of the present application.

[0031] Figure 4 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0032] Figure 5 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0033] Figure 6 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0034] Figure 7 It is a schematic outline diagram of the divided reaction chamber after expansion processing according to an embodiment of the present application.

[0035] Figure 8 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0036] Figure 9 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0037] Figure 10 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0038] Figure 11 It is a schematic diagram of the amplitude of the biochip image converted from Fourier transform to the frequency domain according to an embodiment of the present application.

[0039] Figure 12 It is a schematic diagram of constructing a filter according to an embodiment of the present application.

[0040] Figure 13 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0041] Figure 14 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0042] Figure 15 This is another flowchart of the biochip image analysis method according to the embodiment of the present application.

[0043] Figure 16 It is a module diagram of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0044] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and should not be understood as limiting the present application.

[0045] See also Figure 1 and Figure 2 The present application provides an embodiment of a biochip image analysis method for identifying high-throughput, low-signal-to-noise ratio matrix biochip fluorescence images, thereby realizing automated analysis of chamber location and sample positive / negative determination.

[0046] The biochip image analysis method of the embodiment of the present application can be used in the biochip image analysis device 10 of the embodiment of the present application, that is, the biochip image analysis device 10 of the embodiment of the present application can use the biochip image analysis method of the embodiment to identify high-throughput, low signal-to-noise ratio matrix biochip fluorescence images, and realize automated analysis of chamber position positioning and sample positive and negative determination.

[0047] In certain embodiments, the biochip analysis comprises:

[0048] Step S1, acquiring a biochip image and performing preprocessing to obtain a preprocessed image;

[0049] Step S2, performing angle deflection correction on the pre-processed image to obtain a deflection-corrected image;

[0050] Step S3 , performing enhancement processing on the deflection-corrected image and identifying the positive and negative nature of the region of interest in the pre-processed image based on the enhanced image.

[0051] Specifically, the biochip image analysis device 10 includes an acquisition module 11, a correction module 12, and a processing module 13. Step S1 can be implemented by the acquisition module 11, step S2 can be implemented by the correction module 12, and step S3 can be implemented by the processing module 13. In other words, the acquisition module 11 can be used to acquire the biochip image to obtain a preprocessed image. The correction module 12 can be used to randomly select a detection area and detect the reaction chambers within the detection area to obtain a deflection-corrected image. The processing module 13 can be used to enhance the deflection-corrected image and identify the positive or negative of the region of interest in the preprocessed image based on the enhanced image.

[0052] The biochip image analysis method and device of the embodiment of the present application can achieve effective recognition of high-throughput, low signal-to-noise ratio matrix biochip fluorescence images. Specifically, the improved processing method solves the problem of uneven fluorescence illumination of the microscopic chip through filtering; and realizes automated analysis of chamber position positioning and sample positive and negative determination through image enhancement processing.

[0053] See also Figure 3 In some embodiments, step S1 includes:

[0054] Step S11, obtaining the original image, camera intrinsic parameter matrix and distortion coefficient; and

[0055] Step S12: Correcting the original image according to the camera intrinsic parameter matrix and the distortion coefficient to obtain a biochip image.

[0056] Specifically, step S11 and step S12 can be implemented by the acquisition module 11. That is, the acquisition module 11 can be used to acquire the original image, the camera intrinsic parameter matrix and the distortion coefficient, and to correct the original image according to the camera intrinsic parameter matrix and the distortion coefficient to obtain the biochip image.

[0057] It is understood that by obtaining the camera intrinsic parameter matrix and distortion coefficients, the distortion generated in the original image captured by the camera can be corrected, so that the corrected biochip image can more realistically display the characteristics of the biochip. This can help ensure the effectiveness and accuracy of biochip analysis.

[0058] In some examples, the biochip may be a quadrilateral, with a plurality of reaction chambers arranged in an array on the biochip. It should be noted that in the embodiments of the present application, the region where the reaction chambers are located in the biochip image is described as the region of interest.

[0059] In certain embodiments, the method for analyzing a biochip image includes: using a calibration plate and a traditional calibration method to calibrate a camera used for shooting to obtain a camera intrinsic parameter matrix and distortion coefficients.

