CN119295493B - Tumor medical image processing method and system of tumor ablation treatment system - Google Patents
Tumor medical image processing method and system of tumor ablation treatment system Download PDFInfo
- Publication number
- CN119295493B CN119295493B CN202411825006.2A CN202411825006A CN119295493B CN 119295493 B CN119295493 B CN 119295493B CN 202411825006 A CN202411825006 A CN 202411825006A CN 119295493 B CN119295493 B CN 119295493B
- Authority
- CN
- China
- Prior art keywords
- region
- edge
- abnormal
- area
- index
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 206010028980 Neoplasm Diseases 0.000 title claims abstract description 107
- 238000011298 ablation treatment Methods 0.000 title claims abstract description 30
- 238000003672 processing method Methods 0.000 title claims abstract description 20
- 230000002159 abnormal effect Effects 0.000 claims abstract description 379
- 238000004458 analytical method Methods 0.000 claims abstract description 116
- 230000008859 change Effects 0.000 claims abstract description 75
- 230000011218 segmentation Effects 0.000 claims abstract description 53
- 238000000034 method Methods 0.000 claims abstract description 33
- 238000012216 screening Methods 0.000 claims description 32
- 238000012545 processing Methods 0.000 claims description 21
- 238000003708 edge detection Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 claims 9
- 230000001186 cumulative effect Effects 0.000 claims 4
- 230000002596 correlated effect Effects 0.000 claims 3
- 230000005856 abnormality Effects 0.000 description 14
- 230000029058 respiratory gaseous exchange Effects 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000010606 normalization Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000010317 ablation therapy Methods 0.000 description 2
- 238000002591 computed tomography Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000011282 treatment Methods 0.000 description 2
- 208000032843 Hemorrhage Diseases 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 210000003484 anatomy Anatomy 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000012010 growth Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000017074 necrotic cell death Effects 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 238000003325 tomography Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 230000004614 tumor growth Effects 0.000 description 1
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
本发明涉及区域分割技术领域，具体涉及一种肿瘤消融治疗系统的肿瘤医学影像处理方法及系统。该方法对患病部位CT图像进行初次分割得到组织区域，筛选出初始异常区域，根据初始异常区域与其分析区域的形状变化程度的差异，以及初始异常区域与其分析区域所在患病部位CT图像之间图像数量的差异，筛选最终异常区域，然后筛选误分割区域；根据误分割区域与其分析区域的边缘像素点的位置分布差异调整误分割区域的边缘像素点的梯度值，之后利用调整后的梯度值重新对患病部位CT图像进行区域分割。本发明有效提高对患病部位CT图像进行区域分割的准确性。
The present invention relates to the field of regional segmentation technology, and specifically to a tumor medical image processing method and system for a tumor ablation treatment system. The method performs initial segmentation on the CT image of the diseased part to obtain the tissue region, screens out the initial abnormal region, and screens the final abnormal region based on the difference in the degree of shape change between the initial abnormal region and its analysis region, and the difference in the number of images between the CT image of the diseased part where the initial abnormal region and its analysis region are located, and then screens the mis-segmented region; the gradient value of the edge pixel point of the mis-segmented region is adjusted based on the difference in the position distribution of the edge pixel point of the mis-segmented region and its analysis region, and then the adjusted gradient value is used to re-segment the CT image of the diseased part. The present invention effectively improves the accuracy of regional segmentation of the CT image of the diseased part.
Description
Technical Field
The invention relates to the technical field of region segmentation, in particular to a tumor medical image processing method and system of a tumor ablation treatment system.
Background
Ablation therapy is a minimally invasive tumor therapy mode which is rapidly developed in recent years, and because different tissue parts have high structural complexity and functional diversity, boundaries of various tissues in a computed tomography (Computed Tomography, CT) image are often vague, and especially transition regions of tumors and normal tissues, the boundaries of the tumors and surrounding tissues are difficult to determine due to the complexity, so that when a tumor ablation therapy system is used for treating tumors in a patient, accurate segmentation of the tumors and the normal tissues is very important.
The existing method generally uses a region segmentation algorithm to segment normal tissues and tumor regions in a tumor CT image, but the segmentation effect on tumors and surrounding tissues in the CT image is poor due to the fact that boundaries of the tumor tissues are blurred and artifacts possibly exist in the CT image, so that the region segmentation of the CT image is inaccurate.
Disclosure of Invention
In order to solve the technical problem that the region segmentation algorithm is inaccurate for the region segmentation of CT images due to boundary blurring and artifacts of tumor tissues, the invention aims to provide a tumor medical image processing method and system of a tumor ablation treatment system, and the adopted technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a tumor medical image processing method of a tumor ablation treatment system, the method comprising:
Acquiring CT images of a plurality of layers of affected parts of a patient;
Obtaining analysis areas of each initial abnormal area, and screening final abnormal areas from the initial abnormal areas according to differences of the shape change degree of each initial abnormal area and the analysis areas thereof and differences of the image quantity between each initial abnormal area and the CT image of the diseased part where the analysis areas are located;
Screening the misclassification area from the final abnormal area according to the difference and the shape difference of the edge fluctuation degree of the final abnormal area and the analysis area thereof;
According to the position distribution difference of the edge pixel points of each misclassification area and the analysis area thereof, the gradient value of the edge pixel point of each misclassification area is adjusted, and the CT image of the affected part is subjected to area segmentation again by utilizing the adjusted gradient value.
