X-ray image processing system for enteritis prognosis analysis

Abstract

The invention relates to the technical field of image processing, in particular to an X-ray image processing system for enteritis prognosis analysis, which comprises the following steps of: obtaining a gray level image according to multi-threshold segmentation; obtaining the gray level fluctuation degree of the gray level image according to the histogram of the gray level image; determining a pixel point aggregation degree through connected domain analysis; obtaining an initial lesion probability by combining the gray level fluctuation degree and the pixel point aggregation degree of the gray level image; according to the fitting result of each gray level image, obtaining the probability of existence of intestinal tract gas; adjusting the initial lesion probability according to the intestinal tract gas probability to obtain the lesion probability; obtaining transformation parameters of the gray level image according to the lesion probability; linear transformation is conducted on the X-ray image according to the transformation parameters, and an enhanced intestinal tract X-ray image is obtained and visualized. The contrast ratio between a lesion area and the surroundings is enhanced, the visual effect of shadow caused by gas is weakened, and prognosis analysis of doctors is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and particularly to an X-ray image processing system for enteritis prognosis analysis. Background Art

[0002] Radiation enteritis is a radiation disease, which is a digestive tract injury caused by the human body being exposed to a large amount or for a long time to radiation. It is necessary to comprehensively evaluate the patient's condition and perform treatment intervention. With the progress of technology, the prognosis analysis system for radiation enteritis in hospitals is becoming more and more perfect. It can not only evaluate the severe state of patients but also reflect the data during the treatment process of patients, facilitating doctors to give rehabilitation opinions in a timely manner. Among them, the X-ray examination of the intestine is one of the important evaluation criteria. The accurate shooting and uploading of X-ray images directly affect the accuracy of doctors' disease evaluation and diagnosis. However, since the state of the intestine of patients with radiation enteritis may affect the absorption and leakage of the contrast agent, thereby reducing the contrast effect of the image and weakening the contrast of the image, it not only increases the diagnostic difficulty of doctors but also affects the analysis of the treatment effect. Therefore, it is necessary to enhance the X-ray images collected in the prognosis analysis system of radiation enteritis to increase their contrast and enhance the visualization results. To enhance intestinal X-ray images, it is difficult to avoid image distortion based on the enhancement of the entire image, and the enhancement effect is poor. Summary of the Invention

[0003] The present invention provides an X-ray image processing system for enteritis prognosis analysis to solve the existing problems.

[0004] The following technical solutions are adopted for an X-ray image processing system for enteritis prognosis analysis of the present invention: An embodiment of the present invention provides an X-ray image processing system for enteritis prognosis analysis, and the system includes the following modules: An image preprocessing module, configured to collect intestinal X-ray images of patients with radiation enteritis and perform grayscale processing; divide the obtained grayscale image into multiple grayscale-level images by using multi-threshold segmentation; A lesion probability determination module, configured to determine the grayscale fluctuation degree of multiple grayscale-level images according to the fluctuation of the frequency in the grayscale histogram of the grayscale-level images; quantify the pixel aggregation degree of the grayscale-level images by using the number, size, and overlapping property of the connected domains of the multiple grayscale-level images as indicators; obtain the initial lesion probability of the grayscale-level images by combining the grayscale fluctuation degree and the pixel aggregation degree of the grayscale-level images; A lesion probability adjustment module, configured to perform elliptical fitting according to the shape characteristics of the shadow area formed by intestinal gas and obtain the probability of the presence of intestinal gas in the grayscale-level images according to the fitting result of each grayscale-level image; adjust the initial lesion probability of the grayscale-level images according to the intestinal gas probability, and obtain the adjusted image lesion probability as the final lesion probability; An image enhancement module, configured to obtain transformation parameters for linearly transforming and enhancing each grayscale image according to the lesion probability, and enhance the X-ray image according to the obtained parameters; A lesion segmentation and recognition module, configured to segment the enhanced image.

