Liver CT image enhancement processing method and system

Abstract

The present invention discloses a method and system for enhancing liver CT images. The method includes: collecting N liver CT images of a specific liver taken from different cross-sections in the plain scan phase and the enhanced phase respectively, where 50 ≤ N ≤ 200; respectively processing and enhancing m liver CT images in the plain scan phase and m liver CT images in the enhanced phase by the filtered back projection method to obtain A-class enhanced images in the plain scan phase and A-class enhanced images in the enhanced phase, where m < N; respectively processing and enhancing the remaining k liver CT images in the plain scan phase and k liver CT images in the enhanced phase by the adaptive statistical iterative reconstruction method to obtain B-class enhanced images in the plain scan phase and B-class enhanced images in the enhanced phase, where k = N - m; and respectively integrating the A-class enhanced images in the plain scan phase with the B-class enhanced images in the plain scan phase, and integrating the A-class enhanced images in the enhanced phase with the B-class enhanced images in the enhanced phase, and jointly constructing a three-dimensional liver model.

Description

Technical Field

[0001] The present invention relates to a medical image processing system, and in particular to a liver CT image processing system and method. Background Art

[0002] In recent years, CT has become a common imaging method for abdominal images, and the number of patients undergoing CT examinations has increased rapidly each year. Therefore, the current development trend of CT is to obtain clearer images and avoid patients from being exposed to radiation multiple times by improving CT image enhancement processing technology.

[0003] At present, the enhancement methods commonly used in medical image enhancement processing include filtered back projection (FBP), adaptive statistical iterative reconstruction (ASiR), and model-based iterative reconstruction (VEO). Among them, filtered back projection is the most widely used, with high noise index and radiation dose, ignoring the problem that the original data is affected by quantum noise and electronic noise during the acquisition process, and even amplifying the noise of the original image when normalizing the data, resulting in linear artifacts and artificial artifacts. Adaptive statistical iterative reconstruction (ASiR) and model-based iterative reconstruction (VEO) can significantly reduce noise while improving spatial resolution. However, the enhancement process is time-consuming and inefficient, and is not suitable for processing a large number of CT images.

[0004] For example, a method for image enhancement disclosed in Chinese patent application No. 20131012340.3 includes: inputting an image, filtering the image with a first type of filter to obtain a filter map, and obtaining a first high-frequency information map based on the image and the filter map; performing normalized linear structure detection on the filter map, and performing threshold transformation to obtain an enhanced position map, filtering the image with a second type of filter to obtain an enhanced position map; and obtaining an enhanced map of the image based on the first high-frequency information map, the noise map enhancement position map, and the enhancement coefficient map. This patent application is based on linear structure detection and distance transformation, and can enhance images without amplifying noise. However, the image enhancement method of this patent application undergoes two filtering processes, and the transformation process will cause the image to lose some pixel information, affecting the restoration of the final result.

[0005] For another example, a method and apparatus for image enhancement disclosed in Chinese Patent No. 20131004602.9. The image enhancement method includes the steps of: extracting the dark and bright regions of the original image, and enhancing the average intensity of the local details in the dark and bright regions, adjusting the overall measure of the obtained image with enhanced local details to obtain an image with adjusted overall brightness; and adjusting the saturation of the image with adjusted overall brightness. This patent application can achieve a better image enhancement effect, making the enhanced image more in line with human visual characteristics. However, the image enhancement method and apparatus of this patent application cannot meet the accuracy requirements of medical CT images, and the obtained images are relatively rough and not conducive to doctors observing and determining lesions.

[0006] For another example, a digital image enhancement disclosed in Chinese Patent Application No. 20088000010.7. The enhanced image includes adjusting its brightness, and then adjusting the contrast of the image with adjusted brightness. Adjusting the brightness is using an estimate of the real-world brightness. Contrast adjustment involves extracting background and contrast information, adjusting the contrast information based on the background information through an adjustment stamp, and combining the background and corrected contrast information to obtain an output image. However, the digital image enhancement of this patent application is achieved by adjusting brightness and contrast, and the obtained enhanced image has low clarity and is not conducive to doctors observing and determining lesions.