Further, the screening the initial abnormal region in the CT image of the affected part from the tissue region includes:
For each layer of affected part CT image, acquiring an adjacent area of each tissue area in the affected part CT image, wherein edge pixel points exist on the edge of the adjacent area and are the edge pixel points on the edge of each tissue area;
Calculating the sum of the absolute values of the differences between each tissue region and the gray-scale value of all adjacent regions of the tissue region respectively, and recording the sum as a gray-scale specific index of each tissue region;
Acquiring the arithmetic average difference of gray values of pixel points in each tissue region;
Acquiring an abnormal index of each tissue region according to the gray specific index and the arithmetic average difference, wherein the gray specific index and the arithmetic average difference are in positive correlation with the abnormal index;
and taking the tissue region larger than the first preset threshold value as an initial abnormal region in the CT image of the affected part.
Further, the screening the final abnormal region from the initial abnormal region includes:
For each initial abnormal region, marking the total number of pixel points in the initial abnormal region as an area index of a corresponding region, and taking the total number of pixel points at the edge of the initial abnormal region as an edge index of the corresponding region;
And acquiring an abnormal change index of each initial abnormal region according to the difference of the image quantity and the difference of the image change index between the CT images of the affected part where each initial abnormal region and the analysis region are located, and taking the initial abnormal region larger than a second preset threshold value as a final abnormal region in the CT images of the affected part.
Further, the acquiring the abnormal change index of each initial abnormal region includes:
For each initial abnormal region, recording the number of images between the initial abnormal region and the CT image of the affected part where each analysis region is located as an interval image index of the initial abnormal region and each analysis region;
And carrying out normalization processing on products of the absolute values of the differences of the interval change indexes of the initial abnormal region and all the analysis regions and the absolute values of the differences of the interval image indexes to obtain the abnormal change index of each initial abnormal region.
Further, the screening the misclassified area from the final abnormal area includes:
Calculating the average value of the absolute value of the difference value of the chain code value of each edge pixel point on the edge of the final abnormal region and the adjacent edge pixel point, and marking the average value as the local fluctuation index of each edge pixel point on the edge of the final abnormal region;
Selecting an analysis area from the analysis areas of the final abnormal area as an adjacent interval area of the final abnormal area, and acquiring a shape edge difference index of the final abnormal area according to the difference of the edge fluctuation index of the final abnormal area and the adjacent interval area and the difference of the abnormal index;
And taking the average value of the shape edge difference indexes of all the final abnormal areas in each abnormal area sequence as a judging threshold value of the corresponding abnormal area, and taking the final abnormal area as a misclassification area if the shape edge difference index of the final abnormal area is larger than the judging threshold value of the abnormal area sequence where the final abnormal area is located.
Further, according to the difference of the position distribution of the edge pixel points of each misclassification area and the analysis area thereof, the gradient value of the edge pixel point of each misclassification area is adjusted, and the adjusted gradient value is utilized to conduct the region segmentation on the CT image of the affected part again, which comprises the following steps:
acquiring an abnormal edge possible index of each edge pixel point on the edge of each misclassification area according to the position distribution difference of the edge pixel points of each misclassification area and the analysis area;
Taking the sum of the abnormal edge possible index and a constant 1 as an adjustment coefficient of each edge pixel point on the edge of each misclassification area;
for each layer of diseased part CT image, carrying out edge detection on the diseased part CT image to obtain gradient values of edge pixels on the edge of the mistakenly segmented area in the diseased part CT image, and weighting the gradient values by utilizing the adjustment coefficient to obtain optimized gradient values of the edge pixels on the edge of the mistakenly segmented area;
And based on the optimized gradient value, re-dividing the region of the CT image of the affected part by using a watershed algorithm.
Further, the obtaining the abnormal edge possible index of each edge pixel point on the edge of each misclassification area includes:
for each final abnormal region, taking the distance between each edge pixel point on the edge of the final abnormal region and the centroid of the final abnormal region as the analysis distance of the corresponding edge pixel point;
Determining a starting point on the edge of the final abnormal region, and starting from the starting point, sequentially arranging edge pixel points on the edge of the final abnormal region to obtain an edge point sequence of the final abnormal region;
For each misclassification area, matching the misclassification area with the edge point sequence of each analysis area by using a DTW algorithm, and taking the average value of the analysis distances of all edge pixel points matched in the edge point sequence of the analysis area by each edge pixel point in the edge point sequence of the misclassification area as the comparison distance of each edge pixel point in the edge point sequence of the misclassification area;
And calculating the accumulated sum of the absolute difference values of the analysis distance and the comparison distance of each edge pixel point on the edge of the error segmentation area, and carrying out negative correlation and normalization processing on the accumulated sum to obtain an abnormal edge possible index of the corresponding edge pixel point.