[0005] Preferably, the specific method for obtaining the grayscale image is as follows: Using a multi-threshold segmentation algorithm to segment the grayscale image to obtain thresholds, and denoting the j-th threshold among them as and , dividing the grayscale values into multiple levels according to the obtained thresholds, which are respectively , where represents the -th threshold, and dividing the grayscale levels of all pixel points whose grayscale values are in the interval into the -th grayscale level. Marking the pixel points obtained in the -th grayscale level as 1, and marking the values of other pixel points as 0, using the obtained binary image as a mask image, and then multiplying the mask image by the grayscale image to obtain the -th grayscale image, and obtaining grayscale images.

[0006] Preferably, determining the overall grayscale fluctuation degree of multiple grayscale images according to the fluctuation of the frequency in the grayscale histogram of the grayscale image includes the following specific method: Obtaining the grayscale histogram of each grayscale image, and quantifying the grayscale fluctuation degree of each grayscale image according to the variance of the grayscale value frequency in the grayscale image histogram. The specific formula is as follows: where represents the -th segmentation threshold; represents the grayscale fluctuation degree of the -th grayscale image; represents the -th grayscale value in the interval corresponding to the -th grayscale image, and , represents the number of grayscale values included in the interval , ; represents the grayscale average value in the -th grayscale image; represents in the The occurrence frequency of the gray value in the histogram of the gray-level image.

[0007] Preferably, the aggregation degree of pixel points of the gray-level image is quantified according to the number, size, and overlap of connected regions of multiple gray-level images; the initial lesion probability of the gray-level image is obtained by combining the gray-level fluctuation degree and the aggregation degree of pixel points, and the specific formula is: Among them, represents the probability of the existence of a lesion area in the th gray-level image; represents the pixel gray-level fluctuation degree in the th gray-level image; represents the number of connected regions in the th gray-level image; represents the average area of the connected regions formed in the th gray-level image; represents the average value of the total central distance formed between pairwise connected regions in the th gray-level image; represents the aggregation degree of pixel points quantified according to the size, number, and overlap of connected regions of each gray-level image; represents the exponential function with the natural constant as the base.

[0008] Preferably, the probability of the existence of intestinal gas in the gray-level image is obtained according to the fitting result of each gray-level image, and the specific formula is: Among them, represents the average ellipse fitting effect of the th gray-level image, represents the best fitting result of the th connected region in the th gray-level image, represents the number of connected regions in the gray-level image and , represents the number of connected regions of the th gray-level image.

[0009] Preferably, the initial lesion probability of the gray-level image is adjusted according to the intestinal gas probability, and the adjusted image lesion probability is obtained as the final lesion probability, and the specific formula is: Among them, Indicates the lesion probability of the th grayscale image adjusted by the probability of intestinal gas presence, Indicates the probability of intestinal gas presence after normalization processing, Indicates the lesion probability of the initial th grayscale image.

[0010] Preferably, the method for obtaining the transformation parameters for linearly transforming and enhancing each grayscale image according to the lesion probability includes the following specific steps: According to the principle that the greater the lesion probability of the grayscale image, the greater the degree of enhancement, and vice versa, to enhance the contrast, it is necessary to reduce the grayscale value of the area with a large lesion probability and increase the grayscale value of the area with a small lesion probability. The specific calculation formula is as follows: Among them, Indicates the transformation coefficient when performing linear transformation and enhancement on the th grayscale function, is a hyperparameter; Indicates the tangent function; Indicates the hyperbolic tangent function.

[0011] Preferably, the method for obtaining the transformation parameters for linearly transforming and enhancing each grayscale image according to the lesion probability and enhancing the X-ray image according to the obtained parameters includes the following specific steps: Construct a linear transformation and enhancement formula according to the transformation parameters of each grayscale image as follows: Among them, Indicates the pixel value of the output image of the th grayscale image, Indicates the grayscale value of the th pixel point in the th grayscale image, is a hyperparameter, taking , is the average grayscale value of the th grayscale image.