[0060] Specifically, the biochip image analysis device 10 may include a calibration module 15 , which may be used to calibrate a camera for shooting using a calibration plate and a traditional calibration method to obtain a camera intrinsic parameter matrix and distortion coefficients.

[0061] When calibrating the camera parameters through a calibration plate, the calibration plate can have a predetermined pattern, such as a grid pattern or a black and white square pattern. The camera captures an image of the calibration plate at a certain shooting distance. In this way, the image of the calibration plate can be compared with the pattern of the calibration plate. According to the offset of the corresponding feature points in the image of the calibration plate and the pattern of the calibration plate, combined with the shooting distance, the camera intrinsic parameter matrix and distortion parameters related to the camera shooting are obtained.

[0062] It should be noted that in some embodiments, the camera intrinsic parameter matrix and distortion coefficients may be pre-calibrated and pre-stored in the camera or biochip image analysis device 10. In this way, the biochip analysis device can obtain the corresponding camera intrinsic parameter matrix and distortion coefficients from the camera, or determine the camera intrinsic parameter matrix and distortion coefficients based on the camera number or model. Of course, in other embodiments, the biochip image analysis device 10 may also detect the camera intrinsic parameter matrix and distortion coefficients corresponding to the corresponding camera before each acquisition of a biochip image. In this way, the validity of the camera intrinsic parameter matrix and distortion coefficients can be guaranteed.

[0063] In certain embodiments, the original image is a fluorescent image of a biochip where a biochemical reaction has occurred.

[0064] When a biological sample to be tested is loaded onto a biochip and a biochemical reaction occurs, a fluorescence image of the corresponding biochip can be captured using specific equipment. It is understood that in a fluorescence image, different reaction chambers may display the same or different colors and brightness.

[0065] See also Figure 4 In some embodiments, the preprocessed image includes a high-frequency component image, and step S1 includes:

[0066] Step S13, performing Gaussian filtering on the biochip image to obtain a low-frequency component image; and

[0067] Step S14: subtract the low-frequency component image from the biochip image to obtain a high-frequency component image.

[0068] Specifically, step S13 and step S14 can be implemented by the acquisition module 11. That is, the acquisition module 11 can be used to perform Gaussian filtering on the biochip image to obtain a low-frequency component image, and to subtract the low-frequency component image from the biochip image to obtain a high-frequency component image.

[0069] In this way, a low-frequency component image is obtained by using Gaussian filtering, and then the low-frequency component in the biochip image is subtracted to obtain a high-frequency component image, thereby realizing high-frequency filtering and solving the problem of uneven fluorescence illumination of the microscopic chip.

[0070] Of course, in other embodiments, the pre-processed image is not limited to the high-frequency component image discussed above, and may include a grayscale image, a low-frequency component image, an edge-detected image, and the like, as needed. The grayscale image may be obtained by grayscaling the image, the low-frequency component image may be obtained by extracting the low-frequency component, and the edge-detected image may be obtained by extracting the edge of the image. Furthermore, the pre-processed image may be obtained by processing one or more of the aforementioned processing methods in a predetermined order, without specific limitation herein.

[0071] See also Figure 5 In some embodiments, step S2 includes:

[0072] Step S21, selecting a preset number of detection areas in the pre-processed image;

[0073] Step S22, detecting the center and radius of the region of interest within the detection area using Hough circle transform; and

[0074] Step S23: Draw a circle according to the center and radius of the region of interest to determine the region of interest and segment the region of interest.

[0075] Specifically, step S21, step S22, and step S23 can be implemented by the correction module 12. That is, the correction module 12 can be used to select a preset number of detection areas in the pre-processed image, and to use the Hough circle transform to detect the center and radius of the region of interest within the detection area, and to draw a circle based on the center and radius of the region of interest to determine the region of interest and segment the region of interest.