[0007] Therefore, it is an urgent problem in the industry to provide a method and system for enhancing liver CT images that can reduce the radiation dose while ensuring the imaging quality remains unchanged. Summary of the Invention

[0008] The object of the present invention is to provide a method and system for enhancing liver CT images, which can shorten the time-consuming of the enhancement process while ensuring the imaging quality remains unchanged.

[0009] According to one aspect of the present invention, a method for enhancing liver CT images is provided, including the following steps: (1) collecting N liver CT images of a specific liver taken from different cross-sections in the plain scan period and the enhanced period respectively, where 50 ≤ N ≤ 200; (2) respectively processing and enhancing m liver CT images in the plain scan period and m liver CT images in the enhanced period by the filtered back-projection method to obtain A-class enhanced images in the plain scan period and A-class enhanced images in the enhanced period, where m < N; (3) respectively processing and enhancing the remaining k liver CT images in the plain scan period and k liver CT images in the enhanced period by the adaptive statistical iterative reconstruction method to obtain B-class enhanced images in the plain scan period and B-class enhanced images in the enhanced period, where k = N - m; and (4) respectively integrating the A-class enhanced images in the plain scan period with the B-class enhanced images in the plain scan period, and integrating the A-class enhanced images in the enhanced period with the B-class enhanced images in the enhanced period, and jointly constructing a three-dimensional liver model.

[0010] Optionally, the liver CT image during the plain scan period is obtained by combining CT positioning with a scanning range determined by ultrasound or other imaging.

[0011] Optionally, the scanning tube voltage of the liver CT image in the enhancement period is 80 and 140 mAs (irradiation dose per unit time), and the pitch is set to 1.375.

[0012] Optionally, the image slice thickness of the liver CT image is set to 0.5-0.8 mm. Optimally, the image slice thickness of the liver CT image is set to 0.625 mm.

[0013] Optionally, m:k=1:1~5.

[0014] Preferably, m:k=1:1.

[0015] Optionally, step (2) includes: (2.1) selecting any angle as the initial angle Initial angle (2.2) performing a weighted operation on the one-dimensional Fourier transform result obtained in step (2.1) to obtain a weighted result; (2.3) performing a one-dimensional inverse Fourier transform on the weighted result obtained in step (2.2) to obtain a modified projection image; (2.4) performing a back-projection on the modified projection image obtained in step (2.3) to obtain a back-projection image; (2.5) performing a back-projection operation on the initial angle The starting point is transformed by 5 to 10 degrees, and the process of steps (2.1) to (2.4) is repeated until the 180-degree transformation back projection is completed, and the overlapping parts of all the back-projected images obtained are used as the reconstructed image.

[0016] Optionally, step (3) includes: (3.1) establishing a square on each liver CT image, wherein the side length of the square is the longest distance of the liver region on the corresponding liver CT image to cover the liver region, establishing a coordinate system with the center of the square as the origin, setting the image function of the square as a binary function in the coordinate system, assigning the function value of the binary function outside the square to 0, and dividing the square into N 2 There are N equal small squares, each of which is one pixel. 2 Assign values ​​from 1 to J to pixels, and set the grayscale value of pixel j to x j , where 1≤j≤J, J=N 2 ; (3.2) for x j Assign an initial value, set the correction index C and the convergence value D, and j Do ART iteration and get the i-th iteration result x j i, i>1; (3.3) Compare the i-th iteration result obtained in step (3.2) with iteration x j i With the initial value x j If the difference exceeds the convergence value D, update the initial value x j , where the updated x j For x j =x j i +C; (3.4) Update the updated x obtained in step (3.3) j Bring it into the next iteration and repeat the process of steps (3.2) and (3.3) until the jth component x of the kth iteration is j (k) When it is less than or equal to the convergence value D, it stops and the adaptive statistical iterative reconstruction is completed.

[0017] Among them, for a certain reconstruction perspective, the adaptive statistical iterative reconstruction method first calculates a comprehensive projection by "front and back projection" on the estimated object image. This is the first estimate of the attenuation along the perspective, but there is a large error. This estimate may simulate the process of X-ray precursors passing through objects in a real CT system. The limited focus size is simulated by setting the initial position of the X-ray precursors in a small area rather than a single point. In the process of modeling the interaction between X-ray photons and objects, the size and size of the reconstructed pixels are considered by calculating the path length of photons entering the voxel in slightly different directions and positions.

