Method for determining fluorescence intensity value, computer device and computer-readable storage medium

Abstract

The present application provides a method for determining a fluorescence intensity value, a computer device, and a computer-readable storage medium. The method comprises: acquiring multiple fluorescence images, any single fluorescent image comprising multiple fluorescent light spots, and the multiple fluorescent light spots in one fluorescent image being generated by fluorescence imaging of bases to be detected in multiple nucleic acid sequences to be detected in a sequencing cycle; determining coordinate positions of centroids of at least a portion of the multiple fluorescent light spots in the fluorescent image on a sub-pixel point; and on the basis of a coordinate position and pixel value of each pixel point in the fluorescent light spot, calculating a pixel value of the centroid of the fluorescent light spot at the coordinate position of the sub-pixel point by using a biquadratic B-spline approximation method, and using the pixel value as a fluorescence intensity value of the fluorescent light spot.

Description

Method for determining fluorescence intensity value, computer device and computer readable storage medium

[0001] The present disclosure claims priority to a Chinese patent application filed on January 26, 2024, with application number 202410117228.2 and titled “Method for determining fluorescence intensity value, computer device and storage medium”, and a Chinese patent application filed on April 28, 2024, with application number 202410519501.4 and titled “Fluorescence intensity extraction method, gene sequencing method, device and storage medium”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] The present application relates to the field of gene sequencing technology, and in particular to a method for determining a fluorescence intensity value, a computer device, and a computer-readable storage medium. Background Art

[0003] In the field of second-generation nucleic acid sequencing, especially high-throughput sequencing, the mainstream method for identifying bases to be sequenced is to collect fluorescence images from sequencing biochips, analyze the fluorescence intensity values, and use the analyzed fluorescence intensity values ​​as input to identify the bases to be sequenced using a base caller. In this mainstream method, the quality of fluorescence intensity value extraction affects the accuracy of base identification by the base caller, and the accuracy of base identification has a significant impact on downstream genomic analysis and identification. Therefore, a fluorescence intensity value extraction method that is more suitable for base classification is particularly important in second-generation nucleic acid sequencing technology.

[0004] Currently, industrial-grade microscopy is limited by inherent technical bottlenecks, resulting in fluorescence images that fail to fully reflect texture information. Furthermore, improving the signal-to-noise ratio (SNR) of acquired images by enhancing microscopy technology is technically demanding and cost-constraining. Therefore, a high-precision method for extracting fluorescence intensity values ​​from fluorescence images is essential to address this issue, which in turn increases the accuracy requirements for these methods.

[0005] Summary of the Invention

[0006] In view of this, the present application at least provides a method for determining a fluorescence intensity value, a computer device, and a computer-readable storage medium.

[0007] In a first aspect, the present application provides a method for determining a fluorescence intensity value, comprising:

[0008] Acquire multiple fluorescent images; any of the fluorescent images contains multiple fluorescent spots, and the multiple fluorescent spots in one fluorescent image are generated by fluorescent imaging of the bases to be tested in a sequencing cycle of multiple nucleic acid sequences to be tested;

[0009] Determining coordinate positions of centroids of at least some of the multiple fluorescent spots in the fluorescent image at sub-pixel points;

[0010] According to the coordinate position and pixel value of each pixel point in the fluorescent spot, the pixel value of the center of mass of the fluorescent spot at the coordinate position of the sub-pixel point is calculated using the biquadratic B-spline approximation method, and is used as the fluorescence intensity value of the fluorescent spot.

[0011] In a possible implementation manner, the method further includes:

[0012] The type of the base to be detected that generates the fluorescent spot is identified according to the fluorescence intensity value of the fluorescent spot.

[0013] In a possible implementation, determining the coordinate positions of the centroids of at least some of the multiple fluorescent spots in the fluorescent image at sub-pixel points includes:

[0014] determining contour information of at least some of the multiple fluorescent spots based on pixel values ​​of respective pixels in the fluorescent image;

[0015] Based on the contour information of the fluorescent spot, the center of gravity of the connected domain of the fluorescent spot is determined, and the coordinate position of the center of gravity is determined as the coordinate position of the center of mass of the fluorescent spot on the sub-pixel point.

[0016] In a possible implementation, determining the contour information of at least some of the multiple fluorescent spots based on the pixel value of each pixel point in the fluorescent image includes:

[0017] performing an inverting operation on the pixel value of each pixel point of the fluorescent image to obtain an inverted fluorescent image;

[0018] performing a minimum transformation on the inverted fluorescence image to obtain a transformed fluorescence image including a plurality of minimum value pixels;

[0019] superimposing the transformed fluorescence image and the fluorescence image to obtain a target fluorescence image;

[0020] The target fluorescence image is segmented to obtain contour information of the fluorescence spot.

[0021] In a possible implementation, before segmenting the target fluorescence image to obtain the contour information of the fluorescence spot, the method further includes:

[0022] performing binarization processing on the target fluorescence image to obtain a processed target fluorescence image;

[0023] performing an opening operation on the processed target fluorescence image to obtain a noise-reduced target fluorescence image;

[0024] The step of segmenting the target fluorescent image to obtain contour information of each fluorescent spot includes:

[0025] The denoised target fluorescence image is segmented to obtain contour information of the fluorescence spot.

[0026] In a possible implementation, before determining the contour information of at least some of the multiple fluorescent spots based on the pixel values ​​of each pixel in the fluorescent image, the method further includes:

[0027] performing a top-hat transformation on the fluorescence image to obtain a first fluorescence image, and superimposing the fluorescence image with the first fluorescence image to obtain a superimposed fluorescence image;

[0028] performing a black hat transformation on the fluorescent image to obtain a second fluorescent image, and subtracting pixel values ​​at the same position in the superimposed fluorescent image from the second fluorescent image to obtain a processed fluorescent image;

[0029] The determining of the contour information of the fluorescent spot based on the pixel value of each pixel point in the fluorescent image includes:

[0030] The contour information of the fluorescent spot is determined based on the pixel value of each pixel point in the processed fluorescent image.

[0031] In a possible implementation, determining the center of gravity of the connected domain of the fluorescent spot based on the contour information of the fluorescent spot includes:

[0032] determining a minimum circumscribed circle of a connected domain of the fluorescent spot based on the contour information of the fluorescent spot;

[0033] The center of the minimum circumscribed circle is determined, and the center of the minimum circumscribed circle is determined as the center of gravity of the connected domain of the fluorescent spot.

[0034] In one possible implementation, determining the coordinate positions of the centroids of at least some of the multiple fluorescent spots in the fluorescent image at sub-pixel points includes:

[0035] A pixel point having a centroid in the fluorescent image is determined, a parabola fitting is performed on the pixel value of the pixel point having the centroid, and a coordinate position of the centroid of the fluorescent spot on the sub-pixel point is determined according to the fitting result.

[0036] In a possible implementation, determining the pixel point having a centroid in the fluorescent image includes:

[0037] Obtaining the grayscale value of each pixel in the fluorescent image;

[0038] Determine whether the grayscale value of each pixel is greater than or equal to the grayscale value of the area within a preset range from the pixel;

[0039] If yes, it is determined that the pixel point has a centroid; otherwise, it is determined that the pixel point does not have a centroid, and the current pixel point is discarded.

[0040] In one possible implementation, performing parabola fitting on the pixel values ​​of the pixel point where the centroid exists includes:

[0041] Determine the number of parabola fittings according to the domain area within the preset range;

[0042] According to the number of parabola fittings, horizontal parabola fitting and vertical parabola fitting are respectively performed on the grayscale values ​​of the pixel points where the centroid exists.

[0043] In one possible implementation, before determining the number of parabola fitting times according to the domain area within the preset range, the method further includes:

[0044] Pixels with centroids are filtered to obtain integer-level pixels.

[0045] In a possible implementation, determining the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point according to the fitting result includes:

[0046] Calculating a horizontal average value of the horizontal parabola peak points obtained multiple times, and using the horizontal average value as the horizontal coordinate of the center of mass of the fluorescent spot at the sub-pixel point;

[0047] Calculating the longitudinal average value of the peak points of the horizontal parabola and the peak points of the longitudinal parabola obtained multiple times, and using the longitudinal average value as the longitudinal coordinate of the center of mass of the fluorescent spot at the sub-pixel point;

[0048] The coordinate position of the center of mass of the fluorescent light spot on the sub-pixel point is determined based on the abscissa and the ordinate.

[0049] In a possible implementation, the domain area within the preset range includes:

[0050] 4-neighborhood, 8-neighborhood, 20-neighborhood, or 24-neighborhood.

[0051] In a possible implementation, after determining the coordinate position of the center of mass of the fluorescent spot on the sub-pixel point, the method further includes:

[0052] performing image noise reduction processing on the fluorescent spot;

[0053] The method of calculating the pixel value of the centroid of the fluorescent light spot at the coordinate position of the sub-pixel point by using a biquadratic B-spline approximation method according to the coordinate position and pixel value of each pixel point in the fluorescent light spot includes:

[0054] According to the coordinate position and pixel value of each pixel point in the fluorescent light spot after noise reduction processing, the pixel value of the centroid of the fluorescent light spot at the coordinate position on the sub-pixel point is calculated using a biquadratic B-spline approximation method.

[0055] In a possible implementation, performing image noise reduction processing on the fluorescent spot includes:

[0056] Based on the coordinate position of the centroid of the fluorescent light spot on the sub-pixel point, determining a local area to be denoised near the centroid, where the local area to be denoised covers a neighborhood area within a preset range;

[0057] Perform image noise reduction processing on the local area to be noise reduced.

[0058] In a possible implementation, determining the local area to be denoised near the centroid of the fluorescent spot based on the coordinate position of the centroid on the sub-pixel point includes:

[0059] determining, based on a coordinate position of a center of mass of the fluorescent light spot at a sub-pixel point, whether the center of mass of the fluorescent light spot is located in an edge region of the fluorescent image; wherein the edge region is determined based on size information of a convolution kernel used in the image noise reduction process;

[0060] When the centroid of the fluorescent light spot is not located in an edge region of the fluorescent image, determining a pixel point closest to the centroid of the fluorescent light spot from the fluorescent light spot;

[0061] The area around the pixel point with the closest distance is used as the local area to be denoised.

[0062] In one possible implementation, calculating the pixel value of the center of mass of the fluorescent spot at the coordinate position of the sub-pixel point using a biquadratic B-spline approximation method based on the coordinate position and fluorescence intensity value of each pixel point in the fluorescent spot includes:

[0063] Determining a first neighborhood area near the centroid of the fluorescent light spot based on the coordinate position of the centroid on the sub-pixel point;

[0064] Determining a first weight of the first neighborhood area in a horizontal dimension and a second weight in a vertical dimension;

[0065] Based on the first weight of the first neighborhood area in the horizontal dimension, the second weight in the vertical dimension, and the pixel values ​​of each pixel point in the first neighborhood area, the biquadratic B-spline approximation method is used to calculate the pixel value of the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point.

[0066] In a possible implementation, determining a first weight of the first neighborhood area in a horizontal dimension and a second weight in a vertical dimension includes:

[0067] Converting the coordinate position of the centroid at the sub-pixel point from a first coordinate system corresponding to the fluorescent image to a second coordinate system corresponding to the first neighborhood area to obtain a converted coordinate position of the centroid;

[0068] Determine a pixel point closest to the centroid in the first neighborhood area, and determine the pixel point as a central control point of the first neighborhood area; wherein the coordinate position of the central control point is located in a second coordinate system corresponding to the first neighborhood area;

[0069] Determining a first weight of the first neighborhood area in a horizontal dimension based on the horizontal coordinate value of the converted coordinate position of the centroid and the horizontal coordinate value of the coordinate position of the central control point;

[0070] Based on the ordinate value of the converted coordinate position of the centroid and the ordinate value of the coordinate position of the central control point, a second weight of the first neighborhood area in the longitudinal dimension is determined.

