Digestive system department prediction analysis system based on big data analysis

Abstract

The invention discloses a digestive system department prediction analysis system based on big data analysis, and particularly relates to the technical field of digestive system department prediction analysis, the quality monitoring information of endoscope equipment is acquired, the quality degradation coefficient of the endoscope equipment is calculated, the poor image quality caused by equipment aging or faults is avoided, the operation skill difference of operators is quantified, and the accuracy of prediction analysis is improved. An operation skill difference coefficient is calculated, image quality fluctuation caused by misoperation of an operator is reduced, an image acquisition decision model is constructed, an image acquisition decision index is generated, whether image re-acquisition needs to be carried out or not is intelligently judged, misdiagnosis caused by poor image quality is avoided, and the accuracy of image re-acquisition is improved. A color transition coefficient and an underexposure abnormal coefficient are calculated according to color transition information and exposure abnormal information of an initial endoscope image, an image optimization output index is generated, an optimal image is selected from the initial endoscope image set, and it is ensured that only a high-quality image is selected into final analysis. And the influence of low-quality images on subsequent analysis and diagnosis is avoided.

Description

Technical Field

[0001] The present invention relates to the field of gastroenterology prediction and analysis technology, and more specifically, to a gastroenterology prediction and analysis system based on big data analysis. Background Art

[0002] Digestive endoscopy is an important means of diagnosing gastrointestinal diseases and is widely used in early screening, lesion localization and disease assessment of gastrointestinal diseases. However, the difference and inconsistency of endoscopic image quality is one of the main challenges faced in the current diagnosis process. Digestive endoscopy images are affected by many factors, among which differences in equipment quality, operator technical level and individual operating habits are the key factors leading to unstable image quality. Specifically, problems such as uneven illumination, insufficient contrast and poor imaging quality of endoscopic equipment often make it difficult to clearly present the lesion area in the image, which in turn affects the accurate identification and diagnosis of the disease. Technical differences between different doctors or operators in endoscopic examinations, such as inspection angle, lens depth, stability control, etc., often lead to large differences in the quality of endoscopic images obtained by different doctors for the same patient, which in turn affects the discovery and judgment of lesions. Low-quality images not only increase the complexity of subsequent image processing, but also bring great challenges, especially in image denoising, contrast enhancement, and extraction of effective features. Moreover, image quality issues will directly affect the accuracy of the automated diagnostic system, especially for diseases that require urgent diagnosis, such as acute gastric bleeding and intestinal tumors. Image misjudgment may lead to delayed treatment and even endanger the patient's life. Therefore, in the field of gastroenterology, the accuracy and efficiency of disease diagnosis relying on low-quality endoscopic images are greatly restricted. Summary of the invention

[0003] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a gastroenterology prediction and analysis system based on big data analysis to solve the problems raised in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A gastroenterology prediction and analysis system based on big data analysis, including an equipment quality monitoring module, an operation skill monitoring module, an image acquisition decision module, an image information acquisition module, and an image optimization module;

[0006] The equipment quality monitoring module is used to obtain the quality monitoring information of the endoscope equipment and calculate the quality degradation coefficient of the endoscope equipment according to the quality monitoring information of the endoscope equipment;

[0007] An operation skill monitoring module is used to obtain the operation skill information of the operator and calculate the operation skill difference coefficient according to the operation skill information of the operator;

[0008] An image acquisition decision module is used to construct an image acquisition decision model based on the endoscopic equipment quality degradation coefficient and the operation skill difference coefficient, generate an image acquisition decision index, determine whether to re-acquire the endoscopic image based on the image acquisition decision index, and establish an initial endoscopic image set based on the endoscopic images that meet the requirements after re-acquisition;

[0009] An image information acquisition module, used to obtain color transition information and exposure abnormality information of an initial endoscopic image in an initial endoscopic image set, and calculate a color transition coefficient according to the color transition information, and calculate an underexposure abnormality coefficient according to the exposure abnormality information;

[0010] The image optimization module is used to construct an image optimization output model according to the color transition coefficient, the underexposure abnormality coefficient, and the image acquisition decision index, generate an image optimization output index, and output the optimal image in the initial endoscopic image set according to the image optimization output index.

[0011] In a preferred embodiment, the quality monitoring information of the endoscope device is obtained, the quality degradation of the endoscope device is analyzed, and the quality degradation coefficient of the endoscope device is obtained to measure the impact of the degradation of the performance of the endoscope device on the image quality;

[0012] The logic for obtaining the quality degradation coefficient of endoscope equipment is as follows:

[0013] Obtain the light intensity distribution data L(x) of the entire field of view during the use of the endoscope device i ,y i ), whose L(x i ,y i ) represents the light intensity at the two-dimensional coordinate (x, y) in the field of view, i is the order number of the sampling point; calculate the mean light intensity μ L , the expression is as follows i∈{1,2,...,I}; calculate the standard deviation of light intensity σ L , the expression is as follows Calculate the illumination uniformity coefficient U L , the expression is as follows Obtain the capture response time Tideal of the endoscope device to the optical signal during use, and calculate the capture response delay YT. The expression is as follows Where Tresponse is the expected ideal capture response time; obtain the focal length data of the endoscope lens at different positions, and calculate the focal length change rate JL, the expression is as follows Where JLmax is the maximum value of the focal length data, and JLmin is the minimum value of the focal length data; the endoscope equipment quality degradation coefficient NJS is calculated based on the illumination uniformity coefficient, capture response delay, and focal length change rate. The expression is as follows: NJS = a1*U L+a2*YT+a3*JL, where a1, a2, and a3 represent preset proportional coefficients of illumination uniformity coefficient, capture response delay, and focal length change rate, respectively, and a1, a2, and a3 are all greater than 0.