[0018] Optionally, before step (2), the method further includes: directly performing computation processing on the pixel grayscale values ​​of each liver CT image in the spatial domain using a spatial domain method to filter out noise in each liver CT image.

[0019] According to another aspect of the present invention, there is provided a liver CT image enhancement processing system, comprising: a data acquisition module for obtaining N liver CT images of a specific liver taken from different sections at the plain scan stage and the enhanced stage respectively, where 50 ≤ N ≤ 200; and an image enhancement module communicatively connected to the data acquisition module, wherein the image enhancement module includes a filtering enhancement unit, an iterative reconstruction enhancement unit, and an integration unit. The filtering enhancement unit is configured to process and enhance m liver CT images at the plain scan stage and m liver CT images at the enhanced stage by using the filtered back projection method to obtain A-class enhanced images at the plain scan stage and A-class enhanced images at the enhanced stage, where m < N. The iterative reconstruction enhancement unit is configured to process and enhance the remaining k liver CT images at the plain scan stage and k liver CT images at the enhanced stage by using the adaptive statistical iterative reconstruction method to obtain B-class enhanced images at the plain scan stage and B-class enhanced images at the enhanced stage, where k = N - m. The integration unit is communicatively connected to the filtering enhancement unit and the iterative reconstruction enhancement unit respectively, and is configured to integrate the A-class enhanced images at the plain scan stage with the B-class enhanced images at the plain scan stage, and integrate the A-class enhanced images at the enhanced stage with the B-class enhanced images at the enhanced stage, and then construct a liver three-dimensional model.

[0020] Optionally, the filtering enhancement unit includes: a projection subunit, a weighting subunit, a correction subunit, a back projection subunit, and a reconstruction subunit that are communicatively connected in sequence, where the projection subunit is configured to perform ray projection on the pixels of the liver CT image at a selected initial angle to obtain a projection image, and perform one-dimensional Fourier transform on the projection image to obtain a one-dimensional Fourier transform result; the weighting subunit is configured to perform a weighting operation on the obtained one-dimensional Fourier transform result to obtain a weighted result; the correction subunit is configured to perform one-dimensional inverse Fourier transform on the obtained weighted result to obtain a corrected projection image; the back projection subunit is configured to perform back projection on the obtained corrected projection image to obtain a back projection image; the reconstruction subunit is configured to, after starting from the initial angle and changing 5 - 10 degrees each time until 180 degrees of transformation and back projection are completed, use the overlapping part of all the obtained back projection images as the reconstructed image.

[0021] Optionally, the iterative reconstruction enhancement unit includes: a function setting unit, an assignment iteration unit, a comparison update unit, and an iterative reconstruction unit that are communicatively connected in sequence. The function setting unit is configured to establish a square on each liver CT image, the side length of the square is the longest distance of the liver area on the corresponding liver CT image to cover the liver area, establish a coordinate system with the center of the square as the origin, set the image function of the square as a binary function in the coordinate system, assign the function value outside the square to 0, divide the square into N 2 equal small squares, each small square is a pixel, for N2 Assign values ​​from 1 to J to pixels, and set the grayscale value of pixel j to x j , where 1≤j≤J, J=N 2 ; The assignment iteration unit is used to assign x j Assign an initial value, set the correction index C and the convergence value D, and j Do ART iteration and get the i-th iteration result x j i , i>1; the comparison update unit is used to compare the i-th iteration result iteration x obtained by the assignment iteration unit j i With the initial value x j If the difference exceeds the convergence value D, update the initial value x j , where the updated x j For x j =x j i +C; the iterative reconstruction unit is used to compare the updated x obtained by the update unit j Bring it into the next iteration, so that the assignment iteration unit and the comparison update unit repeat the work until the jth component x of the kth iteration j (k) When it is less than or equal to the convergence value D, it stops and the adaptive statistical iterative reconstruction is completed.

[0022] Optionally, the correction index C and / or the convergence value D may be set to 0 or 1.

[0023] Optionally, the system further includes an image smoothing module that is communicatively connected to the data acquisition module and the image enhancement module and is used to filter out noise in each liver CT image.