[0071] In one possible implementation, the method of calculating the pixel value of the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point using a biquadratic B-spline approximation method based on the first weight of the first neighborhood area in the horizontal dimension, the second weight in the vertical dimension, and the pixel value of each pixel point in the first neighborhood area includes:

[0072] Taking the central control point as the center, determining a plurality of pixels to be processed from the first neighborhood area; dividing the plurality of pixels to be processed into a plurality of pixel sets according to a first dimension;

[0073] Based on the to-be-processed pixels contained in each pixel set and the weights matching the first dimension, generating an intermediate pixel value corresponding to the pixel set;

[0074] Determining the pixel value of the coordinate position of the centroid at the sub-pixel point based on the intermediate pixel values ​​corresponding to each of the pixel point sets and the weight of the second dimension matching;

[0075] The first dimension is the horizontal dimension and the second dimension is the vertical dimension, or the first dimension is the vertical dimension and the second dimension is the horizontal dimension.

[0076] The following description of the effects of the computer device and computer-readable storage medium, etc., can be found in the description of the above method and will not be repeated here.

[0077] In a second aspect, the present application provides a computer device comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the method for determining the fluorescence intensity value as described in the first aspect or any embodiment above are performed.

[0078] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method for determining the fluorescence intensity value as described in the first aspect or any embodiment above are executed.

[0079] The embodiment of the present application provides a method for determining a fluorescence intensity value, a computer device, and a computer-readable storage medium. After acquiring multiple fluorescence images, the method can determine the coordinate position of the centroid of each fluorescent spot in the fluorescent image at a sub-pixel point. Since the sub-pixel point is determined as the centroid, the centroid selection accuracy is high, that is, the coordinate position of the centroid of the determined fluorescent spot is more accurate. Then, based on the coordinate position and pixel value of each pixel point in the fluorescent spot, the pixel value of the centroid of the fluorescent spot at the sub-pixel coordinate position can be calculated using the biquadratic B-spline approximation method. Since the pixel value of the centroid of the fluorescent spot at the sub-pixel coordinate position is high, when the pixel value at the centroid position is used as the fluorescence intensity value of the fluorescent spot, the accuracy of the fluorescence intensity value of the fluorescent spot can be improved. In addition, the use of the biquadratic B-spline approximation method to calculate the fluorescence intensity value of the fluorescent spot in the present application can ensure that the determined fluorescence intensity value is more consistent with the data distribution of the actual fluorescent spot, thereby improving the accuracy of the fluorescence intensity value.

[0080] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. The drawings herein are incorporated into and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and, together with the specification, are used to illustrate the technical solutions of the present application. It should be understood that the following drawings only illustrate certain embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings can be obtained based on these drawings without inventive effort.

[0082] FIG1 is a schematic diagram showing a flow chart of a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0083] FIG2 a shows a schematic diagram of a fluorescence image after image morphology processing in a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0084] FIG2 b shows a schematic diagram of a fluorescence image after inversion, including a plurality of minimum value pixels, in a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0085] FIG2c shows a schematic diagram of a target fluorescence image in a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0086] FIG2 d is a schematic diagram showing a segmented image including contour information of a fluorescent spot in a method for determining a fluorescence intensity value provided by an embodiment of the present application;

[0087] FIG2e shows a schematic diagram of a method for determining a fluorescence intensity value provided in an embodiment of the present application, including a centroid located at a sub-pixel point;

[0088] FIG3a shows a schematic diagram of an 8-neighborhood provided in an embodiment of the present application;

[0089] FIG3 b shows a schematic diagram of an initial neighborhood region in a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0090] FIG3 c shows a schematic diagram of a first neighborhood region in a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0091] FIG4 is a schematic diagram showing pixel positions in a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0092] FIG5 a shows a schematic diagram of a convolution kernel used in a noise reduction operation in a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0093] FIG5 b is a schematic diagram showing a convolution kernel used in another noise reduction operation in a method for determining a fluorescence intensity value provided in an embodiment of the present application;

[0094] FIG6 shows a schematic structural diagram of a computer device provided in an embodiment of the present application;

[0095] FIG7 is an overall architecture diagram of a nucleic acid sequencing system provided in an embodiment of the present application;

[0096] FIG8 is a schematic structural diagram of an optical detection system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0097] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of this application.

[0098] Since fluorescence images are collected using industrial-grade microscopic imaging technology, the fluorescence images collected by this technology cannot fully reflect the texture information of the fluorescence images. Related technologies improve the quality of fluorescence images by improving microscopic imaging technology to increase the fluorescence intensity value of the fluorescent spot in the fluorescence image. However, the above method is costly.

[0099] In another related technology, the fluorescence intensity value of the fluorescent spot is calculated by prefabricating reaction micropits and considering the relationship between the light field intensity and the fluorescence intensity. Specifically, the fluorescence image is first morphologically preprocessed to eliminate the influence of the background and the light field, and then the position of the reaction micropit is estimated by the local maximum algorithm, and the center point of the reaction micropit is determined by combining the point spread function. Finally, with the center point of the reaction micropit as the center, the average grayscale value of the surrounding specified area is selected as the fluorescence intensity value of the fluorescent spot; this method requires prefabricating reaction micropits on the sequencing chip, and considering that the physical size of the sequencing chip is small and the sequencing throughput is large, the manufacturing process requirements for the sequencing chip with prefabricated reaction micropits are high, and the manufacturing cost is high.

[0100] In another related technology, the fluorescence intensity value of the fluorescent spot can also be determined based on the template formation and bilinear interpolation method. Specifically, the template coordinates are first generated by the template formation algorithm, and then the bilinear interpolation algorithm is used on the fluorescence image that has been Laplace sharpened to determine the fluorescence intensity value of the fluorescent spot. However, this method uses the bilinear interpolation algorithm to determine the fluorescence intensity value of the fluorescent spot, which assumes that the fluorescence intensity value of the fluorescent spot changes linearly, which does not conform to the distribution law of the fluorescence intensity value of the fluorescent spot. Therefore, the accuracy of the fluorescence intensity value of the fluorescent spot obtained by this method is low.

[0101] Based on the above research, the embodiments of the present application provide a method for determining fluorescence intensity values ​​and a nucleic acid sequence detection device. By acquiring a fluorescence image containing multiple fluorescent spots and determining the coordinate position of the centroid of each fluorescent spot in the fluorescence image at a sub-pixel point, since the centroid is determined to be located at a sub-pixel point, the centroid selection accuracy is high, that is, the coordinate position of the centroid of the determined fluorescent spot is more accurate. Then, based on the coordinate position and pixel value of each pixel point in the fluorescent spot, the biquadratic B-spline approximation method can be used to calculate the pixel value of the centroid of the fluorescent spot at the coordinate position on the sub-pixel point. Since the pixel value of the centroid of the fluorescent spot at the coordinate position on the sub-pixel point is high, when the pixel value at the centroid position is used as the fluorescence intensity value of the fluorescent spot, the accuracy of the fluorescence intensity value of the fluorescent spot can be improved, without the need to improve the microscopic imaging technology or change the sequencing chip, thereby alleviating the problem of high production cost in the related technology. At the same time, compared with the related art that uses the bilinear interpolation method to calculate the fluorescence intensity value of the fluorescent spot, the present application uses the biquadratic B-spline approximation method to calculate the fluorescence intensity value of the fluorescent spot, which can ensure that the determined fluorescence intensity value is more consistent with the data distribution of the actual fluorescent spot, thereby improving the accuracy of the fluorescence intensity value.

[0102] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0103] To facilitate understanding of the embodiments of this application, we first provide a detailed introduction to a method for determining fluorescence intensity values ​​disclosed in the embodiments of this application. The method for determining fluorescence intensity values ​​provided in the embodiments of this application is generally performed by a computer device with certain computing capabilities, such as a nucleic acid sequence detection device. In some possible implementations, this method for determining fluorescence intensity values ​​can be implemented by a processor invoking computer-readable instructions stored in a memory.

[0104] 1 is a flow chart of a method for determining a fluorescence intensity value provided in an embodiment of the present application. The method includes: S101-S103, specifically:

[0105] S101, acquiring multiple fluorescence images; any fluorescence image contains multiple fluorescence spots, wherein the multiple fluorescence spots of one fluorescence image are generated by fluorescence imaging of the bases to be tested in multiple nucleic acid sequences to be tested in one sequencing cycle.

[0106] S102: Determine the coordinate positions of the centroids of at least some of the multiple fluorescent spots in the fluorescent image at sub-pixel points.

[0107] S103 , according to the coordinate position and pixel value of each pixel point in the fluorescent spot, use the biquadratic B-spline approximation method to calculate the pixel value of the centroid of the fluorescent spot at the coordinate position of the sub-pixel point, and use it as the fluorescence intensity value of the fluorescent spot.

[0108] S101 to S103 are described in detail below.

[0109] For S101 and S102:

[0110] Here, an optical image acquisition device can be used to capture images of the nucleic acid sequence to be tested, and image alignment and correction processing can be performed on at least one of the captured images to obtain an aligned image set, and each image in the aligned image set is used as a fluorescence image to be detected; wherein, any fluorescence image contains multiple fluorescent spots, and the multiple fluorescent spots in a fluorescence image are generated by fluorescence imaging of the test bases of multiple nucleic acid sequences to be tested in one sequencing cycle. The fluorescence image can be a grayscale image, that is, the pixel value of each pixel point included in the fluorescence image can be a grayscale value.

[0111] After acquiring the fluorescence image, the coordinate position of the center of mass of each fluorescent spot in the fluorescence image can be determined. Here, the center of mass is located at a sub-pixel point in the fluorescence image, that is, the center of mass is a sub-pixel point obtained indirectly based on the pixel points in the fluorescence image, and its coordinate position value is a decimal, not an integer; the coordinate position of the center of mass can be used to represent the position where the pixel value of each pixel point in the fluorescence spot corresponding to the center of mass is the largest. Therefore, the accuracy of the coordinate position of the center of mass affects the determination of the fluorescence intensity value of the fluorescence spot.

[0112] In actual operation, the coordinate position of the center of mass of the fluorescent spot can be determined by any feasible implementation method. This application does not make any specific restrictions, and the implementation is subject to the achievable method. In this application, the segmented centroid method is used to determine the coordinate position of the center of mass of each fluorescent spot in the fluorescent image at the sub-pixel point.

[0113] In one possible implementation, determining the coordinate positions of the centroids of at least some of the multiple fluorescent spots in the fluorescent image at sub-pixel points may include:

[0114] Step A1, determining contour information of at least some of the multiple fluorescent spots based on the pixel value of each pixel point in the fluorescent image;

[0115] Step A2: determining the center of gravity of the connected domain of the fluorescent spot based on the contour information of the fluorescent spot, and determining the coordinate position of the center of gravity as the coordinate position of the center of mass of the fluorescent spot on the sub-pixel point.

[0116] During implementation, the contour information of at least part of the fluorescent spot in the fluorescent image can be determined based on the pixel values ​​of each pixel point in the fluorescent image. The contour information may include the coordinate positions of the edge pixels of the fluorescent spot. For example, the contour information of the fluorescent spot can be determined using cv2.findContours() in the open source algorithm library OpenCV.

[0117] Furthermore, the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point can be determined based on the contour information of the fluorescent spot. Specifically, a detection frame of the connected domain of the fluorescent spot can be determined based on the contour information of the fluorescent spot, and the coordinate position of the center point of the detection frame, that is, the coordinate position of the center of mass of the connected domain of the fluorescent spot, is determined as the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point.

[0118] Here, the coordinate position of the center of mass of the fluorescent light spot is determined based on the contour information of the fluorescent light spot in the fluorescent image, so that the accuracy of the determined coordinate position of the center of mass can be higher.