[0014] In a preferred embodiment, by acquiring the operation skill information of the operators, analyzing the differences in the operation skills of the operators, and acquiring the operation skill difference coefficient, the influence of the operation skill differences of the operators on the image quality is measured;

[0015] The logic for obtaining the operation skill difference coefficient is as follows:

[0016] Obtain the operator's operating angle when performing endoscopic examination and calculate the operating angle deviation Sθ. The expression is as follows: in It is a vector representation of the actual operating angle, used to capture angle deviations in multiple directions. is the vector representation of the target operation angle; It is the vector dot product operation of the actual operation angle and the target operation angle; is the vector magnitude of the target operation angle, is the modulus of the vector representation of the actual operating angle; obtain the operating depth data of the operator during endoscopic examination, and calculate the average operating depth μd, which is expressed as follows: μd = where sd n Represents the nth depth operation value, n∈{1,2,...,N}, calculates the depth volatility SH, the expression is as follows Where sdmax is the maximum value of the operating depth, and sdmin is the minimum value of the operating depth; obtain the moving speed of the endoscopic device when the operator performs endoscopic examination, and calculate the speed change rate Vrate, which is expressed as follows where v m represents the moving speed of the endoscope device at the mth moment, v m-1 represents the moving speed of the endoscope device at the m-1th moment, m∈{1,2,...,M}, and calculates the trajectory stability coefficient GJ, which is expressed as follows Wherein Vmax is the maximum allowable speed change rate; the operation skill difference coefficient CZX is calculated according to the operation angle offset, depth fluctuation rate, and trajectory stability coefficient, and the expression is as follows CZX=b1*Sθ+b2*SH+b3*GJ, where b1, b2, and b3 represent the preset proportional coefficients of the operation angle offset, depth fluctuation rate, and trajectory stability coefficient, respectively, and b1, b2, and b3 are all greater than 0.

[0017] In a preferred embodiment, an image acquisition decision model is constructed according to the endoscope equipment quality degradation coefficient and the operation skill difference coefficient to generate an image acquisition decision index Idecision. The formula based on the model is as follows: Where c1 and c2 represent the preset proportional coefficients of the endoscopic equipment quality degradation coefficient and the operating skill difference coefficient, respectively, and both c1 and c2 are greater than 0.

[0018] In a preferred embodiment, the image acquisition decision index is compared with a preset image acquisition decision index threshold to determine whether to perform endoscopic image re-acquisition, as follows:

[0019] If the image acquisition decision index is greater than the image acquisition decision index threshold, image re-acquisition is required, and the determination of endoscopic image re-acquisition is repeated according to the image acquisition decision module; if the image acquisition decision index is less than or equal to the image acquisition decision index threshold, the currently acquired endoscopic image is used as the initial endoscopic image, and an initial endoscopic image set is established based on the initial endoscopic image.

[0020] In a preferred embodiment, the color transition of the initial endoscopic image is measured by acquiring color transition information of the initial endoscopic image, analyzing the color transition of the initial endoscopic image, and acquiring a color transition coefficient;

[0021] The logic for obtaining the color transition coefficient is as follows:

[0022] Preprocessing the initial endoscopic image, including removing noise and enhancing contrast; converting the preprocessed initial endoscopic image from RGB color space to HSV color space;

[0023] The Sobel operator is used to calculate the gradient of the initial endoscopic image, as follows: Among them G x is the gradient in the x direction, II is the initial endoscopic image input, Among them G y is the gradient in the y direction; calculate the global average gradient Gavg of the initial endoscopic image, the expression is as follows Among them G x (h) is the gradient in the x direction at the initial endoscopic image pixel h, G y (h) is the gradient in the y direction at the pixel h of the initial endoscopic image, H is the total number of pixels of the initial endoscopic image, h∈{1,2,...,H}; the initial endoscopic image is divided into multiple local regions according to the preset division criteria, and the average gradient Glocal of different local regions is calculated k , calculate the color transition coefficient SGD, the expression is as follows where Glocalk Denote the average gradient of the k-th local region, where k ∈ {1, 2,..., K}.

