A laser speckle-based coagulation point recognition method

Abstract

The present application relates to the technical field of medical detection, in particular to a coagulation point recognition method based on laser speckle, comprising the following steps: S1, generating a circular light spot covering a dialysis kettle as a speckle excitation source by optical fiber laser beam expansion; S2, setting a filter and a polarizer in the speckle reflection path in turn to filter out ambient light and reflected light spots, and collecting a blood flow speckle image matrix; S3, calculating the sliding window gray scale contrast with the speckle image matrix as input and generating a coagulation degree graph matrix; S4, setting a threshold value according to the coagulation degree graph to determine the preliminary coagulation point and calculate the coagulation area; S5, combining the gray scale gradient with the blood flow velocity model to correct the threshold value and enhance the edge, and outputting the optimized coagulation distribution and the corrected area result. The present application realizes the accurate recognition of the coagulation point and the high-precision estimation of the coagulation area through high signal-to-noise ratio laser speckle imaging, preliminary recognition based on the coagulation degree graph, edge contrast enhancement and blood flow model correction.

Description

Technical Field

[0001] This invention relates to the field of medical testing technology, and in particular to a method for identifying coagulation points based on laser speckle. Background Technology

[0002] Coagulation status is a key monitoring indicator in dialysis, cardiac surgery, and extracorporeal blood circulation. Especially in hemodialysis, timely and accurate identification of coagulation points in dialyzers or tubing is of great significance for preventing blood flow obstruction and reducing the risk of thrombosis. Laser speckle imaging, as an optical technology based on dynamic interference, can monitor minute flow changes on tissue surfaces in a non-contact, real-time, and wide-area manner. In recent years, it has been widely used in medical scenarios such as cerebral blood flow and retinal blood flow detection, providing a new means with high temporal resolution and spatial accuracy for assessing blood flow status and velocity.

[0003] While speckle imaging has made progress in the field of blood flow detection, its application in coagulation point identification in scenarios such as dialysis still faces the following key challenges: First, existing methods mostly rely on external detection probes or local imaging perspectives, making it difficult to achieve complete coverage of the entire dialysis vessel or tubing, resulting in blind spots. Second, due to complex background light interference and reflection, the quality of the acquired images is unstable, affecting the accuracy of speckle contrast calculation. Third, traditional grayscale contrast algorithms do not fully incorporate blood flow velocity models, leading to insensitivity to changes in the early coagulation stage, blurry edges of identified coagulation areas, and inaccurate area assessment, making it difficult to meet the requirements of real-time clinical early warning and intervention. Summary of the Invention

[0004] This invention provides a method for identifying coagulation points based on laser speckle. By setting an annular expanded laser spot covering the surface of the dialysis vessel, speckle illumination is achieved from all angles. Filters and polarizers are used to eliminate background light interference and enhance the image signal-to-noise ratio. Furthermore, a coagulation degree map is constructed based on a sliding window grayscale contrast and blood flow velocity distribution model. Combined with an adaptive threshold and edge enhancement mechanism, accurate identification and area correction output of initial coagulation points are achieved. This method can be applied to scenarios such as real-time dialysis monitoring, coagulation early warning systems, and extracorporeal blood circulation safety assurance, improving the accuracy, robustness, and global coverage of coagulation identification.

[0005] A method for identifying coagulation points based on laser speckle includes the following steps:

[0006] S1, the fiber laser is expanded to generate a circular spot covering the target area of ​​the dialysis pot, and the area covered by the circular spot serves as the excitation source for acquiring speckle images;

[0007] S2, In the laser speckle reflection path formed after the circular light spot illuminates the surface of the dialysis pot, a filter and a polarizer are set in sequence. The filter is used to filter out ambient light, and the polarizer is used to remove the reflected light spot on the surface of the dialysis pot, thereby obtaining an imaging area that only includes blood flow speckle. The imaging area is converted into an original speckle image matrix by an image acquisition device.

[0008] S3, taking the original speckle image matrix as input, setting a sliding window of fixed size, calculating the region contrast based on pixel grayscale changes within each window region, and constructing a coagulation degree map matrix, which is used to describe the relative fluidity state of each region;

[0009] S4. Based on the coagulation degree map matrix, set an initial judgment threshold, mark the pixel positions below the initial judgment threshold as preliminary coagulation points, and count the number of all preliminary coagulation points to calculate the area of ​​the preliminary coagulation points.