[0076] Taking the region of interest (ROI) as the area where the reaction chambers are located in the image as an example, when detecting the arrangement of the reaction chambers, it is necessary to determine the position of the reaction chambers within the detection area. Since the reaction chambers are generally circular, the Hough transform can be used to detect the center and radius of the chamber. Furthermore, after determining the position of the reaction chamber based on the chamber center and radius, the reaction chamber can be segmented.

[0077] In some embodiments, step S21 includes: selecting a corresponding detection area within a predetermined area of ​​the pre-processed image.

[0078] The correction module 12 may be configured to select a corresponding detection area within a predetermined area of ​​the pre-processed image.

[0079] The preset area can be set by the user based on experience or automatically selected by the algorithm. Of course, the detection area can also be a randomly selected area in the pre-processed image, which is not specifically limited here.

[0080] In some embodiments, the detection area is a rectangular area, and the detection area includes at least two rows or at least two columns of partial regions of interest.

[0081] It is understood that when performing deflection correction on a high-frequency component image to obtain a deflection-corrected image, it is necessary to determine the deflection angle of the high-frequency component image. Since the reaction chambers on a biochip are generally arranged in an array, i.e., the regions of interest are generally arranged in an array, the image deflection angle can be detected by the arrangement of the chambers. Using a rectangular region facilitates determining the relative deflection angle between the long side of the selected detection region and the arrangement of the reaction chambers.

[0082] The partial regions of interest in the detection area including at least two rows or at least two columns can ensure the detection of the arrangement direction of the reaction chambers.

[0083] It should be noted that the size of the detection area can be flexibly configured according to the area spacing of the regions of interest and the radius of the regions of interest, and is not specifically limited here.

[0084] Of course, in other embodiments, the shape of the detection area may not be limited to the rectangle discussed above, but may be other suitable shapes such as square, triangle, circle, parallelogram, etc. according to actual needs, and no specific limitation is made here.

[0085] In some embodiments, the preset number of detection areas selected each time may be multiple, and the directions of the multiple detection areas may be different, thereby improving the efficiency and accuracy of image deflection angle detection. For example, the preset number of detection areas selected each time may be 9.

[0086] See also Figure 6 and Figure 7 In some embodiments, step S2 includes:

[0087] Step S24, dilating the segmented image to connect adjacent regions of interest in a preset direction;

[0088] Step S25, taking the maximum contour in the detection area after the expansion process and performing principal component analysis to obtain the contour direction; and

[0089] Step S26 , determining an image deflection angle according to the contour direction to correct the pre-processed image to obtain a deflection-corrected image.

[0090] Specifically, steps S24, S25, and S26 can be implemented by the correction module 12. That is, the correction module 12 can be used to dilate the segmented image to connect adjacent regions of interest in a preset direction, perform principal component analysis on the maximum contour within the detection area after dilation to obtain the contour direction, and determine the image deflection angle based on the contour direction to correct the pre-processed image to obtain a deflection-corrected image.

[0091] In step S24, the segmented ROIs can be expanded in a preset direction so that the outlines of the ROIs extend along the preset direction, thereby connecting the outlines of adjacent ROIs. In one example, the preset direction can be the long side direction of the rectangular detection area. Figure 7 The figure shows a schematic outline diagram obtained by expanding the regions of interest in the nine chambers according to a preset direction when there are nine detection regions.

[0092] After the contours of the regions of interest are connected, in step S25, the largest contour within the detection region is selected for PCA analysis to determine the contour direction. It will be appreciated that since the largest contour is typically formed by connecting multiple adjacent regions of interest, in one example, the resulting contour direction can be used as the arrangement direction of the reaction chambers. In particular, when there are multiple detection regions, PCA analysis can be performed on the largest contours within the multiple detection regions to determine the contour direction.

[0093] Therefore, step S25 can determine the image deflection angles of the biochip image and the preprocessed image according to the contour direction, and perform deflection angle correction on the biochip image and / or the preprocessed image to obtain a deflection-corrected image.

[0094] In this way, the present application can solve the problem of deflection angle detection by forming a maximum contour by connecting adjacent interest regions in the same direction and then using the PCA principal component analysis method.