[0024] In a preferred embodiment, by obtaining the exposure anomaly information of the initial endoscopic image, analyzing the underexposure situation of the initial endoscopic image, and obtaining the underexposure anomaly coefficient to measure the anomaly degree of the underexposure of the initial endoscopic image;

[0025] The acquisition logic of the underexposure anomaly coefficient is as follows:

[0026] For each pixel of the initial endoscopic image, calculate the brightness value of each pixel to obtain the set of brightness values of all pixels LD = {ld f} = {ld1, ld2,..., ld F}, where ld f represents the brightness value of the f-th pixel, f ∈ {1, 2,..., F}, and the calculation expression of ld f is as follows: ld f = 0.299 * R f + 0.587 * G f + 0.114 * B f , where R f , G f , and B f respectively represent the values of the red, green, and blue channels of the f-th pixel;

[0027] Sort the set of brightness values LD = {ld f} = {ld1, ld2,..., ld F} in ascending order to obtain the sorted brightness value sequence, and calculate the value position of the brightness threshold according to the sorted brightness value sequence. The expression is as follows: Position = 0.1 * (F - 1), where F is the total number of pixels and Position is the value position of the brightness threshold. If the value position of the brightness threshold is an integer, directly take the brightness value at this position as the brightness threshold; if the value position of the brightness threshold is not an integer, take the weighted average of two adjacent values as the brightness threshold; calculate the underexposure anomaly coefficient QBG, and the expression is as follows where Llow is the brightness threshold, and 1(ld f <Llow) is an indicator function, indicating 1 when ld f <Llow, and 0 otherwise.

[0028] In a preferred embodiment, construct an image preferred output model according to the color transition coefficient, underexposure anomaly coefficient, and image acquisition decision index to generate the image preferred output index Iopt. The formula on which the model is based is as follows Wherein d1 and d2 represent the preset proportional coefficients of the color transition coefficient and the underexposure abnormality coefficient respectively, and both d1 and d2 are greater than 0;

[0029] The initial endoscopic images in the initial endoscopic image set are sorted from large to small based on the image optimization output index, and the initial endoscopic image ranked first is taken as the optimal image.

[0030] Technical effects and advantages of the present invention:

[0031] 1. The present invention obtains the quality monitoring information of the endoscopic equipment in real time through the equipment quality monitoring module, and calculates the quality degradation coefficient of the endoscopic equipment to help doctors understand the equipment status in time and avoid poor image quality due to equipment aging or failure. The operation skill monitoring module quantifies the difference in the operation skills of operators, calculates the operation skill difference coefficient, and reduces the image quality fluctuation caused by improper operation of operators, thereby ensuring the consistency and stability of the image. The image acquisition decision module combines the equipment quality degradation coefficient and the operation skill difference coefficient to construct an image acquisition decision model, and intelligently determines whether image re-collection is required by generating an image acquisition decision index. Image re-collection is performed in time according to actual conditions to avoid misdiagnosis caused by poor image quality and improve the accuracy of diagnosis. The image information acquisition module calculates the color transition coefficient and the underexposure abnormality coefficient according to the color transition information and exposure abnormality information of the initial endoscopic image, and timely evaluates whether the color transition smoothness and exposure in the image are normal, helping to identify possible quality problems in the image. The image optimization module generates an image optimization output index by combining the color transition coefficient and the underexposure abnormality coefficient, selects the optimal image from the initial endoscopic image set, and ensures that only high-quality images are selected into the final analysis to avoid low-quality images affecting subsequent analysis and diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0033] Figure 1 Flowchart of a system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] Embodiment: The present invention provides as Figure 1A gastroenterology prediction and analysis system based on big data analysis is shown, comprising an equipment quality monitoring module, an operation skill monitoring module, an image acquisition decision module, an image information acquisition module, and an image optimization module;

[0036] The equipment quality monitoring module is used to obtain the quality monitoring information of the endoscope equipment and calculate the quality degradation coefficient of the endoscope equipment according to the quality monitoring information of the endoscope equipment;

[0037] An operation skill monitoring module is used to obtain the operation skill information of the operator and calculate the operation skill difference coefficient according to the operation skill information of the operator;

[0038] An image acquisition decision module is used to construct an image acquisition decision model based on the endoscopic equipment quality degradation coefficient and the operation skill difference coefficient, generate an image acquisition decision index, determine whether to re-acquire the endoscopic image based on the image acquisition decision index, and establish an initial endoscopic image set based on the endoscopic images that meet the requirements after re-acquisition;

[0039] An image information acquisition module, used to obtain color transition information and exposure abnormality information of an initial endoscopic image in an initial endoscopic image set, and calculate a color transition coefficient according to the color transition information, and calculate an underexposure abnormality coefficient according to the exposure abnormality information;

[0040] An image optimization module is used to construct an image optimization output model according to a color transition coefficient, an underexposure abnormality coefficient, and an image acquisition decision index, generate an image optimization output index, and output an optimal image in an initial endoscopic image set according to the image optimization output index;

[0041] The equipment quality monitoring module is used to obtain the quality monitoring information of the endoscope equipment and calculate the quality degradation coefficient of the endoscope equipment according to the quality monitoring information of the endoscope equipment;

[0042] The endoscopic equipment quality degradation coefficient is used to measure the impact of endoscopic equipment performance on image quality. It integrates the key performance parameters of endoscopic equipment, evaluates the imaging capabilities of the equipment in actual use, and provides an important reference for image acquisition and optimization; by quantifying the impact of equipment performance on image quality, it achieves dynamic optimization of endoscopic image acquisition quality, effectively avoiding low-quality images caused by equipment performance problems from being used for diagnosis. By calculating the endoscopic equipment quality degradation coefficient, the accuracy and efficiency of the image acquisition process can be significantly improved, thereby reducing the risk of misjudgment of lesions caused by equipment problems such as uneven illumination, insufficient resolution, and low contrast. In addition, by automatically determining whether to re-acquire images, the subjectivity of manual judgment is reduced, the time period for image acquisition is shortened, and more efficient support is provided for diseases that require urgent diagnosis. At the same time, the accuracy and reliability of diagnosis are guaranteed, and the overall quality of endoscopic examinations and medical service levels are ultimately improved. A larger endoscopic equipment quality degradation coefficient indicates that the degradation of endoscopic equipment performance has a greater impact on image quality, and the more endoscopic images need to be re-acquired. A smaller endoscopic equipment quality degradation coefficient indicates that the degradation of endoscopic equipment performance has a smaller impact on image quality, indicating that the quality of the currently acquired endoscopic images is better, which greatly reduces the need for re-acquisition of endoscopic images.