[0010] S5, based on the preliminary coagulation point identification results, extracts the local image grayscale gradient, adjusts the initial judgment threshold in combination with the preset blood flow velocity model, and performs image enhancement processing on the blurred edge areas, outputting the optimized coagulation point distribution map and the corrected area result.

[0011] Optionally, S1 includes:

[0012] S11, start the fiber laser and select the preset wavelength (875nm infrared band) as the speckle excitation source;

[0013] S12, the convex lens group is arranged at the fiber laser emission end, and the beam divergence angle is controlled by adjusting the distance between the lens and the laser source so that the diameter of the expanded beam spot is larger than the diameter of the dialysis vessel surface.

[0014] S13, by fine-tuning the lens focal length and optical axis angle, corrects the shape of the light spot, so that it forms a circular light spot with uniform energy distribution on the surface of the dialysis vessel;

[0015] S14, the formed circular light spot is projected onto the surface of the dialysis vessel, and the irradiation center and coverage boundary are marked according to the structural characteristics of the dialysis vessel.

[0016] Optionally, S2 includes:

[0017] S21, In the reflection path of the laser speckle, a bandpass filter with a center wavelength matching the laser wavelength is installed at the front end of the imaging system. The bandpass filter selects a passband wavelength range of [range to be specified in the original text]. ,in, The wavelength of the laser. To limit the tolerance of the filter band;

[0018] S22, A polarizer is installed along the optical path after the filter to suppress the specular reflection bright spots caused by the curved surface structure of the dialysis pot. The incident angle of the polarizer is adjusted. This makes its polarization direction perpendicular to the principal direction of the reflected light from the pot, thereby increasing the intensity of the reflected light. minimize;

[0019] S23, after installing the filter and polarizer, turn on the laser source and project it onto the target area of ​​the dialysis pot, and adjust the image intensity distribution function. Verify whether there are bright interference areas within the imaging area;

[0020] S24, after completing the filtering and polarization interference suppression, uses a high frame rate CMOS camera to acquire images and generate the original speckle image matrix. .

[0021] Optionally, S3 includes:

[0022] S31, taking the original speckle image matrix as input, sets a sliding window of a fixed size on the image, and slides it region by region in the image according to a set step size. For each region, the distribution characteristics of pixel gray level in the region are statistically analyzed, including the gray level mean and gray level change amplitude, and the contrast of the region is calculated.

[0023] S32, the contrast calculated for each region in the sliding window is mapped to the relative coagulation degree value of that region, and organized in the same distribution to form a coagulation degree map matrix.

[0024] Optionally, S31 includes:

[0025] S311, taking the original speckle image matrix as input, sets a fixed size on the image. sliding window and with step size , Slide along the vertical and horizontal directions region by region to extract the gray values ​​of local areas in turn. ;

[0026] S312, in each sliding window Within this region, calculate the mean and standard deviation of the grayscale values ​​of all pixels.

[0027] S313 calculates the contrast of a local area based on the mean and standard deviation of all pixel grayscale values. .

[0028] Optionally, S32 includes:

[0029] S321, based on local region contrast And normalize it;

[0030] S322 maps the normalized contrast to the relative coagulation degree values ​​of the regions, and generates a coagulation degree map matrix using a reverse mapping method. .

[0031] Optionally, S4 includes:

[0032] S41, using a coagulation degree matrix As input, calculate the average of all values ​​in the coagulation severity map matrix. and multiply by the scaling factor Set initial judgment threshold ;

[0033] S42, traverse all positions in the coagulation degree map matrix. If its relative coagulation degree value Less than the initial judgment threshold If the location is marked as a preliminary coagulation point, a binary decision matrix is ​​constructed. ,in, It is a binary label value;

[0034] S43, for the binary decision matrix Sum all the preliminary coagulation points and count the total number of preliminary coagulation points. Based on the actual physical area of ​​each sliding window Calculate the initial coagulation area .

[0035] Optionally, S5 includes:

[0036] S51, taking the preliminary coagulation point distribution results as input, extracts the local image gray-level gradient distribution for each coagulation region, calculates the local dynamic correction factor by analyzing the difference in gray-level gradients inside and outside the coagulation region, and adaptively adjusts the initial judgment threshold in combination with the preset blood flow velocity model.