[0119] In step A2, when determining the center of gravity of the connected domain of the fluorescent spot based on the contour information of the fluorescent spot, it can include: determining the minimum circumscribed circle of the connected domain of the fluorescent spot based on the contour information of the fluorescent spot; determining the center of the minimum circumscribed circle, and determining the center of the minimum circumscribed circle as the center of gravity of the connected domain of the fluorescent spot.

[0120] During implementation, the minimum circumscribed circle of the connected domain of the fluorescent spot can be determined based on the contour information of the fluorescent spot. In practice, the cv2.minEnclosingCircle() function in the open-source OpenCV library can be used to determine the minimum circumscribed circle of the connected domain of the fluorescent spot. Furthermore, the center of the minimum circumscribed circle can be determined, which determines the center of gravity of the connected domain of the fluorescent spot, and thus the center of mass of the fluorescent spot. The coordinate position of the center of the minimum circumscribed circle is the coordinate position of the center of gravity of the connected domain of the fluorescent spot, that is, the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point.

[0121] The following describes in detail how to determine the contour information of the fluorescent spot.

[0122] In one possible implementation, step A1, when determining the contour information of at least some of the multiple fluorescent spots based on the pixel values ​​of each pixel point in the fluorescent image, may include:

[0123] Step A11, performing an inversion operation on the pixel value of each pixel point of the fluorescence image to obtain an inverted fluorescence image;

[0124] Step A12: performing minimum value detection on the inverted fluorescent image to obtain a plurality of minimum value pixels included in the inverted fluorescent image, and generating a minimum value image based on the target coordinate positions of the plurality of minimum value pixels in the inverted fluorescent image, wherein the pixel value at the target coordinate position in the minimum value image is a first preset value, and the pixel values ​​at other coordinate positions are second preset values;

[0125] Step A13: superimpose the minimum value image and the fluorescence image to obtain a target fluorescence image; segment the target fluorescence image to obtain contour information of the fluorescence spot.

[0126] During implementation, in order to more accurately determine the contour information of the fluorescent spot, the pixel values ​​of each pixel point of the fluorescent image can be inverted to obtain the inverted fluorescent image; in actual operation, if the pixel values ​​of each pixel point of the fluorescent image are grayscale values, then for each pixel point, the inversion operation can be performed by subtracting the pixel value of the pixel point from 255 to obtain the inverted fluorescent image.

[0127] Furthermore, a minimum value detection can be performed on the inverted fluorescence image to obtain multiple minimum value pixels included in the inverted fluorescence image. Considering that the number of minimum value pixels is related to the accuracy of the determined contour information of the fluorescent spot, if the number of minimum value pixels is small, multiple fluorescent spots may correspond to one contour information, that is, the fluorescent spots are not completely segmented. If the number of minimum value pixels is large, one fluorescent spot may correspond to multiple contour information, that is, one fluorescent spot may be over-segmented into multiple contours. Therefore, in actual operation, the H-minimum transformation can be used to determine the minimum pixel points. The H-minimum transformation can control the number of minimum pixel points. The larger the minimum suppression parameter value, the fewer the number of minimum points obtained. Among them, the number of minimum transformations and the minimum suppression parameter included in the H-minimum transformation process can be determined according to actual conditions. For example, the inverted fluorescence image can be subjected to two H-minimum transformations. The minimum suppression parameter of the first H-minimum transformation process can be 18, and the minimum suppression parameter of the second H-minimum transformation process can be 16.

[0128] Then, a minimum image can be generated based on the target coordinate positions of the multiple minimum pixel points in the inverted fluorescence image; wherein the pixel value at the target coordinate position in the minimum image is a first preset value, and the pixel values ​​at the other coordinate positions are second preset values, and the first preset value is greater than the second preset value. For example, the pixel values ​​of the multiple minimum pixel points in the inverted fluorescence image can be adjusted to the first preset value, such as 255, and the pixel values ​​at the other coordinate positions can be adjusted to the second preset value, such as 0, to obtain the minimum image; alternatively, the pixel values ​​of the multiple minimum pixel points in the inverted fluorescence image can be kept unchanged, and the pixel values ​​at the other coordinate positions can be adjusted to 0 to obtain the minimum image.

[0129] After obtaining the minimum value image, the minimum value image and the fluorescence image can be superimposed to obtain a target fluorescence image, wherein the pixel points at the target coordinate positions of multiple minimum value pixels in the target fluorescence image have a high degree of recognition, so that the contour information of the fluorescence spot subsequently obtained based on the target fluorescence image is more accurate.

[0130] Finally, the target fluorescence image can be segmented to obtain the contour information of the fluorescence spot. In actual operation, the target fluorescence image can be segmented using a watershed segmenter to obtain a segmented image, wherein the segmented image includes at least part of the contour information of the fluorescence spot; the watershed segmenter can be, for example, the cv2.watershed() method in the OpenCV library.

[0131] Here, by performing minimum value detection on the fluorescence image after inverting the pixel values, the position of the fluorescent spot can be determined more accurately; then, based on the target coordinate positions of multiple minimum value pixels in the inverted fluorescence image, a minimum value image is generated, and the minimum value image and the fluorescence image are superimposed. The fluorescence spot in the obtained target fluorescence image has a higher degree of recognition, so that the target fluorescence image can be better segmented, so that the contour information of the obtained fluorescent spot is more accurate.

[0132] In actual operation, in order to better segment the target fluorescence image and more accurately obtain the contour information of the fluorescence spot, the target fluorescence image may be subjected to noise reduction processing before segmenting the target fluorescence image.

[0133] In a possible implementation, before segmenting the target fluorescence image to obtain the contour information of the fluorescent spot, the method may further include: binarizing the target fluorescence image to obtain a processed target fluorescence image; and performing an opening operation on the processed target fluorescence image to obtain a denoised target fluorescence image.

[0134] During implementation, before segmenting the target fluorescence image, the target fluorescence image can be binarized to obtain a processed target fluorescence image; and an opening operation can be performed on the processed target fluorescence image to obtain a denoised target fluorescence image. Specifically, the processed target fluorescence image can be eroded to obtain an eroded target fluorescence image, and then the eroded target fluorescence image can be dilated to obtain an expanded target fluorescence image, i.e., a denoised target fluorescence image. Here, the opening operation can be used to eliminate small particle noise points in the target fluorescence image and to disconnect multiple fluorescent spots, thereby improving segmentation efficiency.

[0135] When segmenting the target fluorescence image to obtain the contour information of the fluorescence spot, the method may include segmenting the noise-reduced target fluorescence image to obtain the contour information of the fluorescence spot.

[0136] During implementation, after obtaining the denoised target fluorescence image, the denoised target fluorescence image can be segmented to obtain contour information of at least part of the fluorescence spot. The specific process of segmenting the denoised target fluorescence image can refer to the description of the process of segmenting the target fluorescence image in step A13 above, and will not be repeated here.

[0137] Here, performing an opening operation on the target fluorescence image can effectively reduce the noise in the image, disconnect the adhesion between multiple fluorescence spots, and improve the segmentation efficiency.

[0138] During implementation, considering that there may be fluorescent spots in the fluorescent image that are difficult to identify, and there may be small particle noise points on the fluorescent image, or there may be small holes in the fluorescent spots, the fluorescent image can be preprocessed so that the pixel value of each pixel point in the fluorescent spot on the fluorescent image and the difference between the pixel value of the pixel points in the background area can be enhanced. The background area is the area in the processed fluorescent image excluding the fluorescent spot.

[0139] In a specific implementation, before determining the contour information of the fluorescent spot based on the pixel value of each pixel point in the fluorescent image, the following steps may also be included:

[0140] Step B1, performing a top-hat transformation on the fluorescence image to obtain a first fluorescence image, and superimposing the fluorescence image with the first fluorescence image to obtain a superimposed fluorescence image;

[0141] Step B2: performing black hat transformation on the fluorescence image to obtain a second fluorescence image, and subtracting the pixel values ​​at the same position in the superimposed fluorescence image from the second fluorescence image to obtain a processed fluorescence image.

[0142] During implementation, the top hat transformation process may include: performing binary processing on the fluorescent image to obtain a binary image, and performing an opening operation on the binary image, that is, performing corrosion processing on the binary image to obtain an corroded image, and then performing dilation processing on the corroded image to obtain an image after the opening operation; further, the pixel values ​​of the pixel points at the same coordinate position in the fluorescent image and the image after the opening operation can be subtracted to obtain a first fluorescent image.

[0143] The fluorescence image can then be superimposed with the first fluorescence image to obtain a superimposed fluorescence image. Specifically, the pixel values ​​of the pixels at the same coordinate positions in the fluorescence image and the first fluorescence image can be added to obtain the superimposed fluorescence image. Here, the top-hat transformation can eliminate small particle noise in the fluorescence image and break the adhesion between multiple fluorescent spots.

[0144] The fluorescence image can be subjected to a black hat transformation to obtain a second fluorescence image. Exemplarily, the black hat transformation process may include: performing binary processing on the fluorescence image, then performing a closing operation on the binary image (i.e., dilating the binary image to obtain a dilated image), then performing an erosion operation on the dilated image to obtain a closed image; and then subtracting the closed image from the pixel values ​​of the pixels at the same coordinates in the fluorescence image to generate the second fluorescence image. Here, the black hat transformation can fill small holes within the fluorescence spot and smooth the edges of each fluorescence spot.

[0145] Finally, the pixel values ​​at the same position in the superimposed fluorescence image and the second fluorescence image can be subtracted to obtain a processed fluorescence image; the processed fluorescence image has fewer small particle noise points and fewer small holes in the fluorescence spots, the edges of the fluorescence spots are clearer, and each fluorescence spot is easier to identify.

[0146] Determining the contour information of each fluorescent spot based on the pixel value of each pixel point in the fluorescent image may include: determining the contour information of at least some of the multiple fluorescent spots based on the pixel value of each pixel point in the processed fluorescent image.

[0147] During implementation, since there are fewer small particle noise points in the processed fluorescent image and each fluorescent spot is easier to identify, the accuracy of the contour information of the fluorescent spot can be improved.

[0148] Here, by performing image morphological processing on the fluorescence image, a processed fluorescence image with a more obvious contrast between the fluorescence spot and the background area can be obtained; based on the pixel value of each pixel point in the processed fluorescence image, the contour information of the fluorescence spot is determined with high accuracy.

[0149] The following specifically describes the process of determining the coordinate position of the centroid of the fluorescent spot at the sub-pixel point using the segmentation centroid method with reference to FIG. 2 a to FIG. 2 e .

[0150] After acquiring the fluorescence image, the fluorescence image can be preprocessed to obtain a processed fluorescence image as shown in Figure 2a, wherein the preprocessing process may include: performing a top hat transformation on the fluorescence image to obtain a first fluorescence image, and superimposing the fluorescence image with the first fluorescence image to obtain a superimposed fluorescence image; performing a black hat transformation on the fluorescence image to obtain a second fluorescence image, and subtracting the pixel values ​​at the same position in the superimposed fluorescence image from the second fluorescence image to obtain a processed fluorescence image.

[0151] Then, the pixel values ​​of each pixel point of the processed fluorescence image can be inverted to obtain the inverted fluorescence image as shown in Figure 2b; and the inverted fluorescence image can be subjected to minimum value detection to obtain multiple minimum value pixels included in the inverted fluorescence image, i.e., multiple black pixels in Figure 2b, and a minimum value image can be generated based on the target coordinate positions of the multiple minimum value pixels in the inverted fluorescence image. The minimum value image and the processed fluorescence image are superimposed to obtain the target fluorescence image as shown in Figure 2c.