[0043] Therefore, by obtaining the quality monitoring information of the endoscopic equipment, analyzing the quality degradation of the endoscopic equipment, and obtaining the quality degradation coefficient of the endoscopic equipment, the impact of the degradation of the performance of the endoscopic equipment on the image quality is measured;

[0044] The logic for obtaining the quality degradation coefficient of endoscope equipment is as follows:

[0045] Obtain the light intensity distribution data L(x) of the entire field of view during the use of the endoscope device i ,y i ), whose L(x i ,y i ) represents the light intensity at the two-dimensional coordinate (x, y) in the field of view, i is the order number of the sampling point; calculate the mean light intensity μ L , the expression is as follows Calculate the light intensity standard deviation σ L , the expression is as follows Calculate the illumination uniformity coefficient U L , the expression is as follows Obtain the capture response time Tideal of the endoscope device to the optical signal during use, and calculate the capture response delay YT. The expression is as follows Where Tresponse is the expected ideal capture response time; obtain the focal length data of the endoscope lens at different positions, and calculate the focal length change rate JL, the expression is as follows Where JLmax is the maximum value of the focal length data, and JLmin is the minimum value of the focal length data; the endoscope equipment quality degradation coefficient NJS is calculated based on the illumination uniformity coefficient, capture response delay, and focal length change rate. The expression is as follows: NJS = a1*U L +a2*YT+a3*JL, where a1, a2, and a3 represent preset proportional coefficients of illumination uniformity coefficient, capture response delay, and focal length change rate, respectively, and a1, a2, and a3 are all greater than 0;

[0046] It should be noted that before calculating the quality degradation coefficient of the endoscope equipment, it is necessary to ensure that the illumination uniformity coefficient, capture response delay, and focal length change rate are all normalized. Common normalization methods include Min-Max normalization and Z-Score standardization. a1, a2, and a3 are set according to actual conditions. For example, the expert empowerment method is adopted, that is, experts in related fields are invited to determine the preset proportional coefficients of various indicators through professional opinion surveys and comprehensive evaluations.

[0047] An operation skill monitoring module is used to obtain the operation skill information of the operator and calculate the operation skill difference coefficient according to the operation skill information of the operator;

[0048] The operation skill difference coefficient is used to measure the impact of the operator's operation skill differences on image quality. It can help identify the image quality differences that may be caused by different operators when performing endoscopic examinations. By quantifying the operation skill differences of operators, it is possible to identify image quality problems caused by differences in operation habits. Based on the operation skill difference coefficient, it is automatically determined that image re-acquisition is required to ensure the acquisition of higher quality images. By timely correcting the operator's operation habits and skills, it can effectively reduce the poor image quality caused by individual differences, thereby improving diagnostic accuracy and reducing the risk of misdiagnosis. This method can not only improve the overall image quality, but also improve the credibility of clinical decisions in practical applications, especially when emergency diagnosis or precise evaluation is required. By combining operation skill monitoring with image quality evaluation, the gastroenterology prediction analysis system can better serve clinical needs and reduce the impact of individual operator differences on diagnostic results. A larger operation skill difference coefficient indicates that the operator's operation skill differences have a greater impact on image quality, and the more endoscopic image re-acquisition is needed. A smaller operation skill difference coefficient indicates that the operator's operation performance is more in line with the expected standard, indicating that the quality of the currently acquired endoscopic image is better, which greatly reduces the re-acquisition of endoscopic images.

[0049] Therefore, by obtaining the operator's operating skill information, analyzing the differences in the operators' operating skills, and obtaining the operating skill difference coefficient, the impact of the operator's operating skill difference on the image quality is measured;

[0050] The logic for obtaining the operation skill difference coefficient is as follows:

[0051] Obtain the operator's operating angle when performing endoscopic examination and calculate the operating angle deviation Sθ. The expression is as follows: in It is a vector representation of the actual operating angle, used to capture angle deviations in multiple directions. is the vector representation of the target operation angle; It is the vector dot product operation of the actual operation angle and the target operation angle; is the vector magnitude of the target operation angle, is the modulus of the vector representation of the actual operating angle; obtain the operating depth data of the operator during endoscopic examination, and calculate the average operating depth μd, the expression is as follows where sd n Represents the nth depth operation value, n∈{1,2,...,N}, calculates the depth volatility SH, the expression is as follows Where sdmax is the maximum value of the operating depth, and sdmin is the minimum value of the operating depth; obtain the moving speed of the endoscopic device when the operator performs endoscopic examination, and calculate the speed change rate Vrate, which is expressed as follows where v m represents the moving speed of the endoscope device at the mth moment, v m-1 represents the moving speed of the endoscope device at the m-1th moment, m∈{1,2,...,M}, and calculates the trajectory stability coefficient GJ, which is expressed as follows Where Vmax is the maximum allowable speed change rate; the operation skill difference coefficient CZX is calculated according to the operation angle offset, depth fluctuation rate, and trajectory stability coefficient, and the expression is as follows CZX=b1*Sθ+b2*SH+b3*GJ, where b1, b2, and b3 represent the preset proportional coefficients of the operation angle offset, depth fluctuation rate, and trajectory stability coefficient, respectively, and b1, b2, and b3 are all greater than 0;