[0095] See also Figure 8 In some embodiments, step S2 includes:

[0096] Step S27, increasing the selected area by a preset ratio and randomly selecting a preset number of detection areas in the pre-processed image; and

[0097] Step S28 , repeatedly iteratively detecting the image deflection angle until the image deflection angle is less than a preset angle threshold to obtain a deflection-corrected image.

[0098] Specifically, step S26 and step S27 can be implemented by the correction module 12. That is, the correction module 12 can be used to increase the selected area by a preset ratio and randomly reselect the detection area, and to repeatedly iteratively detect the image deflection angle until the image deflection angle is less than a preset angle threshold to obtain a deflection-corrected image.

[0099] In this way, the image deflection angle is repeatedly detected through detection areas of different sizes, thereby ensuring the accuracy of the image deflection angle.

[0100] In some embodiments, the preset angle threshold value range can be determined by the following conditional formula:

[0101]

[0102] Wherein, θ is the preset angle threshold, dist is the area spacing of the regions of interest, rad is the area radius of the regions of interest, m is the number of rows of the regions of interest in the detection area, and n is the number of columns of the regions of interest in the detection area.

[0103] It should be noted that in some embodiments, the biochip image can be precisely aligned with hardware equipment during capture. This allows the deflection-corrected image to be directly determined based on the precisely aligned biochip image, omitting the process of detecting the image deflection angle. In other embodiments, a marker can be physically set on the biochip. After capturing the biochip image, a relative coordinate system can be constructed by identifying the marker on the biochip, determining the deflection angle of the chip relative to the camera, and then correcting the deflection-corrected image.

[0104] Of course, the angle deflection correction is not limited to the implementation methods discussed above, and a suitable correction method can be selected according to actual conditions, so that the biochip image analysis device 10 can determine whether the area of ​​interest on the chip meets the horizontal or vertical relative position arrangement based on the deflection correction image. No specific limitation is made here.

[0105] See also Figure 9 In some embodiments, step S2 includes:

[0106] Step S27', increasing the selected area by a preset ratio and randomly selecting a preset number of detection areas in the pre-processed image; and

[0107] Step S28 ′: repeatedly iteratively detecting the image deflection angle for a preset number of times to obtain a deflection-corrected image.

[0108] Specifically, step S26' and step S27' can be implemented by the correction module 12. That is, the correction module 12 can be used to increase the selected area by a preset ratio and randomly reselect the detection area, and to repeatedly iterate the deflection angle of the detection image for a preset number of times to obtain a deflection-corrected image.

[0109] In this way, by repeatedly iterating the detection of the image deflection angle for a predetermined number of times using detection areas of different sizes, the accuracy of the image deflection angle can also be ensured. In one example, the predetermined number of times can be preset by the system or set by the user according to actual conditions, for example, the predetermined number of times can be 6 times.

[0110] See also Figure 10 、 Figure 11 and Figure 12 In some embodiments, step S3 includes:

[0111] Step S31, constructing a notch filter; and

[0112] Step S32 : Using a notch filter to filter the deflection-corrected image to obtain a periodic pattern-enhanced image.

[0113] Specifically, step S31 and step S32 can be implemented by the processing module 13. That is, the processing module 13 can be used to construct a notch filter, and to use the notch filter to filter the deflection correction image to obtain a periodic pattern enhanced image.

[0114] In this way, by utilizing the periodic pattern prior of the matrix-type biochip and constructing a notch filter, the noise caused by surface stains, sampling process, and reaction process is reduced to the greatest extent.

[0115] In one example, Figure 10 In the amplitude diagram obtained by Fourier transforming the biochip image into the frequency domain, most of the image information is concentrated in the low-frequency part. Therefore, by filtering out the image information in the central part, most of the image noise can be removed. At the same time, the information in the central vertical and central horizontal directions is the periodic pattern information that is most easily filtered out. In this way, the notch filter constructed in step S31 can be as follows: Figure 11 As shown, it is used to remove image noise of non-periodic patterns.

[0116] The use of a notch filter can reduce the time and space complexity of the algorithm of the biochip image analysis method of the embodiment of the present application, and make the performance requirements of the hardware equipment more relaxed, thereby reducing costs and improving operating efficiency while ensuring the effect.