[0024] Optionally, it also includes: a liver information database and a retrieval module, wherein the liver information database is communicatively connected with the data acquisition module and the image enhancement module, and is used to store a plurality of liver CT images and corresponding liver three-dimensional models; the retrieval module is communicatively connected with the liver information database, and is used for allowing a user to input retrieval information to retrieve a target CT image and a corresponding liver three-dimensional model in the liver information database.

[0025] The beneficial effects of the present invention are as follows: (1) By combining the filtered back projection method and the adaptive statistical iterative reconstruction method to enhance liver CT images, while ensuring the imaging quality remains unchanged, the radiation dose can be reduced by 60% - 80%, and the time-consuming of the enhancement process can be shortened. The enhancement process of all liver CT images can be completed within 2 minutes, improving work efficiency. (2) It serves as a basis for subsequent three-dimensional reconstruction of the liver, enabling the acquisition of more reducible and clearer three-dimensional liver images, providing new ideas for the research and development in the field of liver CT. (3) The liver information database includes a large amount of liver data. Through an open platform for sharing, experts around the world can share digital liver data and discuss cases, which is conducive to communication among doctors and sharing of knowledge, and conforms to the national Internet + strategic direction. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 FIG. shows a schematic flow chart of the method for enhancing liver CT images of the present invention.

[0027] Figure 2 FIG. shows a schematic structural diagram of the system for enhancing liver CT images of the present invention DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present invention, and should not be construed as a limitation to the present invention.

[0029] First, please refer to Figure 1 , as a non-limiting embodiment, the present invention provides a method for enhancing liver CT images. First, in step S1, N liver CT images of a specific liver taken from different cross-sections in the plain scan period and the enhanced period are collected, where 50 ≤ N ≤ 200. Then, in step S2, m liver CT images in the plain scan period and m liver CT images in the enhanced period are respectively processed and enhanced by the filtered back projection method to obtain A-class enhanced images in the plain scan period and A-class enhanced images in the enhanced period, where m < N. Subsequently, in step S3, the remaining k liver CT images in the plain scan period and k liver CT images in the enhanced period are respectively processed and enhanced by the adaptive statistical iterative reconstruction method to obtain B-class enhanced images in the plain scan period and B-class enhanced images in the enhanced period, where k = N - m. Finally, in step S4, the A-class enhanced images in the plain scan period and the B-class enhanced images in the plain scan period are respectively integrated, and after integrating the A-class enhanced images in the enhanced period and the B-class enhanced images in the enhanced period, a three-dimensional liver model is jointly constructed.

[0030] Specifically, step S2 includes: (2.1) Selecting an arbitrary angle as the initial angle Initial angle (2.2) performing a weighted operation on the one-dimensional Fourier transform result obtained in step (2.1) to obtain a weighted result; (2.3) performing a one-dimensional inverse Fourier transform on the weighted result obtained in step (2.2) to obtain a modified projection image; (2.4) performing a back-projection on the modified projection image obtained in step (2.3) to obtain a back-projection image; (2.5) performing a back-projection operation on the initial angle The starting point is transformed by 5 to 10 degrees, and the process of steps (2.1) to (2.4) is repeated until the 180-degree transformation back projection is completed, and the overlapping parts of all the back-projected images obtained are used as the reconstructed image.

[0031] Step S3 includes: (3.1) establishing a square on each liver CT image, wherein the side length of the square is the longest distance of the liver region on the corresponding liver CT image to cover the liver region, establishing a coordinate system with the center of the square as the origin, setting the image function of the square as a binary function in the coordinate system, assigning the function value of the binary function outside the square to 0, and dividing the square into N 2 There are N equal small squares, each of which is a pixel. 2 Assign values ​​from 1 to J to pixels, and set the grayscale value of pixel j to x j , where 1≤j≤J, J=N 2 ; (3.2) for x j Assign an initial value, set the correction index C and the convergence value D, and j Perform ART iteration (algebraic reconstruction method) to obtain the i-th iteration result x j i , i>1; (3.3) Compare the i-th iteration result obtained in step (3.2) with iteration x j i With the initial value x j If the difference exceeds the convergence value D, update the initial value x j , where the updated x j For x j =x j i +C; (3.4) Update the updated x obtained in step (3.3) j Bring it into the next iteration and repeat the process of steps (3.2) and (3.3) until the jth component x of the kth iteration is j (k) When it is less than or equal to the convergence value D, it stops and the adaptive statistical iterative reconstruction is completed.