[0052] It should be noted that before calculating the operation skill difference coefficient, it is necessary to ensure that the operation angle offset, depth volatility, and trajectory stability coefficient are all normalized; b1, b2, and b3 are set according to actual conditions. For example, the expert empowerment method is adopted, that is, experts in related fields are invited to determine the preset proportional coefficients of various indicators through professional opinion surveys and comprehensive evaluations;

[0053] An image acquisition decision module is used to construct an image acquisition decision model based on the endoscopic equipment quality degradation coefficient and the operation skill difference coefficient, generate an image acquisition decision index, determine whether to re-acquire the endoscopic image based on the image acquisition decision index, and establish an initial endoscopic image set based on the endoscopic images that meet the requirements after re-acquisition;

[0054] According to the coefficient of endoscope quality degradation and the coefficient of operation skill difference, an image acquisition decision model is constructed to generate the image acquisition decision index Idecision. The formula of the model is as follows In the formula, c1 and c2 represent the preset proportional coefficients of the endoscope equipment quality degradation coefficient and the operation skill difference coefficient, respectively, and both c1 and c2 are greater than 0;

[0055] It should be noted that before constructing the image acquisition decision model, it is necessary to ensure that the coefficient of endoscopic equipment quality degradation and the coefficient of operation skill difference are normalized; c1 and c2 are set according to the actual situation. For example, the expert empowerment method is adopted, that is, experts in related fields are invited to determine the preset proportional coefficients of various indicators through professional opinion surveys and comprehensive evaluations;

[0056] It can be seen from the above calculation expression that the greater the coefficient of endoscopic equipment quality degradation and the greater the coefficient of difference in operating skills, the greater the image acquisition decision index, indicating that the demand for endoscopic image re-acquisition is greater; conversely, the smaller the coefficient of endoscopic equipment quality degradation and the smaller the coefficient of difference in operating skills, the smaller the image acquisition decision index, indicating that the demand for endoscopic image re-acquisition is smaller;

[0057] The image acquisition decision index is compared with the preset image acquisition decision index threshold to determine whether to perform endoscopic image re-acquisition, as follows:

[0058] If the image acquisition decision index is greater than the image acquisition decision index threshold, it means that the degradation of the performance of the endoscopic device and the difference in the operating skills of the operators have a greater impact on the image quality, and image re-acquisition is required, and the determination of endoscopic image re-acquisition is repeated according to the image acquisition decision module; if the image acquisition decision index is less than or equal to the image acquisition decision index threshold, it means that the degradation of the performance of the endoscopic device and the difference in the operating skills of the operators have a negligible impact on the image quality, and the currently acquired endoscopic image is used as the initial endoscopic image, and an initial endoscopic image set is established based on the initial endoscopic image;

[0059] An image information acquisition module, used to obtain color transition information and exposure abnormality information of an initial endoscopic image in an initial endoscopic image set, and calculate a color transition coefficient according to the color transition information, and calculate an underexposure abnormality coefficient according to the exposure abnormality information;

[0060] The color transition coefficient is used to measure the color transition of the initial endoscopic image. The color transition coefficient can effectively reflect the smooth transition of the color in the image. Especially in digestive endoscopy, the color change in the image is usually used to distinguish different tissues and lesion areas. If the color transition is uneven, it may cause unclear boundaries of certain lesion areas in the image, or even artifacts, which will affect the doctor's diagnosis of the image. By calculating the color transition coefficient, the degree of color change in the image can be quantified, and areas with uneven or unnatural transitions can be identified. Possible quality problems in the image, such as uneven lighting, inaccurate focus, or insufficient image processing, can be discovered in a timely manner, so as to identify images that meet diagnostic needs in advance.

[0061] Therefore, by acquiring the color transition information of the initial endoscopic image, analyzing the color transition of the initial endoscopic image, and acquiring the color transition coefficient, the color transition of the initial endoscopic image is measured;

[0062] The logic for obtaining the color transition coefficient is as follows:

[0063] Preprocessing the initial endoscopic image, including removing noise and enhancing contrast; converting the preprocessed initial endoscopic image from RGB color space to HSV color space;

[0064] It should be noted that the HSV color space is closer to the perception of human eyes, which is convenient for measuring color transition. Converting the initial endoscopic image from RGB color space to HSV color space is an existing mature technology and will not be described in detail here.