[0012] The beneficial effects of the technical solution of the present invention are as follows: By performing gray-level segmentation on the intestinal X-ray images of patients with enteritis, obtaining the images corresponding to each gray level, determining the overall gray-level fluctuation degree of each gray-level image according to the gray-level frequency fluctuation of the gray-level image histogram, and then determining the distribution aggregation degree of the pixel points in the gray-level image and preliminarily determining the probability of lesions in each gray-level image by analyzing the number, size, and overlap of the connected regions in the gray-level image; Considering the influence of intestinal gas, by performing shape fitting on the morphological features of the gas shadow, obtaining the probability of intestinal gas existing in each gray-level image and adjusting the initially determined disease probability by the gas probability to obtain the final lesion probability; Finally, determining the transformation parameters according to the lesion probability and enhancing the image by gray level, improving the adaptability of enhancing the intestinal line image, enhancing the contrast between the lesion area and the surrounding area and weakening the visual effect of the shadow caused by gas. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0014] Figure 1 It is a flowchart of the steps of an X-ray image processing system for enteritis prognosis analysis according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following will, in conjunction with the accompanying drawings and preferred embodiments, describe in detail the specific implementation manner, structure, features, and effects of an X-ray image processing system for enteritis prognosis analysis proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0017] The following will specifically describe the specific solution of an X-ray image processing system for enteritis prognosis analysis provided by the present invention with reference to the accompanying drawings.

[0018] Please refer to Figure 1, which shows the structural block diagram of an X-ray image processing system for enteritis prognosis analysis provided by an embodiment of the present invention. The method includes the following steps: The image preprocessing module 101 collects the intestinal X-ray images of enteritis patients and performs grayscale processing, and divides the images into multiple grayscale images.

[0019] Under the guidance of the doctor's instructions, the correct position is set, the intestinal X-ray image is obtained and preprocessed.

[0020] It should be noted that in order to achieve the regional enhancement of the X-ray image in the embodiment of the present invention, first, the multi-threshold segmentation algorithm is used to divide the grayscale values of the image. The multi-threshold segmentation is a prior art and will not be elaborated here. The calculated threshold is used as the basis for dividing the grayscale interval. The grayscale image is segmented by using the multi-threshold segmentation algorithm to obtain thresholds. In the present invention application, the empirical value J = 9 is taken, and the j-th threshold among them is denoted as and , and according to the obtained thresholds, the grayscale values are divided into multiple levels: , , ,…, ), , where represents the -th threshold, and all pixel points with grayscale values in the interval are divided into the -th grayscale level, so as to obtain grayscale levels.

[0021] It should be noted that since the complexity of the image corresponding to each grayscale level is different, the degree of enhancement for it is also different. Therefore, it is necessary to obtain each grayscale level image to facilitate the analysis of the complexity of its pixel distribution: First, mark the pixel points at the -th grayscale level as 1, and the values of other pixel points as 0. The obtained binary image is used as a mask image, and then the mask image is multiplied by the grayscale image of the intestinal X-ray image to obtain the -th grayscale level image, so as to obtain grayscale images.

[0022] So far, all grayscale levels of the image are obtained by multi-threshold segmentation, and multiple grayscale images are obtained according to the grayscale levels.

[0023] The lesion probability determination module 102 obtains the initial lesion probability of the grayscale image by combining the grayscale fluctuation degree and the pixel point overlap degree of the grayscale image.

[0024] It should be noted that since the gray-scale distribution of normal intestinal X-ray images is relatively uniform, and the gray-scale values of the lesion areas in the X-ray images of patients with radiation enteritis are larger or smaller than those of the normal areas, the complexity of the gray-scale distribution of the images in that gray-scale interval can be reflected according to the frequency fluctuation of the histogram of the gray-scale images. Therefore, in the embodiments of the present invention, the gray-scale distribution fluctuation of each gray-scale image is calculated according to the following formula. Taking the j-th gray-scale image as an example: Obtain the gray-scale mean value of the j-th gray-scale image, denoted as , obtain the gray-scale value of any pixel point i in the j-th gray-scale image, denoted as , obtain the frequency of occurrence of the gray-scale value of any pixel point i in the j-th gray-scale image, and denote it as , denote the number of gray-scale values in the j-th gray-scale image as , and the gray-scale fluctuation degree of the j-th gray-scale image can be calculated according to the following mathematical formula: It should be noted that for the j-th gray-scale image, if the difference between different pixel points and the gray-scale mean value is larger, then the gray-scale value distribution fluctuates more greatly at this time, and the calculated gray-scale fluctuation degree is also larger. At the same time, for the convenience of subsequent calculations, the results of the gray-scale fluctuation degree in this embodiment are linearly normalized to make .