[0037] S52 employs a local contrast enhancement algorithm to highlight true boundary features. It combines the enhanced image information with the existing preliminary coagulation area, and based on the adjusted initial judgment threshold and preliminary coagulation area, it re-determines the coagulation attribution relationship of edge pixels, generates an optimized coagulation point distribution map, and outputs the corrected coagulation area.

[0038] Optionally, S51 includes:

[0039] S511, taking the preliminary coagulation point distribution matrix as input, extracts the local image gray-level gradient distribution within each preliminary coagulation region, including the gray-level gradient magnitude. and direction ;

[0040] S512, for each candidate coagulation region, calculate the average gradient within the region. and external average gradient ,based on , Difference calculation of local dynamic correction factor ;

[0041] S513, based on local dynamic correction factor The relative fluidity index output by the preset blood flow velocity model For the initial judgment threshold Adaptive adjustment is performed to obtain the local correction threshold. .

[0042] Optionally, S52 includes:

[0043] S521, Establish an edge neighborhood at the edge of the initial coagulation area. and for edge neighborhoods Local contrast enhancement is achieved using an adaptive histogram equalization algorithm. Generate enhanced edge image ;

[0044] S522, in the enhanced edge image In the middle, based on the local correction threshold Re-determine the blood clot attribution of edge pixels and update the blood clot distribution matrix. ,in, To optimize the pixel coagulation determination value, and to calculate the corrected coagulation area based on the optimized pixel coagulation determination value. .

[0045] The beneficial effects of this invention are:

[0046] This invention, by employing infrared laser combined with beam expansion and modulation and a combined interference suppression mechanism of filtering and polarization, can achieve high signal-to-noise ratio imaging of laser speckle on the surface of a dialysis chamber, significantly reducing the impact of curved surface reflection and background brightness on speckle image quality, effectively extracting target areas with dynamic flow characteristics, and ensuring that the basic data for image processing is accurate, stable, and covers the complete blood distribution area.

[0047] This invention constructs a coagulation degree map matrix based on the grayscale contrast of a sliding window, generates relative coagulation degree values ​​using reverse mapping, and combines an initial judgment threshold with window area calculation to achieve preliminary identification and quantification of potential coagulation areas. It has advantages such as clear judgment criteria, simple data structure, and flexible parameter adjustment.

[0048] This invention, by introducing a local dynamic correction factor driven by gray-level gradient difference and a blood flow velocity model, combined with adaptive contrast enhancement processing of edge regions, can effectively optimize the accuracy of coagulation point recognition under blurred edges. Furthermore, by correcting area errors based on pixel-based re-determination, it improves the system's recognition robustness and area estimation accuracy under different speckle complexities and coagulation distribution patterns. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a schematic diagram of the identification method flow according to an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram illustrating the dynamic correction of coagulation area in an embodiment of the present invention. Detailed Implementation

[0052] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. Those skilled in the art may employ other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0053] like Figures 1-2 As shown, a method for identifying coagulation points based on laser speckle includes the following steps:

[0054] S1, the fiber laser is expanded to generate a circular spot covering the target area of ​​the dialysis pot. The area covered by the circular spot is used as the excitation source for acquiring speckle images.

[0055] S2, In the laser speckle reflection path formed after the circular light spot illuminates the surface of the dialysis pot, a filter and a polarizer are set in sequence. The filter is used to filter out ambient light, and the polarizer is used to remove the reflected light spot on the surface of the dialysis pot, so as to obtain an imaging area that only includes blood flow speckle. The imaging area is converted into the original speckle image matrix by the image acquisition device.

[0056] S3 takes the original speckle image matrix as input, sets a sliding window of fixed size, calculates the region contrast based on pixel grayscale changes within each window region, and constructs a coagulation degree map matrix, which is used to describe the relative fluidity state of each region.

[0057] S4. Based on the coagulation degree map matrix, set an initial judgment threshold, mark the pixel positions below the initial judgment threshold as preliminary coagulation points, and count the number of all preliminary coagulation points to calculate the area of ​​the preliminary coagulation points.

[0058] S5, based on the preliminary coagulation point identification results, extracts the local image grayscale gradient, adjusts the initial judgment threshold in combination with the preset blood flow velocity model, and performs image enhancement processing on the blurred edge areas, outputting the optimized coagulation point distribution map and the corrected area result.