[0065] The Sobel operator is used to calculate the gradient of the initial endoscopic image, as follows: Among them G x is the gradient in the x direction, II is the initial endoscopic image input, Among them G y is the gradient in the y direction; calculate the global average gradient Gavg of the initial endoscopic image, the expression is as follows Among them G x (h) is the gradient in the x direction at the initial endoscopic image pixel h, G y (h) is the gradient in the y direction at the pixel h of the initial endoscopic image, H is the total number of pixels of the initial endoscopic image, h∈{1,2,...,H}; the initial endoscopic image is divided into multiple local regions according to the preset division criteria, and the average gradient Glocal of different local regions is calculated k , calculate the color transition coefficient SGD, the expression is as follows where Glocal krepresents the average gradient of the kth local region, k∈{1,2,...,K};

[0066] It should be noted that the initial endoscopic image is divided into multiple local areas according to a preset division standard. Common image division standards include division based on color space, division based on image texture, division based on depth information, division based on image edge, etc. In this embodiment, the image is divided into multiple areas based on different color ranges based on image color. The HSV color space is used to divide the hue (H) into different ranges, and each range corresponds to a color area. The calculation method of the average gradient of different local areas is consistent with the calculation method of the global average gradient, which will not be repeated here.

[0067] The underexposure anomaly coefficient is used to measure the degree of exposure anomaly caused by insufficient light in the initial endoscopic image. Underexposure will lead to loss of details in the image, especially in low-light environments, the contrast, color saturation and detail presentation of the image will be affected. The underexposure anomaly coefficient can help identify poor image quality and ensure that the endoscopic image has sufficient readability and accuracy. Underexposure will lead to insufficient brightness in the image, which will affect the visibility of the lesion area. Especially in endoscopic examinations, many small lesions such as tiny ulcers or polyps will be difficult to distinguish or clearly present if the image is underexposed. By calculating the underexposure anomaly coefficient, it is possible to accurately measure whether there are too dark areas in the image or whether the overall brightness is balanced, thereby providing a basis for subsequent image processing or re-acquisition decisions. Through quantitative analysis of the underexposure anomaly coefficient, the system can detect exposure anomalies in a timely manner and respond to them, avoiding the loss of image information due to insufficient exposure. For scenarios that require accurate diagnosis, the evaluation of the underexposure anomaly coefficient not only improves the reliability of image quality, but also reduces the risk of misdiagnosis. Especially in emergency diagnosis, it can detect underexposed images at the first time and re-acquire images to ensure the best image quality, thereby improving diagnostic efficiency and accuracy;

[0068] Therefore, by acquiring the exposure abnormality information of the initial endoscopic image, analyzing the underexposure of the initial endoscopic image, and acquiring the underexposure abnormality coefficient, the abnormal degree of underexposure of the initial endoscopic image is measured;

[0069] The logic for obtaining the underexposure anomaly coefficient is as follows:

[0070] For each pixel of the initial endoscopic image, the brightness value of each pixel is calculated to obtain the brightness value set of all pixels LD = {ld f}={ld1,ld2,...,ld F}, where ld f represents the brightness value of the f-th pixel, f∈{1,2,...,F},ldf The calculation expression of ld is as follows f ld = 0.299 * R f + 0.587 * G f + 0.114 * B f , where R f , G f , B f represent the values of the red, green, and blue channels of the f-th pixel respectively;

[0071] Sort the set of luminance values LD = {ld f} = {ld1, ld2,..., ld F} in ascending order to obtain the sorted luminance value sequence. Calculate the position of the luminance threshold according to the sorted luminance value sequence. The expression is as follows: Position = 0.1 * (F - 1), where F is the total number of pixels and Position is the position of the luminance threshold. If the position of the luminance threshold is an integer, directly take the luminance value at this position as the luminance threshold; if the position of the luminance threshold is not an integer, take the weighted average of two adjacent values as the luminance threshold; calculate the underexposure anomaly coefficient QBG. The expression is as follows where Llow is the luminance threshold, 1(ld f < Llow) is an indicator function, which is 1 when ld f < Llow and 0 otherwise;

[0072] An image optimization module is used to construct an image optimization output model based on the color transition coefficient, underexposure anomaly coefficient, and image acquisition decision index, generate an image optimization output index, and output the optimal image in the initial endoscopic image set according to the image optimization output index;

[0073] Construct an image optimization output model based on the color transition coefficient, underexposure anomaly coefficient, and image acquisition decision index to generate an image optimization output index Iopt. The formula on which the model is based is as follows In the formula, d1 and d2 represent the preset proportionality coefficients of the color transition coefficient and underexposure anomaly coefficient respectively, and both d1 and d2 are greater than 0;

[0074] It should be noted that before constructing the image optimization output model, it is necessary to ensure that both the color transition coefficient and the underexposure anomaly coefficient have been normalized; d1 and d2 are set according to the actual situation. For example, the expert weighting method is adopted, that is, relevant experts are invited to determine the preset proportionality coefficients of each index through professional opinion surveys and comprehensive evaluations;

[0075] It can be seen from the above calculation expression that the larger the color transition coefficient and the larger the underexposure abnormality coefficient, the smaller the image preferred output index, indicating that the quality of the endoscopic image is worse; conversely, the smaller the color transition coefficient and the smaller the underexposure abnormality coefficient, the larger the image preferred output index, indicating that the quality of the endoscopic image is better;

[0076] The initial endoscopic images in the initial endoscopic image set are sorted from large to small based on the image optimization output index, and the initial endoscopic image ranked first is taken as the optimal image;