[0032] As an optional implementation, before step S2, the method further includes: directly performing calculation processing on the pixel grayscale value of each liver CT image in the spatial domain by using a spatial domain method to filter out noise in each liver CT image.

[0033] Based on the same invention concept, please refer to Figure 2 The present invention also provides a liver CT image enhancement processing system, comprising: a data acquisition module 10, an image enhancement module 20, and an image smoothing module 30.

[0034] The data acquisition module 10 is used to obtain 100 liver CT images taken from different cross sections in the plain scan phase and the enhanced phase for a specific liver.

[0035] The image enhancement module 20 includes a filtering enhancement unit 210, an iterative reconstruction enhancement unit 220, and an integration unit 230. The filtering enhancement unit 210 is used to process and enhance the 50 liver CT images in the plain scan period and the 50 liver CT images in the enhancement period by using the filtering back projection method to obtain the plain scan period type A enhanced image and the enhancement period type A enhanced image. The iterative reconstruction enhancement unit 220 is used to process and enhance the remaining 50 liver CT images in the plain scan period and the 50 liver CT images in the enhancement period by using the adaptive statistical iterative reconstruction method to obtain the plain scan period type B enhanced image and the enhancement period type B enhanced image. The integration unit 230 is used to integrate the plain scan period type A enhanced image and the plain scan period type B enhanced image, respectively, and after integrating the enhancement period type A enhanced image and the enhancement period type B enhanced image, construct a three-dimensional liver model.

[0036] As an optional implementation, the filter enhancement unit 210 includes: a projection subunit 2101, a weighting subunit 2102, a correction subunit 2103, a back-projection subunit 2104, and a reconstruction subunit 2105. The projection subunit 2101 is used to select the initial angle The pixels on the liver CT image under the condition of ray projection are projected to obtain a projection image, and a one-dimensional Fourier transform is performed on the projection image to obtain a one-dimensional Fourier transform result. The weighting subunit 2102 is used to perform a weighted operation on the obtained one-dimensional Fourier transform result to obtain a weighted result. The correction subunit 2103 is used to perform a one-dimensional inverse Fourier transform on the obtained weighted result to obtain a corrected projection image. The back-projection subunit 2104 is used to perform back-projection on the obtained corrected projection image to obtain a back-projection image. The reconstruction subunit 2105 is used to reconstruct the image after passing through the initial angle. The transformation is performed 36 times with 5 degrees as the starting point each time until the 180-degree transformation back-projection is completed, and the overlapping parts of all the back-projected images are used as the reconstructed image.

[0037] The iterative reconstruction enhancement unit 220 includes: a function setting unit 2201, an assignment iteration unit 2202, a comparison and update unit 2203, and an iterative reconstruction unit 2204. The function setting unit 2201 is used to establish a square on each liver CT image, the side length of the square is the longest distance of the liver area on the corresponding liver CT image to cover the liver area, and a coordinate system is established with the center of the square as the origin. In the coordinate system, the image function of the square is set as a binary function, and the function value of the binary function outside the square is assigned to 0, and the square is divided into N 2 There are N equal small squares, each of which is a pixel. 2 Assign values ​​from 1 to J to pixels, and set the grayscale value of pixel j to x j , where 1≤j≤J, J=N 2 The assignment iteration unit 2202 is used to assign x j Assign an initial value, set the correction index C and the convergence value D, and j Do ART iteration and get the i-th iteration result x j i , i>1. The comparison and update unit 2203 is used to compare the i-th iteration result iteration x obtained by the assignment iteration unit j i With the initial value x j If the difference exceeds the convergence value D, update the initial value x j , where the updated x j For x j =x j i + C. The iterative reconstruction unit 2204 is used to update the updated x obtained by the comparison and update unit j Bring it into the next iteration, so that the assignment iteration unit and the comparison update unit repeat the work until the jth component x of the kth iteration j (k) When it is less than or equal to the convergence value D, it stops and the adaptive statistical iterative reconstruction is completed.