[0059] S1 includes:

[0060] S11, start the fiber laser, select the preset wavelength (875nm infrared band) as the speckle excitation source, and ensure stable and continuous output of optical power;

[0061] S12, the convex lens group is arranged at the fiber laser emission end. By adjusting the distance between the lens and the laser source, the beam divergence angle is controlled so that the diameter of the expanded beam spot is larger than the surface diameter of the dialysis pot, thus achieving complete coverage of the detection area of ​​the pot.

[0062] S13, by finely adjusting the lens focal length and optical axis angle, corrects the shape of the light spot, so that it forms a circular light spot with uniform energy distribution on the surface of the dialysis vessel, avoiding speckle saturation or reflection distortion caused by excessive local light intensity.

[0063] S14. The formed circular light spot is projected onto the surface of the dialysis vessel. The irradiation center and coverage boundary are marked according to the structural characteristics of the dialysis vessel to ensure that the area covered by the light spot is consistent with the distribution area of ​​potential coagulation points. This area serves as the excitation source for speckle image acquisition.

[0064] Determining the irradiation center and coverage boundary based on the structural characteristics of the dialysis vessel specifically includes:

[0065] (1) Obtain the external geometric parameters of the dialysis vessel (such as the outer diameter of the vessel, center line, and height from the bottom to the sampling window). These parameters are fixed values ​​in the standard parts of the vessel.

[0066] (2) Mount the laser system and camera system on a fixed platform and align them with the geometric center of the dialysis pot (e.g., the center point of the bottom of the pot).

[0067] (3) Align the center of the laser spot with the center line of the pot body with the laser-assisted alignment, and use a positioning screw or ruler for calibration;

[0068] (4) Set the coverage radius (slightly larger than the diameter of the blood channel in the pot or the width of the sampling window) so that the radius of the circular spot covers the entire high-incidence area of ​​coagulation;

[0069] (5) The location of the irradiation center and the coverage boundary are finally determined to control the beam expansion ratio and the incident angle.

[0070] S2 includes:

[0071] S21, In the reflection path of the laser speckle, a bandpass filter with a center wavelength matching the laser wavelength is installed at the front end of the imaging system. The bandpass filter is selected with a passband wavelength range of [range missing]. ,in, The wavelength of the laser. To limit the tolerance of the filter band;

[0072] S22, A polarizer is installed along the optical path after the filter to suppress the specular reflection bright spots caused by the curved surface structure of the dialysis pot. The incident angle of the polarizer is adjusted. This makes its polarization direction perpendicular to the principal direction of the reflected light from the pot, thereby increasing the intensity of the reflected light. Minimize, represented as:

[0073] ;

[0074] in, The initial intensity of the unpolarized reflected light, when At this time, reflected light is suppressed to the greatest extent;

[0075] S23, after installing the filter and polarizer, the laser light source is activated and projected onto the target area of ​​the dialysis pot. At this time, the dynamic speckle generated by blood flow is still retained, while curved surface reflections and bright backgrounds are filtered out. This is achieved by adjusting the image intensity distribution function. Verify the presence of bright interference areas within the imaging region, as shown below:

[0076] ;

[0077] in, For pixels brightness value, The threshold for determining background highlighting. Indicates the interference area; if the imaging area Areas with a coverage rate higher than 95% are considered effective imaging areas;

[0078] ;

[0079] in, , These represent the mean and standard deviation of the grayscale values ​​in the entire original speckle image, respectively. This is the adjustment coefficient;

[0080] S24, after completing the filtering and polarization interference suppression, uses a high frame rate CMOS camera to acquire images and generate the original speckle image matrix. .

[0081] S3 includes:

[0082] S31, taking the original speckle image matrix as input, sets a sliding window of a fixed size on the image, and slides it region by region in the image according to a set step size. For each region, the distribution characteristics of pixel gray level in the region are statistically analyzed, including the gray level mean and gray level change amplitude, and the contrast of the region is calculated.

[0083] S32, the contrast calculated for each region in the sliding window is mapped to the relative coagulation degree value of that region, and organized in the same distribution to form a coagulation degree map matrix.

[0084] S31 includes:

[0085] S311, taking the original speckle image matrix as input, sets a fixed size on the image. sliding window and with step size , Slide along the vertical and horizontal directions region by region to extract the gray values ​​of local areas in turn. , is represented as:

[0086] ;

[0087] in, For the first Line number Column pixel grayscale values, , These are the height and width of the sliding window, respectively.