Further, the acquiring the analysis area of each initial abnormal area includes:
And for each abnormal region sequence, taking the initial abnormal region adjacent to each initial abnormal region as an analysis region of each initial abnormal region in the abnormal region sequence.
Further, the method for carrying out region segmentation on the CT image of the affected part is a watershed algorithm.
In a second aspect, another embodiment of the present invention provides a tumor medical image processing system of a tumor ablation treatment system, the system comprising:
the data acquisition module is used for acquiring CT images of a plurality of layers of affected parts of a patient;
The abnormal region screening module is used for carrying out region segmentation on the CT image of the diseased part to obtain a tissue region, and screening an initial abnormal region in the CT image of the diseased part from the tissue region; acquiring analysis areas of each initial abnormal area, and screening a final abnormal area from the initial abnormal areas according to the difference of the shape change degree of each initial abnormal area and the analysis area thereof and the difference of the image quantity between CT images of the diseased part where each initial abnormal area and the analysis area are positioned;
the misclassification area screening module is used for screening misclassification areas from the final abnormal area according to the difference and the shape difference of the edge fluctuation degree of the final abnormal area and the analysis area;
The region segmentation optimization module is used for adjusting the gradient value of the edge pixel point of each error segmentation region according to the position distribution difference of the edge pixel point of each error segmentation region and the analysis region of each error segmentation region, and re-carrying out region segmentation on the CT image of the affected part by utilizing the adjusted gradient value.
The invention has the following beneficial effects:
In the embodiment of the invention, an initial abnormal region is selected from tissue regions obtained by primary segmentation of CT images of diseased regions, the initial abnormal region represents a tumor region, artifacts appear in the images due to respiration or heartbeat of a patient in a scanning process, region continuity of tissue structures in CT images of adjacent layers is damaged, the tissue structures in CT images of different layers are discontinuous, but the whole tissue structures of the continuous layers still show similar characteristics and regularly change along with respiration and heartbeat of the patient, differences of image quantity between the initial abnormal region and CT images of the diseased regions where the initial abnormal region and the analysis region are respectively located reflect regular change information, differences of shape change degrees of the initial abnormal region and the analysis region reflect structural similar characteristics, a final abnormal region is screened by combining the factors, the regions are regions with abnormal change of the tumor region due to artifacts formed by respiration or heartbeat, edges and shapes of the regions are continuous in the CT images of multiple layers under normal conditions, the regions with abnormal change are determined according to differences and shapes of edge fluctuation degrees of the final abnormal region and the analysis region, and the region segmentation with abnormal segmentation regions are caused to generate regions with deviation. The positions of the tumor areas of the continuous layers are relatively close, the positions of the mistaken segmentation areas and the positions of the analysis areas of the mistaken segmentation areas are relatively close under normal conditions, gradient information of the mistaken segmentation areas is adjusted according to the position segmentation differences of the edge pixel points of the mistaken segmentation areas and the positions of the analysis areas of the mistaken segmentation areas, and the CT images of the diseased parts are subjected to area segmentation again, so that the edges of the tumor areas and the edges of the normal tissue areas are clearer and more accurate.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart illustrating a method for processing tumor medical images of a tumor ablation treatment system according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating steps of a final tissue region screening method according to an embodiment of the present invention;
FIG. 3 is a system architecture diagram of a tumor medical image processing system of a tumor ablation treatment system according to an embodiment of the present invention;
Fig. 4 is a schematic diagram of a computer device of a tumor medical image processing device of a tumor ablation treatment system according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purposes, the following detailed description refers to the specific implementation, structure, characteristics and effects of a tumor medical image processing method and system of a tumor ablation treatment system according to the present invention, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The invention provides a tumor medical image processing method and a tumor medical image processing system for a tumor ablation treatment system, which are specifically described below with reference to the accompanying drawings.
Example 1:
the invention provides a tumor medical image processing method of a tumor ablation treatment system, referring to fig. 1, which shows a step flow chart of the tumor medical image processing method of the tumor ablation treatment system, the method comprises the following steps:
Step S1, acquiring CT images of a plurality of layers of affected parts of a patient.
Specifically, an X-ray tomography scanner is used for scanning a diseased part of a patient to obtain a multi-layer initial diseased part CT image, and in order to reduce the influence of noise on image analysis, denoising is carried out on the initial patient part CT image to obtain a patient part CT image.
It should be noted that, in the embodiment of the present invention, the denoising process is performed by using gaussian filtering, and a specific method is not described herein, which is a technical means well known to those skilled in the art, and the CT image is usually a gray image.
Step S2, carrying out region segmentation on the CT image of the affected part to obtain a tissue region, screening initial abnormal regions in the CT image of the affected part from the tissue region, acquiring analysis regions of each initial abnormal region, and screening final abnormal regions from the initial abnormal regions according to the difference of the shape change degree of each initial abnormal region and the analysis regions thereof and the difference of the image quantity between each initial abnormal region and the CT image of the affected part where the analysis regions are located.