[0025] Then use the above calculation formula to calculate the gray-scale images, obtain the gray-scale fluctuation degree of each gray-scale image as a whole, and provide a basis for further analyzing the local pixel distribution of the gray-scale images later.

[0026] So far, the gray-scale fluctuation degree of each gray-scale image as a whole has been obtained by calculating the fluctuation of the histogram frequency of each gray-scale image.

[0027] Then, by performing connected component analysis on the gray-scale images, the aggregation degree of the pixel point distribution of each gray-scale image is quantified according to the number, size, and overlap of the connected components, and the initial lesion probability of the gray-scale image is determined.

[0028] It should be noted that in the abdominal intestinal X-ray signs, due to the rupture of the gas-containing intestinal loops in patients with enteritis, free gas is generated inside the abdominal cavity, and these free gases appear as low-density aggregated shadow areas in the X-ray images. Therefore, in combination with the obtained gray-scale fluctuation degree of the gray-scale images, the stability of the presence of lesions in each gray-scale image is analyzed, and the probability of local lesions in the gray-scale image is determined by analyzing the aggregation degree of the local pixel points of the gray-scale image.

[0029] It should be noted that in the intestinal X-ray images, the pixel points in the lesion area are relatively dense. Therefore, the connected regions in each grayscale image can be obtained by using connected component analysis first, and then the aggregation degree of local pixel points in the grayscale image can be quantified by analyzing the number of connected components contained in the grayscale image, the corresponding size of the connected components, and the overlap of the connected components.

[0030] It should be noted that when judging the aggregation degree of local pixel points in the grayscale image according to the distribution of connected components in the grayscale image, if the aggregation degree of pixel points is relatively high, it indicates that there may be a large number of connected components with small areas in the grayscale image, and due to the large number of connected component distributions, the overlapping area caused is also large. Therefore, in the embodiments of the present invention, the number of connected components, the area, and the overlap between connected components are used as indicators to adjust the influence of the overall gray level fluctuation degree on the initial lesion probability judgment, so as to enhance the stability of the lesion probability judgment. The calculation formula is as follows: Among them, represents the probability of the existence of a lesion area in the th grayscale image; represents the pixel gray level fluctuation degree in the th grayscale image; represents the number of connected components in the th grayscale image; represents the average area of the connected components formed in the th grayscale image; represents the average value of the total central distance formed between every two connected components in the th grayscale image. If the central distance is smaller, it indicates that the connected components are closer to each other. On the contrary, the connected components are more dispersed, which is used to quantify the overlap between connected components; represents the ratio of the connected component area to the number, and through linear normalization, .

[0031] It should be noted that in the embodiments of the present invention, the change rate of the lesion probability is adjusted by . If is larger, it indicates that the connected component area is larger and the distribution is less, then the pixel points in the grayscale image are more discrete. Then, as the central distance formed between adjacent connected components in the th grayscale image increases, the decreasing rate of the pixel point aggregation degree is more rapid. At this time, for these grayscale images, due to the influence of their surface features, the probability of the existence of a lesion area in itself may not be high; on the contrary, when is smaller, it indicates that there are a large number of small connected components distributed in the image, and the pixel points in the image are more aggregated. Therefore, as the central distance between adjacent connected components in the image The longer it is, the slower the decrease in the aggregation degree of pixel points, and the higher the probability that there are lesions in these grayscale images themselves. Therefore, in the embodiments of the present invention, the accuracy of judging whether there are lesions in an image is improved by the aggregation degree of pixel points in the grayscale image.

[0032] Thus, the initial lesion probability of each grayscale image is obtained.

[0033] The lesion probability adjustment module 103 adjusts the initial lesion probability of the image according to the existence probability of intestinal gas in each level of grayscale image.

[0034] It should be noted that when enhancing the intestinal X-ray image of a patient with radiation enteritis, it is necessary to consider the influence of the gas existing in the intestine itself on determining the lesion area. Because the tiny area formed by the gas may be regarded as the lesion area, in order to weaken the influence of the gas area on the lesion area, in the embodiments of the present invention, according to the characteristic that the aggregation of intestinal gas usually forms an approximate circular or elliptical shadow, the least squares method is used to perform elliptical fitting on each connected domain in the grayscale image, and then the fitting effect is used as the probability of the existence of intestinal gas to adjust the lesion probability calculated in the above steps, so as to improve the judgment accuracy of the lesion area.