[0152] The target fluorescence image can then be binarized to obtain a processed target fluorescence image. An opening operation can then be performed on the processed target fluorescence image to obtain a noise-reduced target fluorescence image. The noise-reduced target fluorescence image can then be segmented to obtain a segmented image, as shown in FIG2d . The segmented image FIG2d includes contour information of at least some of the multiple fluorescent spots.

[0153] Furthermore, the center of gravity of the connected domain of the fluorescent spot can be determined based on the contour information of the fluorescent spot, and the coordinate position of the center of gravity can be determined as the coordinate position of the center of mass of the fluorescent spot on the sub-pixel point. The white pixel point shown in Figure 2e is the center of mass of the fluorescent spot.

[0154] In actual operation, after obtaining the segmented image including the contour information of the fluorescent spot, the processed fluorescent image can be binarized to obtain an intermediate fluorescent image, and the intermediate fluorescent image and the pixel values ​​at the same coordinate position in the segmented image can be ANDed to obtain a segmented image including the fluorescent spot and the contour information of the fluorescent spot, so that it can be judged whether the determined contour information of the fluorescent spot is accurate based on the fluorescent spot in the segmented image.

[0155] In addition to determining the coordinate position of the center of mass of the fluorescent spot on the sub-pixel point based on the contour information of the fluorescent spot, in another embodiment, the pixel point where the center of mass exists in the fluorescent image can also be determined, and a parabola fitting is performed on the pixel value of the pixel point where the center of mass exists, and the coordinate position of the center of mass of the fluorescent spot on the sub-pixel point is determined according to the fitting result.

[0156] It should be noted that after acquiring the fluorescence image, the coordinate position of the center of mass of each fluorescent spot in the fluorescence image can be determined. Here, the center of mass is located on the sub-pixel point in the fluorescence image, that is, the center of mass is a sub-pixel point obtained indirectly based on the pixel points in the fluorescence image, and its coordinate position value is a decimal, not an integer; the coordinate position of the center of mass can be used to represent the position where the pixel value of each pixel point in the fluorescence spot corresponding to the center of mass is the largest. Therefore, the accuracy of the coordinate position of the center of mass affects the determination of the fluorescence intensity value of the fluorescence spot.

[0157] As will be explained later, the centroid of the fluorescent spot is used not only to guide local noise reduction in the image captured by the image acquisition system, but also to extract the fluorescence signal intensity value. Therefore, the sub-pixel centroid coordinates need to be determined with high accuracy.

[0158] Based on this, by determining the pixel points with the center of mass in the fluorescent image, a parabola fitting is performed on the pixel values ​​of the pixel points, thereby determining the coordinate position of the center of mass of each fluorescent spot in the fluorescent image at the sub-pixel point. The sub-pixel point is determined as the center of mass, so that the selection accuracy of the center of mass is high, that is, the coordinate position of the center of mass of the fluorescent spot is determined to be more accurate. Then, based on the coordinate position and pixel value of each pixel point in the fluorescent spot, the pixel value of the center of mass of the fluorescent spot at the sub-pixel coordinate position can be calculated using the biquadratic B-spline approximation method. Since the pixel value of the center of mass of the fluorescent spot at the sub-pixel coordinate position is higher, when the pixel value at the center of mass position is used as the fluorescence intensity value of the fluorescent spot, the accuracy of the fluorescence intensity value of the fluorescent spot can be improved. Moreover, the use of the biquadratic B-spline approximation method to calculate the fluorescence intensity value of the fluorescent spot in this application can ensure that the determined fluorescence intensity value is more consistent with the data distribution of the actual fluorescent spot, thereby improving the accuracy of the fluorescence intensity value.

[0159] In some optional embodiments, determining the pixel point having the centroid in the fluorescent image includes:

[0160] Obtaining the grayscale value of each pixel in the fluorescent image;

[0161] Determine whether the grayscale value of each pixel is greater than or equal to the grayscale value of a region within a preset range from the pixel;

[0162] If yes, it is determined that the pixel point has a centroid; otherwise, it is determined that the pixel point does not have a centroid, and the current pixel point is discarded.

[0163] In this application, the grayscale value of each pixel in the fluorescent image is first obtained, and a determination is made as to whether the grayscale value is greater than or equal to the grayscale values ​​of several surrounding neighborhoods. If so, it is considered that a centroid exists at this pixel; otherwise, it is determined that no centroid exists at this pixel. Figure 4 shows a schematic diagram of the centroid located at a sub-pixel point.

[0164] In some optional embodiments, the domain area within the preset range includes:

[0165] 4-neighborhood, 8-neighborhood, 20-neighborhood, or 24-neighborhood.

[0166] It is understandable that the aforementioned several surrounding neighborhoods include 4 neighborhoods, 8 neighborhoods, 20 neighborhoods, or 24 neighborhoods. For example, FIG3 a is a schematic diagram of 8 neighborhoods.

[0167] In some optional embodiments, performing parabola fitting on the pixel values ​​of the pixel points where the centroid exists includes:

[0168] Determine the number of parabola fittings according to the domain area within the preset range;

[0169] According to the number of parabola fittings, horizontal parabola fitting and vertical parabola fitting are respectively performed on the grayscale values ​​of the pixel points where the centroid exists.

[0170] Before determining the number of parabola fitting times according to the domain area within the preset range, the method further includes:

[0171] Pixels with centroids are filtered to obtain integer-level pixels.

[0172] In this application, parabola is used to fit the grayscale value of the integer pixel point with the centroid, and the neighborhood range is first confirmed, such as 4 neighborhoods, 8 neighborhoods, 20 neighborhoods or 24 neighborhoods. After determining the neighborhood area, the parabola fitting number can be determined according to the neighborhood range. Taking 8 neighborhoods as an example, when the grayscale value of the pixel point obtained by the initial screening is horizontally fitted, 0, 1, 2 pixel grayscale values ​​are used to fit the parabola, 3, 4, 5 pixel grayscale values ​​are used to fit the parabola, and 6, 7, 8 pixel grayscale values ​​are used to fit the parabola; When the grayscale value of the pixel point obtained by the initial screening is vertically fitted, 0, 3, 6 pixel grayscale values ​​are used to fit the parabola, 1, 4, 7 pixel grayscale values ​​are used to fit the parabola, and 2, 5, 8 pixel grayscale values ​​are used to fit the parabola. It is understandable that 4 neighborhoods, 20 neighborhoods or 24 neighborhoods also use the same method to fit horizontal parabolas and vertical parabolas.

[0173] In some optional embodiments, determining the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point according to the fitting result includes:

[0174] Calculating a horizontal average value of the horizontal parabola peak points obtained multiple times, and using the horizontal average value as the horizontal coordinate of the center of mass of the fluorescent spot at the sub-pixel point;

[0175] Calculating the longitudinal average value of the peak points of the horizontal parabola and the peak points of the longitudinal parabola obtained multiple times, and using the longitudinal average value as the longitudinal coordinate of the center of mass of the fluorescent spot at the sub-pixel point;

[0176] The coordinate position of the center of mass of the fluorescent light spot on the sub-pixel point is determined based on the abscissa and the ordinate.

[0177] It can be understood that the horizontal parabola peak points obtained multiple times are averaged, that is, the horizontal average value is obtained, and the horizontal average value is used as the horizontal coordinate of the center of mass of the fluorescent spot on the sub-pixel point. The vertical parabola peak points obtained multiple times are averaged, that is, the vertical average value, and the vertical average value is used as the vertical coordinate of the center of mass of the fluorescent spot on the sub-pixel point, thereby obtaining the coordinate position of the center of mass of the fluorescent spot on the sub-pixel point.

[0178] Specifically, the spatial distribution of the intensity value of the standard spot conforms to the Gaussian model, and conforms to the one-dimensional Gaussian model in one-dimensional space. The one-dimensional Gaussian function is as follows:

[0179] Where Loc represents the coordinates of any position on the one-dimensional model of the standard light spot; f(Loc) represents the intensity value (brightness value, grayscale value) corresponding to the position Loc on the one-dimensional model of the standard light spot; σ represents the beam waist radius of the light spot; μ represents the one-dimensional sub-pixel centroid coordinates of the fluorescent light spot in the horizontal or vertical direction; e represents the base of the natural logarithm;

[0180] For equation (1), x0 has a second-order derivative, and Taylor expansion is performed to obtain:

[0181] Where x0 represents the position coordinate of the second-order derivative on the one-dimensional model of the standard spot; R2(x) indicates that the error is a high-order infinitesimal;

[0182] When the error is R2(x), the one-dimensional Gaussian function can be replaced by a quadratic function, namely a parabola.

[0183] According to the definition of Taylor remainder, R2(x)=o((x-x0) 2 ), so the error is acceptable.

[0184] Specifically, the parabola fitting method is

[0185] Input the horizontal coordinates of three sets of points and the grayscale value of the current pixel;

[0186] Substitute it into the quadratic equation y=ax 2 +bx+c; get a set of three linear equations about the parameters, where x is the coordinate value of the pixel and y is the grayscale value of the pixel;

[0187] Solve equation (3) to find a, b, c, and then use the parabola's horizontal axis peak formula Obtain the parabola peak value. Then calculate the average value of multiple peak values ​​to obtain the horizontal and vertical coordinates of the center of mass of the fluorescent spot at the sub-pixel point.

[0188] For S103:

[0189] During implementation, the pixel value of the center of mass of the fluorescent spot at the coordinate position on the sub-pixel point can be calculated based on the coordinate position and pixel value of each pixel point in the fluorescent spot using the biquadratic B-spline approximation method, and used as the fluorescence intensity value of the fluorescent spot; specifically, the first neighborhood area corresponding to each fluorescent spot in at least part of the fluorescent spots can be determined first.

[0190] For example, the pixel closest to the center of mass of the fluorescent spot can be determined, and a neighborhood area with an 8-neighborhood range centered on this pixel can be determined. The determined neighborhood area with an 8-neighborhood range is used as the initial neighborhood area. In one approach, the initial neighborhood area can be directly determined as the first neighborhood area. In another approach, the initial neighborhood area can be extended to obtain the first neighborhood area, such as the initial neighborhood area shown in Figure 3b and the first neighborhood area shown in Figure 3c; where Z0 to Z8 in Figures 3b and 3c are pixel values. The fluorescence intensity value of the fluorescent spot can then be calculated using the biquadratic B-spline approximation method based on the coordinate positions and pixel values ​​of each pixel in the first neighborhood area.

[0191] During implementation, before calculating the fluorescence intensity value of the fluorescent spot, the fluorescence image may be subjected to noise reduction processing to improve the contrast between the fluorescent spot and the background area in the fluorescent image and reduce the influence of noise in the fluorescent image.

[0192] In a possible implementation, after determining the coordinate positions of the centroids of at least some of the multiple fluorescent spots in the fluorescent image at sub-pixel points, the method may further include: performing image noise reduction processing on the fluorescent spots.

[0193] During implementation, the fluorescence image can be subjected to image noise reduction processing to achieve the purpose of image noise reduction processing of the fluorescent spot. For example, the Laplace operation can be used to perform image noise reduction processing on all pixels of the fluorescence image to obtain a fluorescence image after noise reduction processing.

[0194] In actual operation, the fluorescent spot contained in the fluorescence image after noise reduction processing is also noise-reduced. Therefore, when determining the fluorescence intensity value of the fluorescent spot, the pixel value of the center of mass of the fluorescent spot at the coordinate position on the sub-pixel point can be calculated using the biquadratic B-spline approximation method based on the coordinate position and pixel value of each pixel point in the fluorescence spot after noise reduction processing, and the pixel value at the coordinate position on the sub-pixel point is used as the fluorescence intensity value of the fluorescent spot.

[0195] In actual operation, in order to minimize the amount of noise reduction processing calculations as much as possible, noise reduction processing may not be performed on the entire fluorescence image, but rather on the local area in the fluorescence image, especially the local area of ​​at least part of the fluorescence spot in the fluorescence image. For example, for each pixel point in the local area to be denoised, a convolution processing operation may be performed on the local area to be denoised centered on the pixel point, so that the fluorescence intensity value of the fluorescence spot after noise reduction processing can be determined subsequently.