[0077] The present invention obtains the quality monitoring information of the endoscopic equipment in real time through the equipment quality monitoring module, and helps the doctor to understand the equipment status in time by calculating the quality degradation coefficient of the endoscopic equipment, so as to avoid poor image quality caused by equipment aging or failure. The operation skill monitoring module quantifies the difference in the operation skills of the operators, calculates the operation skill difference coefficient, and reduces the image quality fluctuation caused by improper operation of the operators, thereby ensuring the consistency and stability of the image. The image acquisition decision module combines the equipment quality degradation coefficient and the operation skill difference coefficient to construct an image acquisition decision model, and intelligently judges whether it is necessary to re-acquire the image by generating an image acquisition decision index. The image re-acquisition is performed in time according to the actual situation, so as to avoid misdiagnosis caused by poor image quality and improve the accuracy of diagnosis. The image information acquisition module calculates the color transition coefficient and the underexposure abnormality coefficient according to the color transition information and the exposure abnormality information of the initial endoscopic image, and timely evaluates whether the color transition smoothness and the exposure in the image are normal, so as to help identify the quality problems that may exist in the image. The image optimization module generates the image optimization output index by combining the color transition coefficient and the underexposure abnormality coefficient, selects the optimal image from the initial endoscopic image set, ensures that only high-quality images are selected into the final analysis, and avoids the influence of low-quality images on subsequent analysis and diagnosis.

[0078] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0079] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0080] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0081] 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 computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0082] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A gastroenterology prediction and analysis system based on big data analysis, characterized by: It includes equipment quality monitoring module, operation skill monitoring module, image acquisition decision module, image information acquisition module and image optimization module; The equipment quality monitoring module is used to obtain the quality monitoring information of the endoscope equipment and calculate the quality degradation coefficient of the endoscope equipment according to the quality monitoring information of the endoscope equipment; An operation skill monitoring module is used to obtain the operation skill information of the operator and calculate the operation skill difference coefficient according to the operation skill information of the operator; An image acquisition decision module is used to construct an image acquisition decision model based on the endoscopic equipment quality degradation coefficient and the operation skill difference coefficient, generate an image acquisition decision index, determine whether to re-acquire the endoscopic image based on the image acquisition decision index, and establish an initial endoscopic image set based on the endoscopic images that meet the requirements after re-acquisition; An image information acquisition module, used to obtain color transition information and exposure abnormality information of an initial endoscopic image in an initial endoscopic image set, and calculate a color transition coefficient according to the color transition information, and calculate an underexposure abnormality coefficient according to the exposure abnormality information; The image optimization module is used to construct an image optimization output model according to the color transition coefficient, the underexposure abnormality coefficient, and the image acquisition decision index, generate an image optimization output index, and output the optimal image in the initial endoscopic image set according to the image optimization output index.

2. The gastroenterology prediction and analysis system based on big data analysis according to claim 1, characterized in that: By obtaining the quality monitoring information of the endoscopic equipment, analyzing the quality degradation of the endoscopic equipment, and obtaining the quality degradation coefficient of the endoscopic equipment, the impact of the degradation of the performance of the endoscopic equipment on the image quality is measured; The logic for obtaining the quality degradation coefficient of endoscope equipment is as follows: Obtain the light intensity distribution data L(x i ,y i ), whose L(x i ,y i ) represents the light intensity at the two-dimensional coordinate (x, y) in the field of view, i is the order number of the sampling point; calculate the mean light intensity μ L , the expression is as follows i∈{1,2,...,I}; calculate the standard deviation of light intensity σ L , the expression is as follows Calculate the illumination uniformity coefficient U L , the expression is as follows Obtain the capture response time Tideal of the endoscope device to the optical signal during use, and calculate the capture response delay YT. The expression is as follows Where Tresponse is the expected ideal capture response time; obtain the focal length data of the endoscope lens at different positions, and calculate the focal length change rate JL, the expression is as follows Where JLmax is the maximum value of the focal length data, and JLmin is the minimum value of the focal length data; the endoscope equipment quality degradation coefficient NJS is calculated based on the illumination uniformity coefficient, capture response delay, and focal length change rate. The expression is as follows: NJS = a1*U L +a2*YT+a3*JL, where a1, a2, and a3 represent preset proportional coefficients of illumination uniformity coefficient, capture response delay, and focal length change rate, respectively, and a1, a2, and a3 are all greater than 0.

3. The gastroenterology prediction and analysis system based on big data analysis according to claim 1, characterized in that: By acquiring the operator's operating skill information, analyzing the differences in the operators' operating skills, and obtaining the operating skill difference coefficient, the impact of the operator's operating skill difference on the image quality is measured; The logic for obtaining the operation skill difference coefficient is as follows: Obtain the operator's operating angle when performing endoscopic examination and calculate the operating angle deviation Sθ. The expression is as follows: in It is a vector representation of the actual operating angle, used to capture angle deviations in multiple directions. is the vector representation of the target operation angle; It is the vector dot product operation of the actual operation angle and the target operation angle; is the vector magnitude of the target operation angle, is the modulus of the vector representation of the actual operating angle; obtain the operating depth data of the operator during endoscopic examination, and calculate the average operating depth μd, the expression is as follows where sd n Represents the nth depth operation value, n∈{1,2,...,N}, calculates the depth volatility SH, the expression is as follows Where sdmax is the maximum value of the operating depth, and sdmin is the minimum value of the operating depth; obtain the moving speed of the endoscopic device when the operator performs endoscopic examination, and calculate the speed change rate Vrate, which is expressed as follows where v m represents the moving speed of the endoscope device at the mth moment, v m-1 represents the moving speed of the endoscope device at the m-1th moment, m∈{1,2,...,M}, and calculates the trajectory stability coefficient GJ, which is expressed as follows Wherein Vmax is the maximum allowable speed change rate; the operation skill difference coefficient CZX is calculated according to the operation angle offset, depth fluctuation rate, and trajectory stability coefficient, and the expression is as follows CZX=b1*Sθ+b2*SH+b3*GJ, where b1, b2, and b3 represent the preset proportional coefficients of the operation angle offset, depth fluctuation rate, and trajectory stability coefficient, respectively, and b1, b2, and b3 are all greater than 0.