[0038] The image smoothing module 30 can filter out noise in each liver CT image.

[0039] As another optional implementation, the liver CT image enhancement processing system further includes a liver information database 40 and a retrieval module 50. The liver information database 40 is connected in communication with the data acquisition module 10 and the image enhancement module 20, and is used to store all liver CT images entering the system for enhancement processing and the corresponding liver three-dimensional models obtained after processing. The retrieval module 50 is connected in communication with the liver information database 40, and is used for the user to input retrieval information to retrieve the target CT image and the corresponding liver three-dimensional model in the liver information database.

[0040] Although the preferred embodiments of the present invention have been described in detail herein, it should be understood that the present invention is not limited to the specific structures described and shown in detail herein, and other variations and modifications may be implemented by those skilled in the art without departing from the spirit and scope of the present invention.

Claims

1. A liver CT image enhancement processing method, characterized in that: The following steps are involved: (1) Collect N liver CT images taken from different sections of a specific liver in the plain scan phase and the enhanced phase, where 50≤N≤200; (2) The m liver CT images in the plain scan period and the m liver CT images in the enhanced period are respectively enhanced by the filtered back projection method to obtain the plain scan period type A enhanced image and the enhanced period type A enhanced image, where m <N; (3) respectively enhancing the remaining k liver CT images in the plain scan period and the remaining k liver CT images in the enhanced period using an adaptive statistical iterative reconstruction method to obtain a plain scan period type B enhanced image and an enhanced period type B enhanced image, wherein k = Nm; and (4) integrating the plain scan type A enhanced images with the plain scan type B enhanced images, and integrating the enhanced type A enhanced images with the enhanced type B enhanced images, to jointly construct a three-dimensional liver model; Wherein, step (2) comprises: (2.1) Select any angle as the initial angle Initial angle Performing ray projection on pixels on the liver CT image to obtain a projection image, and performing a one-dimensional Fourier transform on the projection image to obtain a one-dimensional Fourier transform result; (2.2) performing a weighted operation on the one-dimensional Fourier transform result obtained in step (2.1) to obtain a weighted result; (2.3) performing a one-dimensional inverse Fourier transform on the weighted result obtained in step (2.2) to obtain a corrected projection image; (2.4) back-projecting the corrected projection image obtained in step (2.3) to obtain a back-projection image; (2.5) With the initial angle The starting point is transformed by 5 to 10 degrees, and the process of steps (2.1) to (2.4) is repeated until the 180-degree transformation back projection is completed, and the overlapping parts of all the back-projected images obtained are used as the reconstructed image.

2. The liver CT image enhancement processing method according to claim 1, characterized in that: m:k=1:1~5.

3. The liver CT image enhancement processing method according to claim 1, characterized in that: Step (3) includes: (3.1) A square is established on each liver CT image, wherein the side length of the square is the longest distance of the liver region on the corresponding liver CT image so as to cover the liver region, a coordinate system is established with the center of the square as the origin, the image function of the square in the coordinate system is set to a binary function, the function value of the binary function outside the square is assigned to 0, and the square is divided into N 2 There are N equal small squares, each of which is a pixel. 2 Assign values ​​from 1 to J to pixels, and set the grayscale value of pixel j to x j , where 1≤j≤J, J=N 2 ; (3.2) for x j Assign an initial value, set the correction index C and the convergence value D, and j Do ART iteration and get the i-th iteration result x j i ,i > 1; (3.3) Compare the i-th iteration result obtained in step (3.2) with iteration x j i With the initial value x j If the difference exceeds the convergence value D, update the initial value x j , where the updated x j For x j =x j i +C; (3.4) Substitute the updated x obtained in step (3.3) j Bring it into the next iteration and repeat the process of steps (3.2) and (3.3) until the jth component x of the kth iteration is j (k) When it is less than or equal to the convergence value D, it stops and the adaptive statistical iterative reconstruction is completed.

4. The liver CT image enhancement processing method according to any one of claims 1 to 3, characterized in that: Before step (2), the method further includes: using a spatial domain method to directly perform computational processing on the pixel grayscale values ​​of each liver CT image in the spatial domain to filter out noise in each liver CT image.