[0196] In a possible implementation, when performing image noise reduction processing on the fluorescent spot, the following steps may be included:

[0197] Step C1 : determining a local area to be denoised near the centroid based on the coordinate position of the centroid of the fluorescent light spot on the sub-pixel point, where the local area to be denoised covers a neighborhood area within a preset range.

[0198] The neighborhood area within the preset range here may refer to the first neighborhood area, or may include the first neighborhood area and a local area around the first neighborhood area.

[0199] Step C2: performing image noise reduction processing on the local area to be noise reduced.

[0200] For example, a convolution operation may be performed on each pixel in the local area to be denoised, centered on the pixel, so that the fluorescence intensity value of the fluorescent spot after the denoising process can be determined. The local area to be denoised may cover the first neighborhood area, that is, the local area to be denoised may be the first neighborhood area, or the local area to be denoised may include the first neighborhood area and a local area surrounding the first neighborhood area.

[0201] Here, image noise reduction can be performed using the Laplace operation, which essentially performs second-order differential processing on the fluorescence image to enhance areas where grayscale changes are more rapid. This means increasing the difference between the pixel values ​​of the pixels within the fluorescent spot and the background pixels, thereby enhancing the contrast of the fluorescence image. Generally, the second-order differential of grayscale is used to represent the degree of change in areas where grayscale changes are more rapid. For a two-dimensional image f(x, y), its second-order differential can be expressed as the following formula (1):

[0202] Among them, f represents the pixel value of the pixel point, and x and y represent the coordinate position of the pixel point.

[0203] The pixel value of the sharpened pixel in the fluorescence image after noise reduction is as shown in the following formula (2): sharpened pixel = 5 × current-left-right-up-down (2)

[0204] Wherein, current, left, right, up, and down represent the pixel values ​​of the corresponding coordinate positions. The pixel positions corresponding to current, left, right, up, and down are shown in Figure 4.

[0205] According to formula (1) and formula (2), the second-order differential of the fluorescence image can be expressed as: In this case, it can be considered as using the convolution kernel shown in Figure 5a to perform a convolution operation on the pixel at the coordinate position (x, y). Alternatively, the second-order differential of the fluorescence image can also be expressed as: In this case, it can be regarded as performing a convolution operation on the pixel point at the coordinate position (x, y) using the convolution kernel shown in FIG5 b.

[0206] Wherein, f(x, y) in the above process represents the pixel value of the pixel located at the coordinate position (x, y), f(x+1, y) represents the pixel value of the pixel located at the coordinate position (x+1, y), f(x-1, y) represents the pixel value of the pixel located at the coordinate position (x-1, y), f(x, y+1) represents the pixel value of the pixel located at the coordinate position (x, y+1), f(x, y-1) represents the pixel value of the pixel located at the coordinate position (x, y-1), f(x-1, y-1) represents the pixel value of the pixel located at the coordinate position (x-1, y+1), f(x+1, y-1) represents the pixel value of the pixel located at the coordinate position (x+1, y-1), and f(x+1, y+1) represents the pixel value of the pixel located at the coordinate position (x+1, y+1).

[0207] Here, based on the coordinate position of the center of mass of the fluorescent light spot at the sub-pixel point, the local area to be denoised near the center of mass is determined, and image denoising is performed on the local area to be denoised. Compared with performing overall denoising on the fluorescent image, this method can reduce the amount of calculation and generate the denoised fluorescent image more quickly. Moreover, the contrast between the fluorescent light spot after denoising and the background area is more obvious, so that the fluorescence intensity value of the fluorescent light spot subsequently determined based on the denoised fluorescent image is more accurate.

[0208] In one possible implementation, step C1, determining a local area to be denoised near the centroid of the fluorescent light spot based on the coordinate position of the centroid at the sub-pixel point, may include:

[0209] Step C11 , determining whether the centroid of the fluorescent spot is located in the edge area of ​​the fluorescent image; wherein the edge area is determined based on the size information of the convolution kernel used in the image noise reduction process.

[0210] Step C12: if the centroid of the fluorescent spot is not located in the edge area of ​​the fluorescent image, determine the pixel point closest to the centroid of the fluorescent spot from the fluorescent spot.

[0211] In step C13, the area around the nearest pixel is used as the local area to be denoised.

[0212] During implementation, it is possible to determine whether the center of mass of the fluorescent spot is in the edge area of ​​the fluorescent image based on the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point, where the edge area is determined based on the size information of the convolution kernel used in the image noise reduction processing. For example, if the size information of the convolution kernel is 3×3 pixels, the edge area can be obtained by extending 1.5 pixels inward from the edge of the fluorescent image; if the size information of the convolution kernel is 5×5 pixels, the edge area can be obtained by extending 2.5 pixels inward from the edge of the fluorescent image; and so on.

[0213] If the centroid of the fluorescent spot is not located at the edge of the fluorescent image, the pixel closest to the centroid of the fluorescent spot can be determined from the fluorescent spot. For example, the sub-pixel coordinate position of the centroid of the fluorescent spot can be rounded to the nearest integer, and the pixel closest to the centroid of the fluorescent spot can be determined from the fluorescent spot. Alternatively, the sub-pixel coordinate position of the centroid of the fluorescent spot can be rounded to the nearest integer, and the pixel closest to the centroid of the fluorescent spot can be determined from the fluorescent spot. In actual operation, if the pixel closest to the centroid of the fluorescent spot is located in the background area, the pixel closest to the centroid of the fluorescent spot can be re-determined.

[0214] Furthermore, the area around the nearest pixel point can be used as the local area to be denoised; for example, the second neighborhood area centered on the nearest pixel point in the fluorescence image can be determined as the local area to be denoised, wherein the area size of the second neighborhood area is greater than or equal to the area size of the first neighborhood area, so that the determined local area to be denoised can cover the first area, for example, the first neighborhood area is 8 neighborhoods, the second neighborhood area can be 8 neighborhoods, or the second neighborhood area can be 15 neighborhoods.

[0215] Taking into account that when the centroid of the fluorescent spot is in the edge area of ​​the fluorescent image, convolution out of bounds may occur when the fluorescent spot is subjected to image denoising. Therefore, when the centroid of the fluorescent spot is not in the edge area of ​​the fluorescent image, the pixel point closest to the centroid of the fluorescent spot is determined from the fluorescent spot, and the area around the closest pixel point is used as the local area to be denoised. This can ensure that convolution out of bounds will not occur when the fluorescent spot is subjected to image denoising. At the same time, since the closest pixel point is a pixel point, the pixel value of each pixel point in the local area to be denoised can be directly obtained, thereby improving the efficiency of image denoising.

[0216] In a specific implementation, after performing image noise reduction processing on the fluorescence image, the pixel value of the coordinate position of the center of mass of the fluorescence spot at the sub-pixel point can be calculated using the biquadratic B-spline approximation method.

[0217] In one possible implementation, when calculating the pixel value of the centroid of the fluorescent spot at the sub-pixel coordinate position based on the coordinate position and fluorescence intensity value of each pixel in the fluorescent spot using a biquadratic B-spline approximation method, the following steps may be included:

[0218] Step D1, determining a first neighborhood area near the centroid based on the coordinate position of the centroid of the fluorescent spot on the sub-pixel point;

[0219] Step D2, determining a first weight of the first neighborhood area in the horizontal dimension and a second weight in the vertical dimension;

[0220] Step D3, based on the first weight of the first neighborhood area in the horizontal dimension, the second weight in the vertical dimension, and the pixel value of each pixel point in the first neighborhood area, use the biquadratic B-spline approximation method to calculate the pixel value of the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point.

[0221] Specifically, the pixel point closest to the center of mass of the fluorescent spot can be determined, and the neighborhood area with an 8-neighborhood range centered on the pixel point can be determined as the initial neighborhood area. In one method, the initial neighborhood area can be directly determined as the first neighborhood area, and in another method, the initial neighborhood area can be extended to obtain the first neighborhood area.

[0222] Then, the first weight of the first neighborhood area in the horizontal dimension and the second weight in the vertical dimension can be determined. Specifically, the pixel point located at the center of the first neighborhood area can be determined, and the first weight of the first neighborhood area in the horizontal dimension can be determined based on the horizontal distance difference between the pixel point and the centroid. And the second weight of the first neighborhood area in the vertical dimension can be determined based on the vertical distance difference between the pixel point and the centroid. The weights of the first neighborhood area in different dimensions represent the degree of influence of the pixel values ​​of each pixel point in the first neighborhood area on the pixel value of the centroid coordinate position at the sub-pixel point.

[0223] Furthermore, the pixel value of the coordinate position of the centroid of the fluorescent light spot at the sub-pixel point can be calculated based on the first weight of the first neighborhood area in the horizontal dimension, the second weight in the vertical dimension, and the pixel value of each pixel point in the first neighborhood area using the biquadratic B-spline approximation method. For example, based on the first weight of the first neighborhood area in the horizontal dimension and the pixel value of each pixel point in the first neighborhood area, an interpolation operation can be performed from the horizontal dimension to obtain an intermediate pixel value; then, based on the second weight of the first neighborhood area in the vertical dimension and the obtained intermediate pixel value, an interpolation operation can be performed from the vertical dimension to obtain the pixel value of the coordinate position of the centroid at the sub-pixel point. Alternatively, based on the second weight of the first neighborhood area in the vertical dimension and the pixel value of each pixel point in the first neighborhood area, an interpolation operation can be performed from the vertical dimension to obtain an intermediate pixel value; then, based on the first weight of the first neighborhood area in the horizontal dimension and the obtained intermediate pixel value, an interpolation operation can be performed from the horizontal dimension to obtain the pixel value of the coordinate position of the centroid at the sub-pixel point.

[0224] Here, the determined weights of the first neighborhood area in different dimensions can characterize the degree of influence of the pixel values ​​of each pixel point in the first neighborhood area on the pixel value of the coordinate position of the centroid at the sub-pixel point. Therefore, based on the first weight of the first neighborhood area in the horizontal dimension, the second weight in the vertical dimension, and the pixel values ​​of each pixel point in the first neighborhood area, the biquadratic B-spline approximation method can be used to more accurately calculate the pixel value of the coordinate position of the centroid of the fluorescent spot at the sub-pixel point, thereby improving the accuracy of the fluorescence intensity value.

[0225] In one possible implementation, step D2, when determining the first weight of the first neighborhood area corresponding to the centroid in the horizontal dimension and the second weight in the vertical dimension, may include:

[0226] In step D21 , the coordinate position of the centroid at the sub-pixel point is converted from a first coordinate system corresponding to the fluorescent image to a second coordinate system corresponding to the first neighborhood area to obtain the converted coordinate position of the centroid.

[0227] Step D22: determining the pixel point closest to the centroid in the first neighborhood area, and determining the pixel point as the central control point of the first neighborhood area; wherein the coordinate position of the central control point is located in the second coordinate system corresponding to the first neighborhood area;

[0228] Step D23: Determine a first weight of the first neighborhood area in the horizontal dimension based on the horizontal coordinate value indicated by the converted coordinate position of the centroid and the horizontal coordinate value indicated by the coordinate position of the central control point.

[0229] Step D24, determining a second weight of the first neighborhood area in the longitudinal dimension based on the longitudinal coordinate value indicated by the converted coordinate position of the centroid point and the longitudinal coordinate value indicated by the coordinate position of the central control point.