The human tissue structure is complex, different parts have different textures and shapes, and in order to improve the accuracy of tumor feature analysis, the embodiment of the invention uses a watershed algorithm to segment the CT image of the affected part into different tissue areas. The watershed algorithm is a well-known technology for those skilled in the art, and is not described herein.
The affected part presents different expression characteristics in CT images of different layers, and as the characteristics of a tissue region can be influenced by a CT image scanning process, the tissue region obtained by preliminary segmentation can have the condition of unclear boundaries or blurry boundaries, and the unclear region division can lead a doctor to be incapable of accurately positioning the specific position and range of a tumor, thereby influencing the effects of diagnosis and subsequent treatment. Therefore, it is necessary to analyze the gray scale characteristics of the tissue region to obtain an abnormality index, thereby screening the initial abnormality region.
Preferably, in some possible implementation manners of the embodiment of the invention, the method for acquiring the abnormal index includes, for each layer of affected part CT image, acquiring adjacent areas of each tissue area in the affected part CT image, wherein edge pixel points exist on edges of the adjacent areas as edge pixel points on edges of each tissue area, taking an average value of gray values of all pixel points in each tissue area as a gray value of each tissue area, calculating an accumulated sum of difference absolute values of the gray value of each tissue area and the gray value of all adjacent areas respectively, and recording the accumulated sum as a gray specific index of each tissue area, acquiring an arithmetic average difference of gray values of the pixel points in each tissue area, and acquiring the abnormal index of each tissue area according to the gray specific index and the arithmetic average difference, wherein the gray specific index and the arithmetic average difference are in positive correlation.
The tumor area is the analysis focus because the boundaries of malignant tumors are often unclear, and it is necessary to identify the tumor area from the tissue area first for targeted analysis of the tumor. The interior of the tumor may contain different components, such as necrosis, hemorrhage or liquid accumulation, so that the gray level of the tumor area is uneven, and the gray level of the normal tissue area is more uniform, and the gray level characteristics of the tumor area are more obvious compared with the gray level characteristics of the normal tissue area around the tumor area, namely the gray level of the tumor area is obviously higher or lower than the gray level of the normal tissue area around the tumor area.
The adjacent areas of the tissue area are tissue areas around the tissue area, the gray concentration value reflects the whole gray level of the tissue area, the gray characteristic salience degree of the tissue area is obtained through the summation of the absolute values of the difference values of the gray concentration values of the tissue area and all the adjacent areas, and if the gray characteristic salience degree is larger, the abnormal degree of the tissue area is larger, and the possibility that the tissue area is tumor tissue is higher. The arithmetic mean difference of the gray values of the pixel points in the tissue region reflects the gray uniformity degree of the tissue region, and if the arithmetic mean difference is larger, the gray of the tissue region is not uniform, the abnormal degree of the tissue region is larger, and the possibility that the tissue region is tumor tissue is higher. Therefore, the arithmetic mean difference and the gray specific index are both in positive correlation with the anomaly index.
In a specific implementation manner of the embodiment of the present invention, the abnormality index of the tissue area is expressed as:
Wherein G is an abnormality index of each tissue region; t is the total number of adjacent areas of each tissue area; A gray set median value for a t-th contiguous region of each tissue region; u is the total number of pixel points in each tissue region; The gray value of the ith pixel point in each tissue area is obtained; An arithmetic average difference of gray values of pixel points in each tissue region; absolute value function, and Norm is normalization function.
The greater the abnormality index of a tissue region, the greater the likelihood that the tissue region is a tumor region. Therefore, a tissue region greater than a first preset threshold is taken as an initial abnormal region in the CT image of the diseased part, and the initial abnormal region represents tumor tissue in the CT image of the diseased part. It should be noted that, in the embodiment of the present invention, the first preset threshold takes an empirical value of 0.8, and an implementer can set the first preset threshold according to specific situations.
It is known that tumor structures are segmented into multiple layers of diseased portion CT images, each layer of diseased portion CT image may represent anatomical structures and pathological conditions of the tumor structure at a particular depth. The CT images of the affected part are obtained by scanning the body layer by layer, so that the tissue structures have region continuity in the CT images of the adjacent layers, but breathing, heartbeat or other involuntary actions of a patient in the scanning process can cause artifacts to appear in the CT images, so that the change degree of partial regions is larger, the phenomenon that the tissue structures are discontinuous in the CT images of different layers is caused, and the tissue structures of the adjacent layers still show similar characteristics as a whole. Thus, an analysis region of the initial abnormality region is acquired, and the initial abnormality region and its analysis region can be regarded as a tumor structure of the continuous layer.
In the embodiment, the initial abnormal regions corresponding to each other in the CT images of all the affected parts are sequentially arranged to obtain a plurality of abnormal region sequences, and for each abnormal region sequence, the initial abnormal region adjacent to each initial abnormal region is used as an analysis region of each initial abnormal region.