[0035] First, the least squares method is used to perform elliptical fitting on the connected domain of the grayscale image. Assume that the equation of the fitted ellipse is expressed as: where are the horizontal and vertical coordinates of the ellipse equation, are the parameters of the ellipse equation.

[0036] Denote the horizontal and vertical coordinates of the th pixel point of the th connected domain in the th grayscale image as . According to the least squares principle, the objective function of the fitting result of the th connected domain in the th grayscale image is: Then, when reaches the minimum, the fitting effect is the best. Therefore, make the partial derivatives of equal to 0, that is, when , the corresponding ellipse is the best-fitting ellipse, and the relevant parameters of the ellipse can be obtained. Denote the value when reaches the minimum as , denoted as the best ellipse fitting effect of the th connected domain in the th grayscale image.

[0037] It should be noted that since the smaller the value of , the better the effect of ellipse fitting. At this time, the probability that the corresponding connected region is the intestinal gas shadow area is greater. Also, since intestinal gas usually presents as gray or black shadows, the probability that it is divided into the same gray-level image is relatively large. Therefore, in the embodiments of the present invention, the probability of the presence of intestinal gas in each gray-level image is obtained by calculating the average value of the best fitting results: Among them, represents the average ellipse fitting effect of the -th gray-level image, represents the best fitting result of the -th connected region in the -th gray-level image, represents the number of connected regions in the gray-level image and , represents the number of connected regions in the -th gray-level image.

[0038] It should be noted that if the value of is smaller, it indicates that the effect of ellipse fitting is better. At this time, the probability of the presence of gas shadows in the corresponding gray-level image is greater. If the calculated initial lesion probability is relatively large, then due to the relatively large probability of the presence of gas, the possibility of misidentifying the gas region as a lesion region is greater. Therefore, it is necessary to reduce the lesion probability of this gray-level image. Therefore, it is necessary to adjust the initial lesion probability according to the probability of the presence of intestinal gas. In the embodiments of the present invention, the lesion probability of each gray-level image is adjusted in the following form: Among them, represents the lesion probability of the -th gray-level image adjusted by using the probability of the presence of intestinal gas, represents the probability of the presence of intestinal gas after normalization processing, represents the initial lesion probability of the -th gray-level image.

[0039] It should be noted that multiplying by the initial lesion probability to adjust the lesion probability. First, it can be clearly seen that due to the influence of intestinal gas, the probability of the presence of lesions in each gray-level image is correspondingly reduced. At the same time The smaller it is, the better the fitting effect of the connected region of the grayscale image, the greater the probability of the existence of intestinal gas, the greater the corresponding adjustment of the initial lesion probability, and the greater the reduction of the initial lesion probability through multiplication processing. Through the above adjustment formula, the influence of the shadow formed by intestinal gas on the lesion area is effectively weakened, and the over-enhancement of the gas shadow area is prevented.

[0040] Thus, the lesion probability of each grayscale image adjusted according to the gas existence probability is obtained.

[0041] The image enhancement module 104 determines the transformation parameters for the linear transformation enhancement of the grayscale image according to the lesion probability, and performs linear transformation enhancement on the image.

[0042] It should be noted that the greater the lesion probability of the grayscale image, the greater the degree of enhancement required. To make the manifestation of the lesion area more obvious, the pixel values of the area with a large lesion probability are increased through linear transformation; however, for the area with a small lesion probability, since the probability of the shadow formed by intestinal gas is large, it is necessary to reduce the grayscale value of the image with a sufficiently small lesion probability according to the linear transformation, so as to enhance the visual effect of the gas shadow area and prevent the influence on the judgment of the lesion area. Therefore, the embodiments of the present invention construct the following relational expression to determine the transformation parameters: Among them, represents the transformation coefficient when the th grayscale function is linearly transformed and enhanced, is a hyperparameter, representing the limitation on the size of the transformation coefficient. Usually, the value range for adjusting the image contrast is , so the embodiments of the present invention are described with , which is not specifically limited in this embodiment, where can be determined according to the specific implementation situation; a functional relationship between the transformation coefficient and the lesion probability is constructed using the hyperbolic tangent function

[0043] Finally, according to the obtained transformation parameters of each grayscale image, the intestinal X-ray image is enhanced according to the linear transformation algorithm, and the transformation formula is as follows: Among them, represents the pixel value of the output image of the th grayscale image, represents the th grayscale value of the th pixel point in the , is the average gray value of the n-th grayscale image.