[0230] During implementation, the coordinate position of the centroid at the sub-pixel point can be converted from the first coordinate system corresponding to the fluorescent image to the second coordinate system corresponding to the first neighborhood region to obtain the converted coordinate position of the centroid. Specifically, the converted coordinate position of the centroid can be obtained using x-floor(x)+2. That is, if the coordinate position of the centroid in the first coordinate system corresponding to the fluorescent image is (64.78, 125.32), then the converted coordinate position of the centroid is (2.78, 2.32).

[0231] Then, the pixel point closest to the centroid can be determined from the first neighborhood area, and the pixel point can be determined as the central control point of the first neighborhood area; wherein the coordinate position of the central control point is in the second coordinate system corresponding to the first neighborhood area; for example, the converted coordinate position of the centroid can be rounded up. If the converted coordinate position of the centroid is (2.78, 2.32), the coordinate position of the central control point can be (3, 2), that is, floor(2.78+0.5)=3, floor(2.32+0.5)=2; wherein floor(g) represents rounding down g.

[0232] Furthermore, the first weight u of the first neighborhood area in the horizontal dimension can be determined based on the horizontal coordinate value x in the transformed coordinate position of the centroid and the horizontal coordinate value i in the coordinate position of the central control point, as shown in the following formula (3):

[0233] Furthermore, the second weight v of the first neighborhood area in the vertical dimension can be determined based on the ordinate value y in the converted coordinate position of the centroid and the ordinate value j in the coordinate position of the central control point, as shown in the following formula (4):

[0234] Here, the coordinate position of the center of mass at the sub-pixel point is converted from the first coordinate system corresponding to the fluorescent image to the second coordinate system corresponding to the first neighborhood area, so as to determine the central control point of the first neighborhood area and the coordinate position of the central control point based on the converted coordinate position of the center of mass; and the first weight of the first neighborhood area in the horizontal dimension and the second weight in the vertical dimension can be determined based on the converted coordinate position of the center of mass and the coordinate position of the central control point, so as to provide data support for the subsequent determination of the fluorescence intensity value of the fluorescent spot.

[0235] In one possible implementation, step D3, when calculating the pixel value of the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point using a biquadratic B-spline approximation method based on the first weight of the first neighborhood area in the horizontal dimension, the second weight in the vertical dimension, and the pixel value of each pixel point in the first neighborhood area, may include:

[0236] In step D31 , a plurality of pixels to be processed are determined from a first neighborhood area with the central control point as the center; and the plurality of pixels to be processed are divided into a plurality of pixel sets according to a first dimension.

[0237] Step D32: Based on the pixels to be processed contained in each pixel set and the weights matched with the first dimension, an intermediate pixel value corresponding to the pixel set is generated.

[0238] Step D33, based on the intermediate pixel values ​​corresponding to each pixel point set and the weight of the second dimension matching, determine the pixel value of the coordinate position of the center of mass at the sub-pixel point on the fluorescence image; wherein the first dimension is the horizontal dimension and the second dimension is the vertical dimension, or the first dimension is the vertical dimension and the second dimension is the horizontal dimension.

[0239] During implementation, multiple pixel points to be processed can be determined from the first neighborhood area with the central control point as the center; continuing with the above example, the coordinate position of the central control point is (3, 2), then the multiple pixel points to be processed determined from the first neighborhood area may include a pixel point located at (2, 1), a pixel point located at (3, 1), a pixel point located at (4, 1), a pixel point located at (2, 2), a pixel point located at (3, 2), a pixel point located at (4, 2), a pixel point located at (2, 3), a pixel point located at (3, 3), and a pixel point located at (4, 3).

[0240] The plurality of pixels to be processed can then be divided into a plurality of pixel sets according to the first dimension. Taking the first dimension as the horizontal dimension and the second dimension as the vertical dimension as an example, pixel set 1 can include a pixel located at (2, 1), a pixel located at (3, 1), and a pixel located at (4, 1). Pixel set 2 includes a pixel located at (2, 2), a pixel located at (3, 2), and a pixel located at (4, 2). Pixel set 3 includes a pixel located at (2, 3), a pixel located at (3, 3), and a pixel located at (4, 3). Furthermore, based on the pixels to be processed contained in each pixel set and the weights matched with the first dimension, an intermediate pixel value corresponding to the pixel set can be generated. The weight matched with the first dimension is the first weight u. The formula for generating the intermediate pixel value corresponding to the pixel set can be shown as the following formula (5):

[0241] Among them, i and j are the coordinate positions of the central control point; Z i,j is the pixel value of the pixel point at the coordinate position (i, j) in the first neighborhood area; k represents the sequence number of the pixel point set, for example, if the pixel point set is pixel point set 1, then k=1; L(k) is the middle pixel value corresponding to pixel point set k.

[0242] Furthermore, the pixel value of the coordinate position of the center of mass of the fluorescent spot at the sub-pixel point can be determined based on the intermediate pixel values ​​corresponding to each pixel point set and the weight of the second dimension matching; the weight of the second dimension matching is the second weight v, and the formula for determining the pixel value of the coordinate position of the center of mass at the sub-pixel point is shown in the following formula (6):

[0243] Wherein, L(k) is the middle pixel value corresponding to the pixel point set k, and k in formula (6) can be 1, 2, or 3; Q is the pixel value of the coordinate position of the centroid on the sub-pixel point.

[0244] Taking the first dimension as the vertical dimension and the second dimension as the horizontal dimension as an example, pixel point set 1 may include pixel points located at (2, 1), pixel points located at (2, 2), and pixel points located at (2, 3), pixel point set 2 includes pixel points located at (3, 1), pixel points located at (3, 2), and pixel points located at (3, 3), and pixel point set 3 includes pixel points located at (4, 1), pixel points located at (4, 2), and pixel points located at (4, 3).

[0245] The weight of the first dimension matching is the second weight v, and the formula for generating the intermediate pixel value corresponding to the pixel point set can be shown as the following formula (7):

[0246] The weight of the second dimension matching is the first weight u. The formula for determining the pixel value of the coordinate position of the centroid on the sub-pixel point can be shown as the following formula (8):

[0247] In the embodiment of the present application, the quadratic B-spline approximation method is used to determine the pixel value of the coordinate position of the centroid at the sub-pixel point through the known pixel values ​​of each pixel point in the first neighborhood area.

[0248] In a specific implementation, after the fluorescence intensity value of a fluorescent spot in a fluorescent image is determined, the type of the base to be detected that generates the fluorescent spot can be identified according to the fluorescence intensity value of the fluorescent spot.

[0249] Specifically, the fluorescence intensity value of the fluorescent spot can be input into the base detector, and then the type of the base to be detected corresponding to the fluorescent spot can be obtained.

[0250] Those skilled in the art will understand that in the above-mentioned method of the specific implementation method, the writing order of each step does not mean a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0251] Based on the same technical concept, an embodiment of the present application further provides a computer device. The computer device may be located in a nucleic acid sequencing device, or the computer device may be located in a server that is in wireless communication with the nucleic acid sequencing device. In actual operation, if the computer device is located in a server that is in wireless communication with the nucleic acid sequencing device, the nucleic acid sequencing device may, after acquiring the fluorescence image, send the fluorescence image to the server via wireless communication. The computer device located in the server may then perform image processing on the fluorescence image according to the above-mentioned method for determining the fluorescence intensity value to obtain the fluorescence intensity value of the fluorescent spot in the fluorescence image.

[0252] 6 shows a schematic diagram of the structure of a computer device provided in an embodiment of the present application, including a processor 601, a memory 602, and a bus 603. The memory 602 is used to store execution instructions and includes a memory 6021 and an external memory 6022. The memory 6021 is also referred to as internal memory and is used to temporarily store operation data in the processor 601 and data exchanged with an external memory 6022 such as a hard disk. The processor 601 exchanges data with the external memory 6022 through the memory 6021. When the computer device is running, the processor 601 communicates with the memory 602 via the bus 603, so that the processor 601 executes the following instructions:

[0253] Acquire multiple fluorescence images; any of the fluorescence images contains multiple fluorescence spots, wherein the multiple fluorescence spots in one fluorescence image are generated by fluorescence imaging of the bases to be tested in a sequencing cycle of multiple nucleic acid sequences to be tested;

[0254] Determining coordinate positions of centroids of at least some of the multiple fluorescent spots in the fluorescent image at sub-pixel points;

[0255] According to the coordinate position and pixel value of each pixel point in the fluorescent spot, the pixel value of the center of mass of the fluorescent spot at the coordinate position of the sub-pixel point is calculated using the biquadratic B-spline approximation method, and is used as the fluorescence intensity value of the fluorescent spot.

[0256] The specific processing flow of the processor 601 can refer to the description of the above method embodiment and will not be repeated here.

[0257] In addition, embodiments of the present application further provide a computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program executes the steps of the method for determining the fluorescence intensity value described in the above method embodiment. The storage medium may be a volatile or non-volatile computer-readable storage medium.

[0258] An embodiment of the present application also provides a computer program product, which carries a program code. The instructions included in the program code can be used to execute the steps of the method for determining the fluorescence intensity value described in the above method embodiment. For details, please refer to the above method embodiment and will not be repeated here.

[0259] The computer program product may be implemented in hardware, software, or a combination thereof. In one embodiment, the computer program product is implemented as a computer storage medium. In another embodiment, the computer program product is implemented as a software product, such as a software development kit (SDK).

[0260] The above method provided in this application can be applied to a nucleic acid sequencing system, for example, it can be used as a method executed by a computer system in a nucleic acid sequencing system, and the computer system can refer to the above computer device, or a system that carries the above computer-readable storage medium or computer program product.

[0261] Please refer to Figure 7, which shows the overall architecture of a nucleic acid sequencing system according to an embodiment of the present disclosure. As shown in Figure 7, the nucleic acid sequencing system 100 according to an embodiment of the present disclosure includes: a chip 10, a chip platform 20, a reagent storage container 30, a fluidics system 40, an optical detection system 50, and a computer system 60.

[0262] Wherein, one or more sequencing objects are attached to the chip 10;

[0263] The sequencing object here can be a nucleic acid fragment, which includes a base sequence of a certain length, for example, 150 bp in length; the sequencing object can also be a nucleic acid molecule.

[0264] The chip platform 20 is configured to fix and support the chip 10;

[0265] The reagent storage container 30 is configured to store one or more reagents; here, the reagents may exemplarily include polymerase chain reaction (PCR) fluorescent reagents.

[0266] The fluid guiding system 40 is configured to controllably transport the one or more reagents from the reagent storage container 30 to the chip 10 so as to contact and chemically react with the sequencing object, thereby causing the sequencing object to be fluorescently labeled;

[0267] The optical detection system 50 is configured to excite the fluorescent marker on the sequencing object and detect the fluorescent signal generated by the excitation of the fluorescent marker;

[0268] The computer system 60 is configured to obtain the fluorescence image from the optical detection system 50 and identify the nucleic acid sequence of the sequencing object based on the fluorescence image.

[0269] Referring to FIG. 8 , FIG. 8 is a schematic structural diagram of an optical detection system provided in an embodiment of the present disclosure. The optical detection system 50 at least includes a light source component 501 and an imaging assembly 502 .

[0270] The imaging assembly 502 includes at least an aperture plate 5021, a dichroic mirror 5022-1, a microscope 5023, and an image sensor 5024. The aperture plate 5021 is provided with a plurality of light holes 5025. The image sensor 5024 can be exemplarily an industrial camera.

[0271] The light source component 501 emits scattered excitation light, which passes through at least some of the multiple light holes on the aperture plate 5021 to form multiple laser beams and irradiate the dichroic mirror 5022-1; the dichroic mirror 5022-1 reflects the excitation light beams passing through the light holes to the chip 10 to assist the microscope 5023 in focusing; the microscope 5023 collects the fluorescence signals generated by the sequencing objects on the chip 10 after a chemical reaction; the image sensor 5024 senses the fluorescence signals and generates a fluorescence image.

[0272] The dichroic mirror 5022 - 1 can be fixed at a certain angle, for example, by dispensing glue.