It should be noted that, each layer of CT images of the affected part has a corresponding scanning time, and the three-dimensional model of the affected part can be presented by arranging all CT images of the affected part according to the scanning time sequence. Arranging initial abnormal areas corresponding to each other in CT images of all affected parts according to the scanning time sequence of CT images of the affected parts where the initial abnormal areas are located to obtain a plurality of abnormal area sequences, wherein the initial abnormal areas in each abnormal area sequence can form a tumor structure. In the abnormal region sequence, the adjacent previous and next initial abnormal regions of each initial abnormal region are taken as analysis regions of each initial abnormal region for the rest initial abnormal regions except the first and last initial abnormal regions, and the initial abnormal region has two analysis regions. It should be noted that, in this embodiment, the first and last initial abnormal regions in the abnormal region sequence are directly used as the final abnormal regions.
In other embodiments, the centroid of an initial abnormal region in the CT image of the affected part is acquired, a layer of CT image of the affected part is selected as a target CT image, one initial abnormal region is selected from the target CT image as a target region, the pixel points of the centroid of the target region at the same position in the CT images of the other affected parts except the target CT image are used as the reference points of the target region, for each CT image of the other affected parts except the target CT image, whether the centroid of the initial abnormal region exists in the preset range of the reference points of the target region is judged, if yes, the initial abnormal region corresponding to the centroid closest to the reference point of the target region is used as the initial abnormal region corresponding to the CT image of the affected part in the CT image of the affected part, if no, the initial abnormal region corresponding to the target region does not exist in the CT image of the affected part, and the initial abnormal region corresponding to the target region in the CT images of the other affected parts except the target CT image is arranged according to the scanning time sequence of the CT images of the affected part to obtain an abnormal region sequence. According to the method, all abnormal region sequences in CT images of all diseased parts are acquired.
It should be noted that, because the multi-layer CT images of the affected part are obtained by scanning the body layer by layer, the image change between the CT images of adjacent layers is usually small, the positions of the corresponding areas of the same tumor structure in the CT images of different affected parts are basically the same, and if the centroids of the areas are mapped to the same plane, the centroids are more concentrated. The preset range of the reference point in this embodiment is an area formed by taking the reference point as the center and 50 as the radius, and the radius can be set according to the actual situation.
Although the respiration and the heartbeat of the patient can cause the discontinuous CT images of the tissue structures of the continuous layers, the tissue structures of the continuous layers still show similar characteristics and change regularly along with the respiration and the heartbeat of the patient. The difference of the image quantity between the initial abnormal region and CT images of the affected part where the two analysis regions are located reflects regular change information, the difference of the shape change degree of the initial abnormal region and the shape change degree of the analysis regions reflects structural similarity characteristics, the factors are synthesized to analyze, and a final abnormal region is selected from the initial abnormal region, wherein the final abnormal region is a region with abnormal change of a tumor region due to breathing or heartbeat forming artifacts.
Referring to fig. 2, a flowchart illustrating a final abnormal region screening method according to an embodiment of the present invention is shown, where the method includes:
And S210, for each initial abnormal region, recording the total number of pixel points in the initial abnormal region as an area index of the corresponding region, and taking the total number of pixel points at the edge of the initial abnormal region as an edge index of the corresponding region, and acquiring an image change index of each initial abnormal region according to the difference between the area index of each initial abnormal region and the analysis region, the difference between the edge index and the abnormal index of each initial abnormal region.
Along with the scanning direction, the CT image can change gradually, the change degree of the multi-layer tissue area divided by the tissue structure is basically consistent, and the image change of the same tissue structure in CT images of different layers of diseased parts is analyzed through the area and the edge length of the area. The area index reflects the area information of the initial abnormal region, the edge index reflects the edge length information of the initial abnormal region, and if the difference between the area index of the initial abnormal region and the difference between the area index of the analysis region and the edge index of the initial abnormal region are larger, the larger the change degree of the initial abnormal region and the analysis region is, the larger the image change index is.
Since the growth and development of a tumor may cause rapid changes in its morphology, boundary and internal structure, if the initial abnormal region has a larger abnormality index, the initial abnormal region has a larger possibility of representing an actual tumor tissue, and the initial abnormal region has a larger possibility of occurrence of a change, the image change index is larger.
Therefore, the difference between the area index of the initial abnormal region and the area index of the analysis region, the difference between the edge index and the abnormal index of the initial abnormal region are in positive correlation with the image change index. In the embodiment of the invention, the product of the absolute value of the difference value of the area index of each initial abnormal region and the analysis region thereof, the absolute value of the difference value of the edge index and the abnormal index of each initial abnormal region is used as the image change index of each initial abnormal region.
In the embodiment of the present invention, the correlation between the absolute value of the difference between the area indexes of the initial abnormal region and the analysis region thereof, the absolute value of the difference between the edge indexes, and each of the initial abnormal indexes and the image change indexes may also be constructed by other basic mathematical operations, which are not limited and described in detail herein.
Step S220, obtaining an abnormal change index of each initial abnormal region according to the difference of the image quantity and the difference of the image change index between the CT images of the affected part where each initial abnormal region and the analysis region are located, and taking the initial abnormal region larger than a second preset threshold value as a final abnormal region in the CT images of the affected part.