[0044] It should be noted that the intestinal X-ray images are enhanced by linear transformation to different degrees according to grayscale levels, which enhances the adaptability of the linear transformation algorithm for enhancing intestinal X-ray images, prevents over-enhancement or under-enhancement of the images, and can provide doctors with clearer intestinal X-ray images to guide their judgment of the patient's condition and recovery.

[0045] So far, the intestinal X-ray images enhanced by the improved linear transformation enhancement algorithm have been obtained.

[0046] The lesion segmentation and recognition module 105 segments and recognizes lesions based on the enhanced intestinal X-ray images to assist doctors in prognostic analysis.

[0047] It should be noted that after the adaptive linear transformation enhancement of the X-ray images, the contrast between the lesion area and the normal area is enhanced. At this time, the intestinal lesion area can be segmented more completely and accurately, so as to obtain a more accurate lesion area, providing a basis for doctors to conduct prognostic analysis. At the same time, the X-ray images should be saved in the system so that doctors can view and compare the patient's recovery in a timely manner and analyze the condition.

[0048] Specifically, in this embodiment, a manual interaction method is used to segment the X-ray images after adaptive linear transformation enhancement to ensure the accuracy of obtaining the lesion area.

[0049] Through the above steps, the enhancement processing of the X-ray images before the prognostic analysis of enteritis is completed.

[0050] In the embodiment of the present invention, the intestinal X-ray images of enteritis patients are segmented according to grayscale levels to obtain the images corresponding to each grayscale level. The gray fluctuation degree of each grayscale image as a whole is determined according to the gray frequency fluctuation of the grayscale image histogram. Then, the distribution aggregation degree of the pixel points in the grayscale image is determined by analyzing the number, size and overlap of the connected domains in the grayscale image, and the probability of lesions existing in each grayscale image is initially determined; considering the influence of intestinal gas, by shape fitting the morphological characteristics of the gas shadow, the probability of intestinal gas existing in each grayscale image is obtained, and the initially determined disease probability is adjusted by the gas probability to obtain the final lesion probability; finally, the transformation parameters are determined according to the lesion probability and the images are enhanced according to grayscale levels, improving the adaptability of enhancing the intestinal X-ray images, enhancing the contrast between the lesion area and the surrounding area and weakening the visual effect of the shadow caused by gas. Doctors can conduct prognostic analysis based on the X-ray images of enteritis patients with stronger contrast.

[0051] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An X-ray image processing system for enteritis prognosis analysis, characterized in that, The system includes the following modules: An image preprocessing module, which is used to collect the intestinal X-ray images of patients with radiation enteritis and perform grayscale processing to obtain grayscale images of intestinal X-ray images; and divide them into multiple grayscale images by using multi-threshold segmentation; A lesion probability determination module, which is used to determine the gray level fluctuation degree of multiple grayscale images according to the fluctuation of the frequency in the gray level histogram of the grayscale images; Quantify the pixel aggregation degree of the grayscale images by taking the number, size and overlap of the connected regions of multiple grayscale images as indicators; and obtain the initial lesion probability of the grayscale images by combining the gray level fluctuation degree and the pixel aggregation degree of the grayscale images; A lesion probability adjustment module, which is used to perform elliptical fitting according to the shape characteristics of the shadow area formed by intestinal gas and obtain the probability of the existence of intestinal gas in the grayscale images according to the fitting results of each grayscale image; adjust the initial lesion probability of the grayscale images according to the intestinal gas probability, and obtain the adjusted image lesion probability as the final lesion probability; An image enhancement module, which is used to obtain the transformation parameters for linearly transforming and enhancing each grayscale image according to the lesion probability and enhance the X-ray image according to the obtained parameters; A lesion segmentation and recognition module, which is used to segment the enhanced image.