4. The gastroenterology prediction and analysis system based on big data analysis according to claim 1, characterized in that: According to the coefficient of endoscope quality degradation and the coefficient of operation skill difference, an image acquisition decision model is constructed to generate the image acquisition decision index Idecision. The formula of the model is as follows Where c1 and c2 represent the preset proportional coefficients of the endoscopic equipment quality degradation coefficient and the operating skill difference coefficient, respectively, and both c1 and c2 are greater than 0.

5. The gastroenterology prediction and analysis system based on big data analysis according to claim 4, characterized in that: The image acquisition decision index is compared with the preset image acquisition decision index threshold to determine whether to perform endoscopic image re-acquisition, as follows: If the image acquisition decision index is greater than the image acquisition decision index threshold, image re-acquisition is required, and the determination of endoscopic image re-acquisition is repeated according to the image acquisition decision module; if the image acquisition decision index is less than or equal to the image acquisition decision index threshold, the currently acquired endoscopic image is used as the initial endoscopic image, and an initial endoscopic image set is established based on the initial endoscopic image.

6. The gastroenterology prediction and analysis system based on big data analysis according to claim 1, characterized in that: By acquiring color transition information of the initial endoscopic image, analyzing the color transition of the initial endoscopic image, and acquiring a color transition coefficient, the color transition of the initial endoscopic image is measured; The logic for obtaining the color transition coefficient is as follows: Preprocessing the initial endoscopic image, including removing noise and enhancing contrast; converting the preprocessed initial endoscopic image from RGB color space to HSV color space; The Sobel operator is used to calculate the gradient of the initial endoscopic image, as follows: Among them G x is the gradient in the x direction, II is the initial endoscopic image input, Among them G y is the gradient in the y direction; calculate the global average gradient Gavg of the initial endoscopic image, the expression is as follows Among them G x (h) is the gradient in the x direction at the initial endoscopic image pixel h, G y (h) is the gradient in the y direction at the pixel h of the initial endoscopic image, H is the total number of pixels of the initial endoscopic image, h∈{1,2,...,H}; the initial endoscopic image is divided into multiple local regions according to the preset division criteria, and the average gradient Glocal of different local regions is calculated k , calculate the color transition coefficient SGD, the expression is as follows where Glocal k represents the average gradient of the kth local region, k∈{1,2,...,K}.

7. The gastroenterology prediction and analysis system based on big data analysis according to claim 1, characterized in that: By acquiring exposure abnormality information of the initial endoscopic image, analyzing the underexposure of the initial endoscopic image, and acquiring an underexposure abnormality coefficient, the abnormal degree of underexposure of the initial endoscopic image is measured; The logic for obtaining the underexposure anomaly coefficient is as follows: For each pixel of the initial endoscopic image, the brightness value of each pixel is calculated to obtain the brightness value set of all pixels LD = {ld f }={ld1,ld2,...,ld F }, where ld f represents the brightness value of the f-th pixel, f∈{1,2,...,F},ld f The calculation expression is as follows: f =0.299*R f +0.587*G f +0.114*B f , where R f , G f , B f Respectively represent the values ​​of the red, green, and blue channels of the f-th pixel; The brightness value set LD = {ld f }={ld1,ld2,...,ld F } Sort in ascending order to obtain a sorted brightness value sequence, and calculate the value position of the brightness threshold according to the sorted brightness value sequence. The expression is as follows: Position = 0.1*(F-1), where F is the total number of pixels, and Position is the value position of the brightness threshold. If the value position of the brightness threshold is an integer, the brightness value at the position is directly taken as the brightness threshold; if the value position of the brightness threshold is not an integer, the weighted average of the two adjacent values ​​is taken as the brightness threshold; Calculate the underexposure anomaly coefficient QBG, and the expression is as follows where Llow is the luminance threshold, and 1(ld f <Llow) is an indicator function, indicating that it is 1 when ld f <Llow, and 0 otherwise.

8. The gastroenterology prediction and analysis system based on big data analysis according to claim 1, characterized in that: According to the color transition coefficient, underexposure anomaly coefficient, and image acquisition decision index, an image optimization output model is constructed to generate the image optimization output index Iopt. The formula of the model is as follows: Wherein d1 and d2 represent the preset proportional coefficients of the color transition coefficient and the underexposure abnormality coefficient respectively, and both d1 and d2 are greater than 0; The initial endoscopic images in the initial endoscopic image set are sorted from large to small based on the image optimization output index, and the initial endoscopic image ranked first is taken as the optimal image.

Images