Preferably, in some possible implementation manners of the embodiment of the invention, the method for acquiring the abnormal change index comprises the steps of recording the number of images between the initial abnormal region and the CT image of the affected part where each analysis region is located as an interval image index of the initial abnormal region and each analysis region for each initial abnormal region, calculating the absolute value of the difference value of the image change index of the initial abnormal region and each analysis region as an interval change index of the initial abnormal region and each analysis region, and carrying out normalization processing on the product of the absolute value of the difference value of the interval change index of the initial abnormal region and all the analysis regions and the absolute value of the difference value of the interval image index to obtain the abnormal change index of each initial abnormal region.
The CT images can change gradually along the scanning direction, the change degree of the multi-layer tissue areas with the tissue structures divided is basically consistent, the change degree is usually smaller, and the influence of motion artifacts on analysis is considered, because respiration and heartbeat have regularity, along with the scanning, the CT images of the tissue structures of the continuous layers show regular change, namely the number of the images among the CT images of the tissue structures of the continuous layers is relatively close. Therefore, if the number of images between CT images in which the tissue structures of the continuous layers are located is relatively close and the degree of change of the initial abnormal region corresponding to the tissue structures of the continuous layers is relatively close, the probability that the abnormal region in which the tumor region changes due to the breathing or heartbeat forming artifact appears in the initial abnormal region corresponding to the tissue structures of the continuous layers is smaller. When the absolute value of the difference value of the interval change index and the absolute value of the interval image index of the initial abnormal region are both larger, the probability of occurrence of abnormal change of the initial abnormal region is larger.
In a specific implementation manner of the embodiment of the present invention, the abnormal change index of the initial abnormal region is expressed as:
Wherein, B is the abnormal change index of each initial abnormal region, P is the image change index of each initial abnormal region; An image change index for the 1 st analysis area of each initial abnormal area; An image change index for the 2 nd analysis area of each initial abnormal area; a change index of the interval between each initial abnormal region and the 1 st analysis region; A change index of the interval between each initial abnormal region and the 2 nd analysis region; an interval image index of each initial abnormal region and the 1 st analysis region thereof; An interval image index of each initial abnormal region and the 2 nd analysis region; absolute value function, and Norm is normalization function.
If the abnormal change index of the initial abnormal region is larger, the change regularity of the initial abnormal region and the analysis region is poorer, and the possibility of abnormal change of the initial abnormal region is larger, the initial abnormal region which is larger than a second preset threshold value is taken as a final abnormal region in the CT image of the affected part. It should be noted that, in this embodiment, the second preset threshold takes an empirical value of 0.9, and the practitioner can set the second preset threshold according to the specific situation.
And S3, screening the misclassification area from the final abnormal area according to the difference and the shape difference of the edge fluctuation degree of the final abnormal area and the analysis area thereof.
Under normal conditions, the edges and the shape of the region are continuous in different layers of CT images, and if the region variation of the region image of a certain layer in the tissue structure of the continuous layer is obviously larger than that of the region images of other layers, the region division of the region image of the layer may have certain deviation due to various factors. The final abnormal region should be relatively close to the abnormal change degree in the adjacent images, and the abnormal change degree of the region is reflected by the edge fluctuation degree and the shape of the region, so that the mistakenly segmented region is identified.
Preferably, in some possible implementation manners of the embodiment of the invention, the method for acquiring the shape edge difference index comprises the steps of acquiring a chain code value of each edge pixel point on the edge of a final abnormal region for each final abnormal region, calculating a mean value of absolute values of differences between each edge pixel point on the edge of the final abnormal region and the chain code values of adjacent edge pixel points, respectively, and marking the mean value as a local fluctuation index of each edge pixel point on the edge of the final abnormal region, marking the mean value as the edge fluctuation index of all edge pixel points on the edge of the final abnormal region, and marking an analysis region from an analysis region of the final abnormal region as an adjacent interval region of the final abnormal region, and acquiring the shape edge difference index of the final abnormal region according to the difference between the edge fluctuation index of the final abnormal region and the adjacent interval region and the difference between the abnormality indexes. The method for obtaining the chain code value is a known technique and will not be described herein.
The chain code value of the edge pixel point of the final abnormal region reflects the change trend of the edge of the final abnormal region, and the chain code value of each edge pixel point on the edge of the final abnormal region represents the change direction from the current edge pixel point to the next edge pixel point in the clockwise direction. The local fluctuation index of each edge pixel point on the edge of the final abnormal region presents the fluctuation degree of the local edge where the edge pixel point is located, and the average value of the local fluctuation indexes of all the edge pixel points on the edge of the final abnormal region reflects the integral fluctuation characteristic of the edge of the final abnormal region. The abnormality index reflects the area characteristics and edge length characteristics of the final abnormal region, which together determine the shape characteristics of the final abnormal region.