[0117] See also Figure 13 In some embodiments, step S33 includes:

[0118] Step S331, using a box filter to perform smoothing filtering on the periodic pattern enhanced image;

[0119] Step S332, integrating the pixel values ​​of the smoothed image in the horizontal direction and the vertical direction to obtain a first integral curve in the horizontal direction and a second integral curve in the vertical direction, and taking the minimum value point set of the first integral curve and the second integral curve to determine the grid interval lines; and

[0120] Step S333: Divide the grid area according to the grid interval lines.

[0121] Specifically, step S331, step S332, and step S333 can be implemented by the processing module 13. That is, the processing module 13 can be used to perform smoothing filtering on the periodic pattern-enhanced image using a box filter, and to integrate the pixel values ​​of the smoothed image in the horizontal and vertical directions to obtain a first integral curve in the horizontal direction and a second integral curve in the vertical direction, to obtain a minimum point set of the first integral curve and the second integral curve to determine the grid interval lines, and to divide the grid area according to the grid interval lines.

[0122] In this way, using box filter smoothing can help reduce other noise interference images in the image and then divide the grid area.

[0123] In some embodiments, the length and width of the box filter operator satisfy the following conditional equations:

[0124]

[0125] Among them, b is the length and width of the box filter operator, dist is the area spacing of the region of interest, and rad is the area radius of the region of interest.

[0126] See also Figure 14 In other embodiments, step S3 includes:

[0127] Step S34, detecting a region of interest in the deflection-corrected image using Hough circle transform; and

[0128] Step S35: Draw a circle based on the detected region of interest to approximate the region and obtain a grid region.

[0129] Specifically, step S34 and step S35 can be implemented by the processing module 13, that is, the processing module 13 can be used to use the Hough circle transform to detect the region of interest in the deflection correction image, and to draw a circle to approximate the detected region of interest to obtain a grid area.

[0130] That is to say, for the division of the grid area, the Hough transform method can also be used to detect the interest area in the deflection correction image, and then a circle is drawn according to the detected interest area to approximate it to obtain the grid area, thereby realizing the division of the grid area.

[0131] Of course, the division of the grid area is not limited to the implementation method discussed above, and other division methods can be used as needed, which is not specifically limited here.

[0132] See also Figure 15 In some embodiments, step S4 includes:

[0133] Step S41, traversing the grid area, and calculating the mean square error of pixel values ​​in each grid area corresponding to the pre-processed image;

[0134] Step S42, when the mean square error is greater than the variance threshold, marking the corresponding region of interest sample as positive; and

[0135] Step S43: When the mean square error is not greater than the variance threshold, mark the corresponding ROI sample as negative.

[0136] Specifically, step S41, step S42, and step S43 can be implemented by the recognition module 14. That is, the recognition module 14 can be used to traverse the grid area, calculate the mean square error of the pixel values ​​in each grid area corresponding to the preprocessed image, and mark the corresponding region of interest sample as positive when the mean square error is greater than the variance threshold, and mark the corresponding region of interest sample as negative when the mean square error is not greater than the variance threshold.

[0137] In this way, the mean square error of the pixel values ​​of each grid is compared with the variance threshold to achieve the division between positive and negative.

[0138] In certain embodiments, the method for analyzing a biochip image includes outputting a positive-negative identification result of a reaction chamber.

[0139] Specifically, the biochip image analysis device 10 includes an output module (not shown), which can be used to output the positive / negative identification result of the reaction chamber.

[0140] An embodiment of the present application further provides an image analysis method, comprising the steps of the biochip image analysis method of any of the above embodiments.

[0141] The biochip image analysis method and biochip image analysis device 10 of the embodiment of the present application can achieve effective recognition of high-throughput, low signal-to-noise ratio matrix biochip fluorescence images. Specifically, the problem of uneven fluorescence illumination of the microscopic chip is solved by high-frequency filtering; the problem of deflection angle detection is solved by the PCA principal component analysis method of the maximum contour of adjacent equidirectional chambers; and the periodic pattern prior of the matrix biochip is utilized to construct a notch filter to minimize the noise caused by surface stains, sampling process, and reaction process, thereby successfully realizing the automated analysis of chamber position positioning and sample positive and negative determination.