[0273] In one possible embodiment, the light source component 501 may be, for example, a light emitting diode (LED) or a semiconductor laser (LD). The aperture piece 5021 may be made of an opaque material, for example, a metal sheet. The scattered excitation light emitted by the light source component 501 can only pass through the light-through hole portion, while the non-light-through hole portion of the aperture piece 5021 will block the excitation light.

[0274] In one possible embodiment, the imaging assembly 502 further includes an attenuation plate 5026, which is disposed parallel to the aperture plate 5021 and configured to attenuate the intensity of the excitation light beam. Referring to FIG2 , the attenuation plate 5026 can be disposed behind the aperture plate 5021, so that the excitation light beam, after passing through the light aperture, can directly illuminate the attenuation plate 5026.

[0275] In a possible embodiment, the imaging assembly 502 further includes a convex lens 5027, which is arranged parallel to the attenuation plate 5026 and located between the attenuation plate 5026 and the dichroic mirror 5022-1, and is configured to collimate the excitation light beam into a parallel beam.

[0276] The microscope 5023 may include an objective lens 50231 and a tube lens 50232 . The objective lens 50231 is located below the dichroic mirror 5022 - 1 , the tube lens 50232 is located above the dichroic mirror 5022 - 1 , and the image sensor 5024 is located above the tube lens 50232 .

[0277] A motor is configured on one side of the objective lens 50231, for example, a voice coil motor. The motor on the objective lens 50231 can be used to control the objective lens 50231 to move up and down to achieve focusing.

[0278] During sequencing, the image sensor 5024 (e.g., an industrial camera) in the imaging component 502 needs to image the fluorescence signals of different areas of the chip 10 through relative displacement with the chip 10 in different sequencing cycles to obtain a fluorescence image. In one sequencing cycle, multiple sequencing objects on the chip 10 are excited by the fluorescent markers they carry and generate corresponding fluorescence signals. The fluorescence signal generated by each sequencing object is presented in the form of a fluorescent spot in the fluorescence image, so a fluorescent image contains multiple fluorescent spots. After extracting the fluorescence intensity value of the fluorescent spot, the base type of the sequencing object that generated the fluorescent spot can be identified based on its fluorescence intensity value. Whether the fluorescent spot can be accurately identified and the fluorescence intensity value of the fluorescent spot can be accurately extracted is crucial to whether the base type can be accurately identified.

[0279] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system and device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. In the several embodiments provided in this application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0280] The units described as separate components may or may not be physically separate, and 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 these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0281] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0282] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0283] The above are only specific embodiments of the present application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A computer device, characterized in that, Comprising: A processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device runs, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the following method for determining fluorescence intensity values is performed: Obtain a fluorescence image; the fluorescence image contains a plurality of fluorescence spots, wherein the plurality of fluorescence spots in one fluorescence image are generated by fluorescence imaging of the bases to be measured of a plurality of nucleic acid sequences to be measured in one sequencing cycle; Determine the coordinate positions of the centroids of at least some of the fluorescence spots among the plurality of fluorescence spots in the fluorescence image on sub-pixel points; According to the coordinate positions and pixel values of each pixel point in the fluorescence spot, use the biquadratic B-spline approximation method to calculate the pixel value at the coordinate position of the centroid of the fluorescence spot on the sub-pixel point, and use it as the fluorescence intensity value of the fluorescence spot.

2. The computer device according to claim 1, wherein When the machine-readable instructions are executed by the processor, the following method is also performed: Identify the type of the base to be measured that generates the fluorescence spot according to the fluorescence intensity value of the fluorescence spot.

3. The computer device according to claim 1, characterized in that, In the method executed by the processor, the determining the coordinate positions of the centroids of at least some of the fluorescence spots among the plurality of fluorescence spots in the fluorescence image on sub-pixel points includes: Based on the pixel values of each pixel point in the fluorescence image, determine the contour information of at least some of the fluorescence spots among the plurality of fluorescence spots; Based on the contour information of the fluorescence spot, determine the centroid of the connected domain of the fluorescence spot, and determine the coordinate position of the centroid as the coordinate position of the centroid of the fluorescence spot on the sub-pixel point.

4. The computer device according to claim 3, characterized in that, In the method executed by the processor, the based on the pixel values of each pixel point in the fluorescence image, determining the contour information of at least some of the fluorescence spots among the plurality of fluorescence spots includes: Perform an inversion operation on the pixel values of each pixel point in the fluorescence image to obtain an inverted fluorescence image; Perform a minimum value detection on the inverted fluorescence image to obtain a plurality of minimum value pixel points included in the inverted fluorescence image, and generate a minimum value image based on the target coordinate positions of the plurality of minimum value pixel points in the inverted fluorescence image. The pixel value at the target coordinate position in the minimum value image is a first preset value, and the pixel values at other coordinate positions are second preset values; Overlay the minimum value image and the fluorescence image to obtain a target fluorescence image; segment the target fluorescence image to obtain the contour information of the fluorescence spot.

5. The computer device according to claim 4, wherein, In the method executed by the processor, before segmenting the target fluorescence image to obtain the contour information of the fluorescence spot, it further includes: Perform a binarization process on the target fluorescence image to obtain a processed target fluorescence image; Perform an opening operation on the processed target fluorescence image to obtain a denoised target fluorescence image; The segmenting the target fluorescence image to obtain the contour information of the fluorescence spot includes: Segment the denoised target fluorescence image to obtain the contour information of the fluorescence spot.

6. The computer device according to any one of claims 3 to 5, characterized in that In the method executed by the processor, before determining the contour information of the fluorescence spot based on the pixel values of each pixel point in the fluorescence image, it further includes: Performing a top-hat transform on the fluorescence image to obtain a first fluorescence image, and superimposing the fluorescence image and the first fluorescence image to obtain a superimposed fluorescence image; Performing a black-hat transform on the fluorescence image to obtain a second fluorescence image, and subtracting the pixel values at the same positions in the superimposed fluorescence image and the second fluorescence image to obtain a processed fluorescence image; The determining the contour information of the fluorescence spot based on the pixel values of each pixel point in the fluorescence image includes: Determining the contour information of the fluorescence spot based on the pixel values of each pixel point in the processed fluorescence image.

7. The computer device according to any one of claims 3 to 5, characterized in that In the method executed by the processor, the determining the centroid of the connected region of the fluorescence spot based on the contour information of the fluorescence spot includes: Determining the minimum circumscribed circle of the connected region of the fluorescence spot based on the contour information of the fluorescence spot; Determining the center of the minimum circumscribed circle, and determining the center of the minimum circumscribed circle as the centroid of the connected region of the fluorescence spot.

8. The computer device according to claim 1, wherein In the method executed by the processor, determining the coordinate positions of the centroids of at least some of the multiple fluorescence spots in the fluorescence image on sub-pixel points includes: Determining the pixel points with centroids in the fluorescence image, performing parabolic fitting on the pixel values of the pixel points with centroids, and determining the coordinate positions of the centroids of the fluorescence spots on sub-pixel points according to the fitting results.

9. The computer device according to claim 8, characterized in that, In the method executed by the processor, the determining the pixel points with centroids in the fluorescence image includes: Obtaining the gray values of each pixel point in the fluorescence image; Judging whether the gray value of each pixel point is greater than or equal to the gray value of the neighborhood region within a preset range from the pixel point; If so, determining that the pixel point has a centroid, otherwise determining that the pixel point does not have a centroid and discarding the current pixel point.

10. The computer device according to claim 8, characterized in that, In the method executed by the processor, performing parabolic fitting on the pixel values of the pixel points with centroids includes: Determining the number of times of parabolic fitting according to the neighborhood region within a preset range; Performing horizontal parabolic fitting and vertical parabolic fitting on the gray values of the pixel points with centroids respectively according to the number of times of parabolic fitting.

11. The computer device according to claim 10, characterized in that, In the method executed by the processor, before determining the number of times of parabolic fitting according to the neighborhood region within a preset range, it further includes: Screening the pixel points with centroids to obtain integer-level pixel points.

12. The computer device according to claim 10 or 11, characterized in that, In the method executed by the processor, determining the coordinate positions of the centroids of the fluorescence spots on sub-pixel points according to the fitting results includes: Calculating the horizontal average value of the horizontal parabolic peak points obtained multiple times, and taking the horizontal average value as the abscissa of the centroid of the fluorescence spot on the sub-pixel point; Calculating the vertical average value of the horizontal parabolic peak points and the peak points of the vertical parabola obtained multiple times, and taking the vertical average value as the ordinate of the centroid of the fluorescence spot on the sub-pixel point; Determining the coordinate positions of the centroids of the fluorescence spots on the sub-pixel point based on the abscissa and ordinate.

13. The computer device according to any one of claims 9 to 11, characterized in that, The domain area within the preset range includes: 4-neighborhood, 8-neighborhood, 20-neighborhood or 24-neighborhood.

14. The computer device according to claim 1, wherein In the method executed by the processor, after determining the coordinate positions of at least some of the centroids of the plurality of fluorescent spots in the fluorescent image on sub-pixel points, it further includes: Performing image noise reduction processing on the fluorescent spots; The calculating the pixel value at the coordinate position of the centroid of the fluorescent spot on the sub-pixel point by using the biquadratic B-spline approximation method according to the coordinate positions and pixel values of each pixel point in the fluorescent spot includes: Calculating the pixel value at the coordinate position of the centroid of the fluorescent spot on the sub-pixel point by using the biquadratic B-spline approximation method according to the coordinate positions and pixel values of each pixel point in the fluorescent spot after noise reduction processing.

15. The computer device according to claim 14, characterized in that, In the method executed by the processor, the performing image noise reduction processing on the fluorescent spots includes: Based on the coordinate position of the centroid of the fluorescent spot on the sub-pixel point, determining a local area to be noise-reduced near the centroid, and the local area to be noise-reduced covers the neighborhood area within the preset range; Performing image noise reduction processing on the local area to be noise-reduced.

16. The computer device according to claim 15, wherein In the method executed by the processor, the determining the local area to be noise-reduced near the centroid based on the coordinate position of the centroid of the fluorescent spot on the sub-pixel point includes: Based on the coordinate position of the centroid of the fluorescent spot on the sub-pixel point, determining whether the centroid of the fluorescent spot is in the edge area of the fluorescent image; wherein the edge area is determined based on the size information of the convolution kernel used in the image noise reduction processing; In the case where the centroid of the fluorescent spot is not in the edge area of the fluorescent image, determining the pixel point closest to the centroid of the fluorescent spot from the fluorescent spot; Taking the area around the pixel point closest in distance as the local area to be noise-reduced.

17. The computer device according to claim 1, characterized in that, In the method executed by the processor, the calculating the pixel value at the coordinate position of the centroid of the fluorescent spot on the sub-pixel point by using the biquadratic B-spline approximation method according to the coordinate positions and fluorescence intensity values of each pixel point in the fluorescent spot includes: Based on the coordinate position of the centroid of the fluorescent spot on the sub-pixel point, determining a first neighborhood area near the centroid; Determining a first weight in the horizontal dimension and a second weight in the vertical dimension of the first neighborhood area; Based on the first weight in the horizontal dimension, the second weight in the vertical dimension of the first neighborhood area, and the pixel values of each pixel point in the first neighborhood area, calculating the pixel value at the coordinate position of the centroid of the fluorescent spot on the sub-pixel point by using the biquadratic B-spline approximation method.