In the embodiment of the invention, in the abnormal region sequence, the adjacent next initial abnormal region of each final abnormal region is used as the adjacent interval region of the final abnormal region. If the difference of the edge fluctuation index and the difference of the abnormality index of the final abnormality region and the adjacent interval region are larger, the more obvious the edge fluctuation and the shape characteristic change of the final abnormality region in the adjacent layer are, the larger the shape edge difference index is. Therefore, the difference in the edge fluctuation index and the difference in the abnormality index of the final abnormal region and the adjacent interval region thereof and the shape edge difference index are in positive correlation.
In the embodiment of the invention, the product of the absolute value of the difference value of the edge fluctuation index of the final abnormal region and the adjacent interval region and the absolute value of the difference value of the abnormal index is used as the shape edge difference index of the final abnormal region. The larger the shape edge index, the greater the likelihood that the final abnormal region is divided by mistake.
And taking the average value of the shape edge difference indexes of all the final abnormal areas in each abnormal area sequence as a judging threshold value of the corresponding abnormal area, and taking the final abnormal area as a misclassification area if the shape edge difference index of the final abnormal area is larger than the judging threshold value of the abnormal area sequence where the final abnormal area is positioned.
And S4, adjusting the gradient value of the edge pixel point of each misclassification area according to the position distribution difference of the edge pixel point of each misclassification area and the edge pixel point of each misclassification area, and carrying out area segmentation on the CT image of the affected part again by utilizing the adjusted gradient value.
The position of the tumor region of the continuous layer is relatively close, the position segmentation difference of the edge pixel points of the mistaken segmentation region and the analysis region thereof is used for analyzing to obtain an abnormal edge possible index for measuring the possibility that the edge pixel points belong to the actual tumor edge, and the gradient value of the edge pixel points of the mistaken segmentation region is adjusted by using the abnormal edge possible index, so that the accuracy of region segmentation is improved.
Preferably, in some possible implementation manners of the embodiment of the present invention, the method for acquiring the abnormal edge possible indicator includes:
The method comprises the steps of determining the distance between each edge pixel point on the edge of a final abnormal region and the centroid of each edge pixel point, determining the starting point on the edge of the final abnormal region, sequentially arranging the edge pixel points on the edge of the final abnormal region from the starting point to obtain an edge point sequence of the final abnormal region, matching the edge point sequence of each analysis region with the error segmentation region by using a DTW algorithm for each error segmentation region, taking the average value of the analysis distances of all edge pixel points matched in the edge point sequence of each edge pixel point of the error segmentation region as the comparison distance of each edge pixel point in the edge point sequence of the error segmentation region, calculating the sum of the difference absolute values of the analysis distances of each edge pixel point on the edge of the error segmentation region and the comparison distances of the edge pixel points, carrying out negative correlation and normalization processing on the sum to obtain the possible index of the abnormal edge of the corresponding edge pixel points. The dynamic time warping algorithm (DYNAMIC TIME WARPING, DTW) is a well-known technique for those skilled in the art, and will not be described herein.
The method for determining the starting point on the edge of the final abnormal region comprises the steps of selecting the largest analysis distance corresponding to the edge-dividing pixel point as an initial point for the analysis distance of all edge pixel points on the edge of the final abnormal region, recording a line segment connecting each initial point with the mass center of the final abnormal region as a screening line segment of each initial point, acquiring the included angles between the screening line segments of all the initial points and the horizontal right direction, and taking the initial point corresponding to the screening line segment corresponding to the smallest included angle as the starting point on the edge of the final abnormal region.
In this embodiment, from the start point on the edge of the final abnormal region, edge pixel points on the edge of the final abnormal region are sequentially arranged in the clockwise direction, to obtain the edge point sequence of the final abnormal region. The difference of the analysis distance between each edge pixel point in the edge point sequence of the misclassification area and the edge pixel point matched with the edge pixel point in the edge point sequence of the analysis area presents the position difference between each edge pixel point of the misclassification area and the edge pixel point matched with the edge pixel point, and if the difference is larger, the greater the deviation degree of the misclassification area compared with the analysis area is indicated, the smaller the possibility that the edge pixel point belongs to the actual tumor edge of the area edge is. Therefore, the accumulated sum of the absolute value of the difference between the analysis distance of each edge pixel point on the edge of the misclassified area and the comparison distance thereof is in negative correlation with the possible index of the abnormal edge.
It should be noted that, each edge pixel point has two comparison distances, if a certain edge pixel point of the misclassified area has no matched edge pixel point in one or two analysis areas, that is, the edge pixel point has no or only one comparison distance, the possible index of the abnormal edge of the edge pixel point in the misclassified area is set as a constant 0.
According to the scheme, the watershed algorithm is adopted to carry out region segmentation on the CT image of the affected part, and the gray level image is generally converted into the gradient image for analysis, so that the gradient value of the edge pixel point of the mistakenly segmented region needs to be adjusted by using the possible indexes of the abnormal edge, and the accuracy of region segmentation is improved. If the abnormal edge of the edge pixel point is larger, the possibility that the edge pixel point belongs to the edge of the tumor tissue is larger, the edge of the tumor tissue is blurred, the gradient value is larger, the edge of the mistakenly segmented area is clearer, and the accuracy of area segmentation can be improved by adjusting the gradient value of the edge pixel point.