[0088] S312, in each sliding window Within this region, the mean and standard deviation of the grayscale values ​​of all pixels are calculated to quantify the brightness concentration and fluctuation intensity of the region, expressed as:

[0089] ;

[0090] ;

[0091] in, , The first The mean and standard deviation of gray levels in each sliding window region This represents the total number of pixels within the window.

[0092] S313 calculates the contrast of a local area based on the mean and standard deviation of all pixel grayscale values. The light intensity fluctuations in this region are characterized by:

[0093] ;

[0094] in, It is a small positive number.

[0095] S32 includes:

[0096] S321, based on local region contrast The contrast values ​​are then normalized to ensure they fall within a uniform range, thus eliminating bias caused by differences in brightness across different areas. This is represented as:

[0097] ;

[0098] in, For normalized contrast, , These are the minimum and maximum values ​​in the entire contrast matrix, respectively. To prevent division by zero of small constants;

[0099] S322 maps the normalized contrast to the relative coagulation degree values ​​of the regions, and generates a coagulation degree map matrix using a reverse mapping method. , is represented as:

[0100] ;

[0101] in, The wavelength of the laser;

[0102] .

[0103] S4 includes:

[0104] S41, using a coagulation degree matrix As input, calculate the average of all values ​​in the coagulation severity map matrix. and multiply by the scaling factor Set initial judgment threshold This is used to distinguish potentially clotted areas from normal areas, and is represented as:

[0105] ;

[0106] in, , These represent the number of rows and columns in the coagulation severity map matrix, respectively. For the first The relative coagulation level values ​​for each window area;

[0107] S42, traverse all positions in the coagulation degree map matrix. If its relative coagulation degree value Less than the initial judgment threshold If the location is marked as a preliminary coagulation point, a binary decision matrix is ​​constructed. ,in, This is a binary label value that indicates whether a point of coagulation is present.

[0108] S43, for the binary decision matrix Sum all the preliminary coagulation points and count the total number of preliminary coagulation points. Based on the actual physical area of ​​each sliding window Calculate the initial coagulation area , is represented as:

[0109] .

[0110] S5 includes:

[0111] S51, taking the preliminary coagulation point distribution results as input, extracts the local image gray-level gradient distribution for each coagulation region, calculates the local dynamic correction factor by analyzing the difference in gray-level gradients inside and outside the coagulation region, and adaptively adjusts the initial judgment threshold in combination with the preset blood flow velocity model.

[0112] S52 employs a local contrast enhancement algorithm to highlight true boundary features. It combines the enhanced image information with the existing preliminary coagulation area, and based on the adjusted initial judgment threshold and preliminary coagulation area, it re-determines the coagulation attribution relationship of edge pixels, generates an optimized coagulation point distribution map, and outputs the corrected coagulation area.

[0113] S51 includes:

[0114] S511, taking the preliminary coagulation point distribution matrix as input, extracts the local image gray-level gradient distribution within each preliminary coagulation region, including the gray-level gradient magnitude. and direction , is represented as:

[0115] ;

[0116] ;

[0117] in, This represents the original grayscale value of the corresponding pixel. , They are grayscale values ​​in , The gradient component in the direction;

[0118] S512, for each candidate coagulation region, calculate the average gradient within the region. and external average gradient ,based on , Difference calculation of local dynamic correction factor , is represented as:

[0119] ;

[0120] in, It is a tiny positive number;

[0121] S513, based on local dynamic correction factor The relative fluidity index output by the preset blood flow velocity model For the initial judgment threshold Adaptive adjustment is performed to obtain the local correction threshold. , is represented as:

[0122] ;

[0123] in, , These are the corresponding adjustment coefficients;

[0124] The blood flow velocity model is represented as follows:

[0125] ;

[0126] in, , The first , Pixel position in frame image grayscale value, The total number of frames in the time window. In pixels A local window centered on the center. It is a very small positive number.