5. A liver CT image enhancement processing system, characterized in that: include: A data acquisition module, which is used to obtain N liver CT images taken from different sections in a plain scan phase and an enhanced phase for a specific liver, wherein 50≤N≤200; and An image enhancement module is communicatively connected with the data acquisition module, wherein the image enhancement module includes a filtering enhancement unit, an iterative reconstruction enhancement unit and an integration unit, The filtering enhancement unit is used to enhance the m liver CT images in the plain scan period and the m liver CT images in the enhanced period by using the filtering back projection method to obtain the A-type enhanced image in the plain scan period and the A-type enhanced image in the enhanced period, wherein the m <N; The iterative reconstruction and enhancement unit is used to process and enhance the remaining k liver CT images in the plain scan period and the k liver CT images in the enhanced period using an adaptive statistical iterative reconstruction method to obtain a plain scan period B-type enhanced image and an enhanced period B-type enhanced image, wherein k=Nm; The integration unit is respectively connected to the filtering enhancement unit and the iterative reconstruction enhancement unit for respectively integrating the type A enhanced image during the plain scan period with the type B enhanced image during the plain scan period, and after integrating the type A enhanced image during the enhanced period with the type B enhanced image during the enhanced period, constructing a three-dimensional liver model; The filtering enhancement unit includes: a projection subunit, a weighting subunit, a correction subunit, a back-projection subunit, and a reconstruction subunit which are sequentially connected to each other. The projection subunit is used to select the initial angle Performing ray projection on pixels on the liver CT image to obtain a projection image, and performing one-dimensional Fourier transform on the projection image to obtain a one-dimensional Fourier transform result; The weighting subunit is used to perform a weighted operation on the obtained one-dimensional Fourier transform result to obtain a weighted result; The correction subunit is used for performing a one-dimensional inverse Fourier transform on the obtained weighted result to obtain a corrected projection image; The back-projection subunit is used to back-project the obtained corrected projection image to obtain a back-projection image; The reconstruction subunit is used to reconstruct the The transformation is performed by 5 to 10 degrees each time as the starting point until 180 degrees of transformation back-projection is completed, and the overlapping parts of all the back-projected images obtained are used as the reconstructed image.

6. The liver CT image enhancement processing system according to claim 5, characterized in that: The iterative reconstruction enhancement unit comprises: a function setting unit, an assignment iteration unit, a comparison and update unit, and an iterative reconstruction unit which are sequentially connected in communication. The function setting unit is used to: establish a square on each liver CT image, the side length of the square is the longest distance of the liver area on the corresponding liver CT image to cover the liver area, establish a coordinate system with the center of the square as the origin, set the image function of the square as a binary function in the coordinate system, assign 0 to the function value of the binary function outside the square, and divide the square into N 2 There are N equal small squares, each of which is a pixel. 2 Assign values ​​from 1 to J to pixels, and set the grayscale value of pixel j to x j , where 1≤j≤J, J=N 2 ; The assignment iteration unit is used to j Assign an initial value, set the correction index C and the convergence value D, and j Do ART iteration and get the i-th iteration result x j i ,i > 1; The comparison and update unit is used to compare the i-th iteration result iteration x obtained by the assignment iteration unit. j i With the initial value x j If the difference exceeds the convergence value D, update the initial value x j , where the updated x j For x j =x j i +C; The iterative reconstruction unit is used to update the updated x obtained by the comparison and update unit j Bring it into the next iteration, so that the assignment iteration unit and the comparison update unit repeat the work until the jth component x of the kth iteration j (k) When it is less than or equal to the convergence value D, it stops and the adaptive statistical iterative reconstruction is completed.

7. The liver CT image enhancement processing system according to claim 5, characterized in that: It also includes an image smoothing module that is in communication with the data acquisition module and the image enhancement module and is used to filter out noise in each liver CT image.

8. The liver CT image enhancement processing system according to claim 5, characterized in that: Also includes: Liver information database and retrieval module, The liver information database is in communication with the data acquisition module and the image enhancement module, and is used to store a plurality of liver CT images and corresponding liver three-dimensional models; The retrieval module is in communication with the liver information database and is used for allowing a user to input retrieval information to retrieve a target CT image and a corresponding three-dimensional liver model in the liver information database.

Images