2. The X-ray image processing system for enteritis prognosis analysis according to claim 1, wherein The method of using multi-threshold segmentation to divide into multiple grayscale images specifically includes: Use the multi-threshold segmentation algorithm to obtain a mask image for the grayscale image of the intestinal X-ray image, and multiply the mask image by the grayscale image of the intestinal X-ray image to obtain multiple grayscale images.

3. The X-ray image processing system for enteritis prognosis analysis according to claim 2, wherein The multi-threshold segmentation algorithm specifically includes: Set the number of thresholds of the multi-threshold segmentation algorithm to J, where J is a first preset value; Denote the j-th threshold among them as and , and divide the gray values into multiple levels: , , ,…, ), , where represents the -th threshold, and divide the gray levels of all pixel points whose gray values are in the interval into the -th gray level; Mark the obtained pixel points at the -level gray level as 1, and mark the values of other pixel points as 0.

4. The X-ray image processing system for enteritis prognosis analysis according to claim 1, wherein The method for specifically obtaining the gray level fluctuation degree of multiple grayscale images according to the fluctuation of the frequency in the gray level histogram of the grayscale images includes: Obtain the gray level mean value in each grayscale image, record the difference between the gray level value of each pixel point in each grayscale image and the gray level mean value as the first difference, obtain the frequency of each gray level value in each grayscale image, and record it as the first frequency, record the product of the square of the first difference and the first frequency as the first product, and record the cumulative sum of the first products as the gray level fluctuation degree.

5. The X-ray image processing system for enteritis prognosis analysis according to claim 1, wherein, Quantify the pixel aggregation degree of the grayscale images by taking the number, size and overlap of the connected regions of multiple grayscale images as indicators; and obtain the initial lesion probability of the grayscale images by combining the gray level fluctuation degree and the pixel aggregation degree of the grayscale images, specifically including: Record the ratio of the average area of the connected regions of each grayscale image to the number of connected regions of each grayscale image as the first ratio, obtain the average value of the distances between any two connected regions in each grayscale image, and record it as the first mean value, calculate the value of the negative exponential function with the natural constant e as the base by using the product of the first mean value and the first ratio, and record it as the first function value, and record the product of the gray level fluctuation degree of each grayscale image and the first function value as the initial lesion probability.

6. The X-ray image processing system for enteritis prognosis analysis according to claim 1, wherein The specific formula for obtaining the probability of the existence of intestinal gas in the grayscale images according to the fitting results of each grayscale image includes: Among them, represents the average elliptical fitting effect of the th grayscale image, represents the best fitting result of the th connected component in the th grayscale image, represents the number of connected components in the grayscale image and , represents the number of connected components of the th grayscale image.

7. The X-ray image processing system for enteritis prognosis analysis according to claim 1, characterized in that, Adjusting the initial lesion probability of the grayscale image according to the intestinal gas probability, and obtaining the adjusted image lesion probability as the final lesion probability, the specific method included is as follows: Among them, represents the lesion probability of the th grayscale image adjusted by the intestinal gas presence probability, represents the intestinal gas presence probability after normalization processing, represents the lesion probability of the initial th grayscale image.

8. The X-ray image processing system for enteritis prognosis analysis according to claim 1, wherein Obtaining the transformation parameters for linearly transforming and enhancing each grayscale image according to the lesion probability, the specific formula included is as follows: Among them, represents the transformation coefficient when performing linear transformation enhancement on the th gray-level function, being a hyperparameter; represents the tangent function; represents the hyperbolic tangent function.

9. The X-ray image processing system for enteritis prognosis analysis according to claim 1, characterized in that, Obtaining the transformation parameters for linearly transforming and enhancing each grayscale image according to the lesion probability and enhancing the X-ray image according to the obtained parameters, the specific method included is as follows: Constructing the linear transformation enhancement formula according to the transformation parameters of each grayscale image as follows: Among them, represents the pixel value of the output image of the grayscale image, represents the th grayscale value of the is a hyperparameter, taking , is the average grayscale value of the th grayscale image.

Images