[0127] S52 includes:

[0128] S521, Establish an edge neighborhood at the edge of the initial coagulation area. and for edge neighborhoods Local contrast enhancement is achieved using an adaptive histogram equalization algorithm. Generate enhanced edge image Specifically, it includes:

[0129] (1) Binary decision matrix based on the initial coagulation region Extract its edge pixel set It includes all pixels on the edge contour, and is represented as:

[0130] ;

[0131] ;

[0132] in, It is a Sobel horizontal convolution kernel. It is a Sobel vertical convolution kernel, and * indicates a two-dimensional convolution operation. It is the grayscale value of the input image pixels. , These are the horizontal and vertical gradient images, respectively;

[0133] ;

[0134] ;

[0135] in, The edge strength threshold;

[0136] (2) For each edge pixel Construct a fixed-size edge neighborhood around it. And extract the neighborhood grayscale sub-image. , is represented as:

[0137] ;

[0138] ;

[0139] in, The neighborhood radius, The grayscale values ​​are those in the original image.

[0140] (3) For each neighborhood grayscale sub-image Regional application of adaptive histogram equalization algorithm The process generates locally enhanced sub-images, and then merges the enhancement results to form an edge-enhanced image. , is represented as:

[0141] ;

[0142] ;

[0143] in, For enhanced pixel grayscale, grayscale level is The number of pixels, The size of the neighborhood window. This represents the upper limit of grayscale levels. To enhance the grayscale value at the corresponding location in the local block, For all pixels The set of local enhancement block indexes For weighting;

[0144] S522, in the enhanced edge image In the middle, based on the local correction threshold Re-determine the blood clot attribution of edge pixels and update the blood clot distribution matrix. ,in, To optimize the pixel coagulation determination value, and to calculate the corrected coagulation area based on the optimized pixel coagulation determination value. , is represented as:

[0145] ;

[0146] ;

[0147] in, This is a preliminary coagulation assessment result. This represents the physical area corresponding to a single pixel.

[0148] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0149] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying coagulation points based on laser speckle, characterized in that, Includes the following steps: S1, the fiber laser is expanded to generate a circular spot covering the target area of ​​the dialysis pot, and the area covered by the circular spot serves as the excitation source for acquiring speckle images; S2, In the laser speckle reflection path formed after the circular light spot illuminates the surface of the dialysis pot, a filter and a polarizer are set in sequence. The filter is used to filter out ambient light, and the polarizer is used to remove the reflected light spot on the surface of the dialysis pot, thereby obtaining an imaging area that only includes blood flow speckle. The imaging area is converted into an original speckle image matrix by an image acquisition device. S3, taking the original speckle image matrix as input, setting a sliding window of fixed size, calculating the region contrast based on pixel grayscale changes within each window region, and constructing a coagulation degree map matrix, which is used to describe the relative fluidity state of each region; S4. Based on the coagulation degree map matrix, set an initial judgment threshold, mark the pixel positions below the initial judgment threshold as preliminary coagulation points, and count the number of all preliminary coagulation points to calculate the area of ​​the preliminary coagulation points. S5, based on the preliminary coagulation point identification results, extracts the local image grayscale gradient, adjusts the initial judgment threshold in combination with the preset blood flow velocity model, and performs image enhancement processing on the blurred edge areas, outputting the optimized coagulation point distribution map and the corrected area result.

2. The method for identifying coagulation points based on laser speckle according to claim 1, characterized in that, S1 includes: S11, start the fiber laser and select the preset wavelength as the speckle excitation source; S12, the convex lens group is arranged at the fiber laser emission end, and the beam divergence angle is controlled by adjusting the distance between the lens and the laser source so that the diameter of the expanded beam spot is larger than the diameter of the dialysis vessel surface. S13, by fine-tuning the lens focal length and optical axis angle, corrects the shape of the light spot, so that it forms a circular light spot with uniform energy distribution on the surface of the dialysis vessel; S14, the formed circular light spot is projected onto the surface of the dialysis vessel, and the irradiation center and coverage boundary are marked according to the structural characteristics of the dialysis vessel.

3. The method for identifying coagulation points based on laser speckle according to claim 2, characterized in that, S2 includes: S21, In the reflection path of the laser speckle, a bandpass filter with a center wavelength matching the laser wavelength is installed at the front end of the imaging system. The bandpass filter selects a passband wavelength range of [range to be specified in the original text]. ,in, The wavelength of the laser. To limit the tolerance of the filter band; S22, A polarizer is installed along the optical path after the filter to suppress the specular reflection bright spots caused by the curved surface structure of the dialysis pot. The incident angle of the polarizer is adjusted. This makes its polarization direction perpendicular to the principal direction of the reflected light from the pot, thereby increasing the intensity of the reflected light. minimize; S23, after installing the filter and polarizer, turn on the laser source and project it onto the target area of ​​the dialysis pot, and adjust the image intensity distribution function. Verify whether there are bright interference areas within the imaging area; S24, after completing the filtering and polarization interference suppression, uses a high frame rate CMOS camera to acquire images and generate the original speckle image matrix. .