18. The computer device according to claim 17, characterized in that, In the method executed by the processor, the determining the first weight in the horizontal dimension and the second weight in the vertical dimension of the first neighborhood area includes: Converting the coordinate position of the centroid on the sub-pixel point from the first coordinate system corresponding to the fluorescent image to the second coordinate system corresponding to the first neighborhood area to obtain the converted coordinate position of the centroid; Determine the pixel point in the first neighborhood region that is closest to the centroid, and determine the pixel point as the central control point of the first neighborhood region; wherein, the coordinate position of the central control point is located under the second coordinate system corresponding to the first neighborhood region; Based on the abscissa value of the transformed coordinate position of the centroid and the abscissa value of the coordinate position of the central control point, determine the first weight of the first neighborhood region in the horizontal dimension; Based on the ordinate value of the transformed coordinate position of the centroid and the ordinate value of the coordinate position of the central control point, determine the second weight of the first neighborhood region in the vertical dimension.

19. The computer device according to claim 17, characterized in that, In the method executed by the processor, the calculating the pixel value of the coordinate position of the centroid of the fluorescence spot on the sub-pixel point by using the biquadratic B-spline approximation method based on the first weight of the first neighborhood region in the horizontal dimension, the second weight in the vertical dimension, and the pixel values of each pixel point in the first neighborhood region includes: Centered on the central control point, determine a plurality of to-be-processed pixel points from the first neighborhood region; divide the plurality of to-be-processed pixel points into a plurality of pixel point sets according to the first dimension; Based on the to-be-processed pixel points included in each pixel point set and the weight matching the first dimension, generate an intermediate pixel value corresponding to the pixel point set; Based on the intermediate pixel values corresponding to each of the pixel point sets and the weight matching the second dimension, determine the pixel value of the coordinate position of the centroid on the sub-pixel point; Wherein, the first dimension is the horizontal dimension and the second dimension is the vertical dimension, or the first dimension is the vertical dimension and the second dimension is the horizontal dimension.

20. A method for determining a fluorescence intensity value, characterized in that, Includes: Obtain a fluorescence image; the fluorescence image includes a plurality of fluorescence spots, wherein the plurality of fluorescence spots in one fluorescence image are generated by fluorescence imaging of the to-be-detected bases of a plurality of to-be-detected nucleic acid sequences in one sequencing cycle; Determine the coordinate positions of the centroids of at least some of the plurality of fluorescence spots in the fluorescence image on the sub-pixel points; According to the coordinate positions and pixel values of each pixel point in the fluorescence spot, calculate the pixel value at the coordinate position of the centroid of the fluorescence spot on the sub-pixel point by using the biquadratic B-spline approximation method, and use it as the fluorescence intensity value of the fluorescence spot.

21. The method according to claim 20, wherein The method further includes: Identify the type of the to-be-detected base that generates the fluorescence spot according to the fluorescence intensity value of the fluorescence spot.

22. The method according to claim 20, wherein The determining the coordinate positions of the centroids of at least some of the plurality of fluorescence spots in the fluorescence image on the sub-pixel points includes: Based on the pixel values of each pixel point in the fluorescence image, determine the contour information of at least some of the plurality of fluorescence spots; Based on the contour information of the fluorescence spot, determine the centroid of the connected domain of the fluorescence spot, and determine the coordinate position of the centroid as the coordinate position of the centroid of the fluorescence spot on the sub-pixel point.

23. The method according to claim 22, wherein The based on the pixel values of each pixel point in the fluorescence image, determining the contour information of at least some of the plurality of fluorescence spots includes: Invert the pixel values of each pixel point in the fluorescence image to obtain an inverted fluorescence image; Perform minimum value detection on the inverted fluorescence image to obtain a plurality of minimum value pixel points included in the inverted fluorescence image, and generate a minimum value image based on the target coordinate positions of the plurality of minimum value pixel points in the inverted fluorescence image. The pixel value at the target coordinate position in the minimum value image is a first preset value, and the pixel values at other coordinate positions are second preset values; Overlay the minimum value image and the fluorescence image to obtain a target fluorescence image; segment the target fluorescence image to obtain the contour information of the fluorescence spot.

24. The method according to claim 23, wherein Before segmenting the target fluorescence image to obtain the contour information of the fluorescence spot, it further includes: Perform binarization processing on the target fluorescence image to obtain a processed target fluorescence image; Perform opening operation on the processed target fluorescence image to obtain a denoised target fluorescence image; The segmenting the target fluorescence image to obtain the contour information of the fluorescence spot includes: Segment the denoised target fluorescence image to obtain the contour information of the fluorescence spot.

25. The method according to any one of claims 22 to 24, characterized in that Before determining the contour information of the fluorescence spot based on the pixel values of each pixel point in the fluorescence image, it further includes: Perform top-hat transformation processing on the fluorescence image to obtain a first fluorescence image, and overlay the fluorescence image and the first fluorescence image to obtain an overlaid fluorescence image; Perform black-hat transformation processing on the fluorescence image to obtain a second fluorescence image, and subtract the pixel values at the same positions in the overlaid fluorescence image and the second fluorescence image to obtain a processed fluorescence image; The determining the contour information of the fluorescence spot based on the pixel values of each pixel point in the fluorescence image includes: Determine the contour information of the fluorescence spot based on the pixel values of each pixel point in the processed fluorescence image.

26. The method according to any one of claims 22 to 24, characterized in that, The determining the centroid of the connected component of the fluorescence spot based on the contour information of the fluorescence spot includes: Determine the minimum circumscribed circle of the connected component of the fluorescence spot based on the contour information of the fluorescence spot; Determine the center of the minimum circumscribed circle, and determine the center of the minimum circumscribed circle as the centroid of the connected component of the fluorescence spot.

27. The method according to claim 20, characterized in that, Determine the coordinate positions of the centroids of at least some of the fluorescence spots in the plurality of fluorescence spots in the fluorescence image on sub-pixel points, including: Determine the pixel points with centroids in the fluorescence image, perform parabolic fitting on the pixel values of the pixel points with centroids, and determine the coordinate positions of the centroids of the fluorescence spots on sub-pixel points according to the fitting results.

28. The method according to claim 27, wherein The determining the pixel points with centroids in the fluorescence image includes: Obtain the gray values of each pixel point in the fluorescence image; Judge whether the gray value of each pixel point is greater than or equal to the gray value of the neighborhood area within a preset range from the pixel point; If so, determine that the pixel point has a centroid, otherwise determine that the pixel point does not have a centroid and discard the current pixel point.

29. The method according to claim 20, characterized in that, Performing parabolic fitting on the pixel values of the pixel points with centroids includes: Determine the number of parabola fittings according to the domain area within the preset range; Perform horizontal parabola fitting and vertical parabola fitting on the gray values of the pixel points with centroids respectively according to the number of parabola fittings.

30. The method according to claim 29, characterized in that, Before determining the number of parabola fittings according to the domain area within the preset range, it further includes: Screen the pixel points with centroids to obtain integer-level pixel points.

31. The method according to claim 29 or 30, characterized in that, Determine the coordinate position of the centroid of the fluorescence spot on the sub-pixel points according to the fitting result, including: Obtain the horizontal average value of the horizontal parabola peak points obtained multiple times, and use the horizontal average value as the abscissa of the centroid of the fluorescence spot on the sub-pixel points; Obtain the vertical average value of the horizontal parabola peak points and the peak points of the vertical parabola obtained multiple times, and use the vertical average value as the ordinate of the centroid of the fluorescence spot on the sub-pixel points; Determine the coordinate position of the centroid of the fluorescence spot on the sub-pixel points based on the abscissa and ordinate.

32. The method according to any one of claims 28 to 30, characterized in that, The domain area within the preset range includes: 4-neighborhood, 8-neighborhood, 20-neighborhood or 24-neighborhood.

33. The method according to claim 1, wherein After determining the coordinate positions of at least some of the centroids of the multiple fluorescence spots in the fluorescence image on the sub-pixel points, the method further includes: Perform image noise reduction processing on the fluorescence spots; The calculating the pixel value at the coordinate position of the centroid of the fluorescence spot on the sub-pixel points by using the biquadratic B-spline approximation method according to the coordinate positions and pixel values of the pixel points in the fluorescence spot includes: Calculate the pixel value at the coordinate position of the centroid of the fluorescence spot on the sub-pixel points by using the biquadratic B-spline approximation method according to the coordinate positions and pixel values of the pixel points in the fluorescence spot after noise reduction processing.

34. The method according to claim 33, characterized in that, The performing image noise reduction processing on the fluorescence spots includes: Based on the coordinate position of the centroid of the fluorescence spot on the sub-pixel points, determine the local area to be noise-reduced near the centroid, and the local area to be noise-reduced covers the neighborhood area within the preset range; Perform image noise reduction processing on the local area to be noise-reduced.

35. The method according to claim 34, characterized in that, The determining the local area to be noise-reduced near the centroid based on the coordinate position of the centroid of the fluorescence spot on the sub-pixel points includes: Based on the coordinate position of the centroid of the fluorescence spot on the sub-pixel points, determine whether the centroid of the fluorescence spot is in the edge area of the fluorescence image; wherein the edge area is determined based on the size information of the convolution kernel used in the image noise reduction processing; In the case where the centroid of the fluorescence spot is not in the edge area of the fluorescence image, determine the pixel point closest to the centroid of the fluorescence spot from the fluorescence spot; Use the area around the pixel point closest to the centroid as the local area to be noise-reduced.

36. The method according to claim 20, wherein The calculating the pixel value at the coordinate position of the centroid of the fluorescence spot on the sub-pixel points by using the biquadratic B-spline approximation method according to the coordinate positions and fluorescence intensity values of the pixel points in the fluorescence spot includes: Based on the coordinate position of the centroid of the fluorescence spot on the sub-pixel points, determine the first neighborhood area near the centroid; Determine a first weight of the first neighborhood region in the horizontal dimension and a second weight in the vertical dimension; Based on the first weight of the first neighborhood region in the horizontal dimension, the second weight in the vertical dimension, and the pixel values of each pixel point in the first neighborhood region, use the biquadratic B-spline approximation method to calculate the pixel value of the coordinate position of the centroid of the fluorescence spot on the sub-pixel point.

37. The method according to claim 36, wherein The determining the first weight of the first neighborhood region in the horizontal dimension and the second weight in the vertical dimension includes: Convert the coordinate position of the centroid on the sub-pixel point from the first coordinate system corresponding to the fluorescence image to the second coordinate system corresponding to the first neighborhood region to obtain the converted coordinate position of the centroid; Determine the pixel point closest to the centroid from the first neighborhood region, and determine the pixel point as the central control point of the first neighborhood region; wherein, the coordinate position of the central control point is in the second coordinate system corresponding to the first neighborhood region; Based on the abscissa value of the converted coordinate position of the centroid and the abscissa value of the coordinate position of the central control point, determine the first weight of the first neighborhood region in the horizontal dimension; Based on the ordinate value of the converted coordinate position of the centroid and the ordinate value of the coordinate position of the central control point, determine the second weight of the first neighborhood region in the vertical dimension.

38. The method according to claim 36, characterized in that, The calculating the pixel value of the coordinate position of the centroid of the fluorescence spot on the sub-pixel point by using the biquadratic B-spline approximation method based on the first weight of the first neighborhood region in the horizontal dimension, the second weight in the vertical dimension, and the pixel values of each pixel point in the first neighborhood region includes: Taking the central control point as the center, determine a plurality of pixels to be processed from the first neighborhood region; Divide the plurality of pixels to be processed into a plurality of pixel point sets according to the first dimension; Generate an intermediate pixel value corresponding to the pixel point set based on the pixels to be processed included in each pixel point set and the weight matching the first dimension; Based on the intermediate pixel values corresponding to each pixel point set respectively and the weight matching the second dimension, determine the pixel value of the coordinate position of the centroid on the sub-pixel point; Wherein, the first dimension is the horizontal dimension and the second dimension is the vertical dimension, or the first dimension is the vertical dimension and the second dimension is the horizontal dimension.

39. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, it executes the steps of the method for determining the fluorescence intensity value according to any one of claims 1 to 19.

Images