[0142] See also Figure 16 The embodiment of the present application also provides a computer device 100, including a processor 110 and a memory 120, wherein the memory 120 stores a computer program 122, and when the computer program 122 is executed by the processor 110, the method for analyzing the biochip image of any of the above embodiments is implemented.

[0143] In one example, when the computer program 122 is executed by the processor 110, the following steps are implemented:

[0144] Step S1, acquiring a biochip image to process and obtain a pre-processed image;

[0145] Step S2, randomly selecting a detection area and detecting the reaction chambers within the detection area to obtain a deflection correction image;

[0146] Step S3 , performing enhancement processing on the deflection-corrected image and identifying the positive and negative nature of the region of interest in the pre-processed image based on the enhanced image.

[0147] The computer device 100 of the embodiment of the present application executes the computer program 122 through the processor 110, which can effectively recognize the fluorescent images of high-throughput and low-signal-to-noise ratio matrix biochips. Specifically, the improved processing method solves the problem of uneven fluorescence illumination of the microscopic chip through filtering; and successfully realizes the automated analysis of chamber position positioning and sample positive and negative determination through grid division.

[0148] An embodiment of the present application further provides a storage medium having a computer program stored thereon, which, when executed by one or more processors, implements the biochip image analysis method of any of the above embodiments.

[0149] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with an embodiment or example is included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0150] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0151] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0152] In addition, the functional units in the various embodiments of the present application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The above-mentioned storage medium can be a read-only memory, a magnetic disk or an optical disk, etc.

[0153] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for analyzing a biochip image, characterized in that: include: acquiring a biochip image and performing preprocessing to obtain a preprocessed image; performing angle deflection correction on the preprocessed image to obtain a deflection-corrected image; and performing enhancement processing on the deflection-corrected image and identifying the positive and negative nature of the region of interest in the pre-processed image based on the enhanced image; The pre-processed image includes a high-frequency component image, and the acquiring of the biochip image to obtain the pre-processed image includes: Performing Gaussian filtering on the biochip image to obtain a low-frequency component image; and The high-frequency component image is obtained by subtracting the low-frequency component image from the biochip image.

2. The method for analyzing biochip images according to claim 1, wherein: The acquiring of the biochip image to obtain the pre-processed image comprises: Get the original image, camera intrinsic parameter matrix and distortion coefficients; and The original image is corrected according to the camera intrinsic parameter matrix and the distortion coefficient to obtain the biochip image.

3. The method for analyzing biochip images according to claim 2, wherein: The biochip image analysis method includes: using a calibration plate and a traditional calibration method to calibrate a camera used for shooting, so as to obtain the camera internal parameter matrix and the distortion coefficient.

4. The method for analyzing biochip images according to claim 2, wherein: The original image is a fluorescent image of the biochip where a biochemical reaction has occurred.

5. The method for analyzing biochip images according to claim 1, wherein: The performing angle deflection correction on the pre-processed image to obtain a deflection-corrected image comprises: selecting a preset number of detection areas in the preprocessed image; Detecting the center and radius of the region of interest within the detection area using Hough circle transform; and A circle is drawn according to the center and radius of the region of interest to determine the region of interest and segment the region of interest.

6. The method for analyzing biochip images according to claim 5, characterized in that: The selecting a preset number of detection areas in the pre-processed image comprises: The corresponding detection area is selected within a predetermined area of ​​the preprocessed image.

7. The method for analyzing biochip images according to claim 5, wherein: The detection area is a rectangular area, and the detection area includes at least two rows or at least two columns of parts of the regions of interest.

8. The method for analyzing biochip images according to claim 5, wherein: The performing angle deflection correction on the pre-processed image to obtain a deflection-corrected image comprises: Performing dilation processing on the segmented image to connect adjacent regions of interest in a preset direction; Taking the maximum contour in the detection area after the expansion process and performing principal component analysis to obtain the contour direction; and An image deflection angle is determined according to the contour direction to correct the pre-processed image to obtain the deflection-corrected image.