4. The method for identifying coagulation points based on laser speckle according to claim 3, characterized in that, S3 includes: S31, taking the original speckle image matrix as input, sets a sliding window of a fixed size on the image, and slides it region by region in the image according to a set step size. For each region, the distribution characteristics of pixel gray level in the region are statistically analyzed, including the gray level mean and gray level change amplitude, and the contrast of the region is calculated. S32, the contrast calculated for each region in the sliding window is mapped to the relative coagulation degree value of that region, and organized in the same distribution to form a coagulation degree map matrix.

5. The method for identifying coagulation points based on laser speckle according to claim 4, characterized in that, S31 includes: S311, taking the original speckle image matrix as input, sets a fixed size on the image. sliding window and with step size , Slide along the vertical and horizontal directions region by region to extract the gray values ​​of local areas in turn. ; S312, in each sliding window Within this region, calculate the mean and standard deviation of the grayscale values ​​of all pixels. S313 calculates the contrast of a local area based on the mean and standard deviation of all pixel grayscale values. .

6. The method for identifying coagulation points based on laser speckle according to claim 5, characterized in that, S32 includes: S321, based on local region contrast And normalize it; S322 maps the normalized contrast to the relative coagulation degree values ​​of the regions, and generates a coagulation degree map matrix using a reverse mapping method. .

7. The method for identifying coagulation points based on laser speckle according to claim 6, characterized in that, S4 includes: S41, using a coagulation degree matrix As input, calculate the average of all values ​​in the coagulation severity map matrix. and multiply by the scaling factor Set initial judgment threshold ; S42, traverse all positions in the coagulation degree map matrix. If its relative coagulation degree value Less than the initial judgment threshold If the location is marked as a preliminary coagulation point, a binary decision matrix is ​​constructed. ,in, It is a binary label value; S43, for the binary decision matrix Sum all the preliminary coagulation points and count the total number of preliminary coagulation points. Based on the actual physical area of ​​each sliding window Calculate the initial coagulation area .

8. The method for identifying coagulation points based on laser speckle according to claim 7, characterized in that, S5 includes: S51, taking the preliminary coagulation point distribution results as input, extracts the local image gray-level gradient distribution for each coagulation region, calculates the local dynamic correction factor by analyzing the difference in gray-level gradients inside and outside the coagulation region, and adaptively adjusts the initial judgment threshold in combination with the preset blood flow velocity model. S52 employs a local contrast enhancement algorithm to highlight true boundary features. It combines the enhanced image information with the existing preliminary coagulation area, and based on the adjusted initial judgment threshold and preliminary coagulation area, it re-determines the coagulation attribution relationship of edge pixels, generates an optimized coagulation point distribution map, and outputs the corrected coagulation area.

9. The method for identifying coagulation points based on laser speckle according to claim 8, characterized in that, S51 includes: S511, taking the preliminary coagulation point distribution matrix as input, extracts the local image gray-level gradient distribution within each preliminary coagulation region, including the gray-level gradient magnitude. and direction ; S512, for each candidate coagulation region, calculate the average gradient within the region. and external average gradient ,based on , Difference calculation of local dynamic correction factor ; S513, based on local dynamic correction factor The relative fluidity index output by the preset blood flow velocity model For the initial judgment threshold Adaptive adjustment is performed to obtain the local correction threshold. .

10. The method for identifying coagulation points based on laser speckle according to claim 9, characterized in that, S52 includes: S521, Establish an edge neighborhood at the edge of the initial coagulation area. and for edge neighborhoods Local contrast enhancement is achieved using an adaptive histogram equalization algorithm. Generate enhanced edge image ; S522, in the enhanced edge image In the middle, based on the local correction threshold Re-determine the blood clot attribution of edge pixels and update the blood clot distribution matrix. ,in, To optimize the pixel coagulation determination value, and to calculate the corrected coagulation area based on the optimized pixel coagulation determination value. .

Images