9. The method for analyzing biochip images according to claim 8, wherein: The performing angle deflection correction on the pre-processed image to obtain a deflection-corrected image comprises: increasing the selected area by a preset ratio and reselecting a preset number of the detection areas in the pre-processed image; and The image deflection angle is detected repeatedly and iteratively until the image deflection angle is less than a preset angle threshold to obtain the deflection-corrected image.

10. The method for analyzing biochip images according to claim 9, wherein: The preset angle threshold value range is determined by the following conditional formula: ； Wherein, θ is the preset angle threshold, dist is the area spacing of the regions of interest, rad is the area radius of the regions of interest, m is the number of rows of the regions of interest in the detection area, and n is the number of columns of the regions of interest in the detection area.

11. The method for analyzing biochip images according to claim 8, wherein: The performing angle deflection correction on the pre-processed image to obtain a deflection-corrected image comprises: increasing the selected area by a preset ratio and reselecting a preset number of the detection areas in the pre-processed image; and The deflection angle of the image is repeatedly detected for a preset number of times to obtain the deflection-corrected image.

12. The method for analyzing biochip images according to claim 1, wherein: The performing enhancement processing on the deflection-corrected image and identifying the positive and negative nature of the region of interest in the pre-processed image according to the enhanced image comprises: Constructing a notch filter; and The deflection-corrected image is filtered using the notch filter to obtain a periodic pattern-enhanced image.

13. The method for analyzing biochip images according to claim 12, wherein: The performing enhancement processing on the deflection-corrected image and identifying the positive and negative nature of the region of interest in the pre-processed image according to the enhanced image comprises: Using a box filter to perform smoothing filtering on the periodic pattern enhanced image; Integrating pixel values ​​in the horizontal and vertical directions of the smoothed image to obtain a first integral curve in the horizontal direction and a second integral curve in the vertical direction, and taking a set of minimum value points of the first integral curve and the second integral curve to determine grid interval lines; and The grid areas are divided according to the grid interval lines.

14. The method for analyzing biochip images according to claim 13, wherein: The length and width of the box filter operator satisfy the following conditional formula: ； in, is the length and width of the box filter operator, dist is the area spacing of the region of interest, and rad is the area radius of the region of interest.

15. The method for analyzing biochip images according to claim 13, wherein: The performing enhancement processing on the deflection-corrected image and identifying the positive and negative nature of the region of interest in the pre-processed image according to the enhanced image comprises: Traversing the grid area, and calculating the mean square error of pixel values ​​in each grid area corresponding to the preprocessed image; When the mean square error is greater than the variance threshold, the region of interest corresponding to the mark is positive; and When the mean square error is not greater than the variance threshold, the corresponding region of interest is marked as negative.

16. The method for analyzing biochip images according to claim 1, wherein: The biochip image analysis method includes: Outputting the positive / negative identification result of the region of interest.

17. A method for analyzing an image, characterized in that: The method comprises the steps of the biochip image analysis method according to any one of claims 1 to 16.

18. A biochip image analysis device, characterized in that: include: An acquisition module, configured to acquire a biochip image and perform preprocessing to obtain a preprocessed image; a correction module, configured to perform angle deflection correction on the preprocessed image to obtain a deflection-corrected image; a processing module configured to perform enhancement processing on the deflection-corrected image and identify the positive and negative nature of the region of interest in the pre-processed image based on the enhanced image; The pre-processed image includes a high-frequency component image, and the acquiring of the biochip image to obtain the pre-processed image includes: Performing Gaussian filtering on the biochip image to obtain a low-frequency component image; and The high-frequency component image is obtained by subtracting the low-frequency component image from the biochip image.

19. A computer device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the method for analyzing the biochip image according to any one of claims 1 to 16 or the method for analyzing the image according to claim 17 is implemented.

20. A storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed by one or more processors, the method for analyzing a biochip image according to any one of claims 1 to 16 or the method for analyzing an image according to claim 17 is implemented.

Images