The method for adjusting the gradient value comprises the steps of taking the sum value of an abnormal edge possible index and a constant 1 as an adjustment coefficient of each edge pixel point on the edge of each misclassified area, and carrying out edge detection on a CT image of a diseased part for each layer of CT image of the diseased part to obtain the gradient value of the edge pixel point on the edge of the misclassified area in the CT image of the diseased part, and weighting the gradient value by utilizing the adjustment coefficient to obtain the optimized gradient value of the edge pixel point on the edge of the misclassified area.
For each diseased part CT image, the optimized gradient value of the edge pixel point of the mistakenly segmented area in the diseased part CT image is kept unchanged, the optimized gradient image of the diseased part CT image is obtained, and the optimized gradient image is subjected to area segmentation by using a watershed algorithm, so that the more accurate area segmentation of the diseased part CT image is realized, and the edges of the tumor area and the normal tissue area are clearer and more accurate.
It should be noted that, in this embodiment, a Canny operator is selected to perform edge detection on the CT image of the affected part, where the Canny operator is a well-known technology of those skilled in the art, and is not described herein again.
Through the states of the same tissue structure in different CT images, a doctor can evaluate the tumor characteristics and the tumor growth trend more intuitively, assist the doctor in judging the range of the position of the tumor structure, and can formulate a more reasonable tumor ablation treatment scheme and optimize the treatment plan.
The present invention has been completed.
Example 2:
The invention provides a tumor medical image processing system of a tumor ablation treatment system, please refer to fig. 3, which shows a system structure diagram of the tumor medical image processing system of the tumor ablation treatment system provided by an embodiment of the invention, the system comprises:
The data acquisition module 510 is used for acquiring CT images of a plurality of layers of affected parts of a patient;
The abnormal region screening module 520 is configured to perform region segmentation on the CT image of the diseased region to obtain a tissue region, and screen an initial abnormal region in the CT image of the diseased region from the tissue region; acquiring analysis areas of each initial abnormal area, and screening a final abnormal area from the initial abnormal areas according to the difference of the shape change degree of each initial abnormal area and the analysis area thereof and the difference of the image quantity between CT images of the diseased part where each initial abnormal area and the analysis area are positioned;
The misclassification area screening module 530 is configured to screen misclassification areas from the final abnormal area according to the difference and the shape difference of the edge fluctuation degree of the final abnormal area and the analysis area thereof;
the region segmentation optimization module 540 is configured to adjust a gradient value of an edge pixel point of each misclassification region according to a position distribution difference of the edge pixel point of each misclassification region and an edge pixel point of an analysis region thereof, and re-segment the CT image of the affected part by using the adjusted gradient value.
It should be noted that, in the apparatus provided in the foregoing embodiment, only the division of the foregoing functional modules is illustrated, and in practical application, the foregoing functional allocation may be performed by different functional modules according to needs, that is, the internal structure of the computer apparatus is divided into different functional modules, so as to perform all or part of the functions described above. In addition, the embodiment of the tumor medical image processing system of the tumor ablation treatment system and the embodiment of the tumor medical image processing method of the tumor ablation treatment system provided in the foregoing embodiments belong to the same concept, and detailed implementation processes of the embodiment of the method are detailed in the embodiment of the method, and are not described herein again.
Example 3:
Fig. 4 is a schematic diagram of a computer device of a tumor medical image processing device of a tumor ablation treatment system according to an embodiment of the present invention. Illustratively, as shown in FIG. 4, the computer device includes a memory 601, a processor 602, and a computer program 603 stored in the memory 601 and running on the processor 602, wherein the processor 602, when executing the computer program 603, causes the computer device to perform any of the tumor medical image processing methods of the tumor ablation treatment system described above.
In addition, the embodiment of the application also protects a device, which can comprise a memory and a processor, wherein executable program codes are stored in the memory, and the processor is used for calling and executing the executable program codes to execute the tumor medical image processing method of the tumor ablation treatment system provided by the embodiment of the application.
In this embodiment, the functional modules of the apparatus may be divided according to the above method example, for example, each functional module may be corresponding to one processing module, or two or more functions may be integrated into one processing module, where the integrated modules may be implemented in a hardware form. It should be noted that, in this embodiment, the division of the modules is schematic, only one logic function is divided, and another division manner may be implemented in actual implementation.
It should be understood that the apparatus provided in this embodiment is used to perform the tumor medical image processing method of a tumor ablation treatment system, so that the same effects as those of the implementation method can be achieved.
In case of an integrated unit, the apparatus may comprise a processing module, a memory module. When the device is applied to equipment, the processing module can be used for controlling and managing the actions of the equipment. The memory module may be used to support devices executing inter-program code, etc.
Wherein the processing module may be a processor or controller that may implement or execute the various exemplary logic blocks, modules and circuits contained in connection with the present disclosure. A processor may also be a combination of computing functions, including for example one or more microprocessors, digital Signal Processing (DSP) and microprocessor combinations, etc., and a memory module may be a memory.
It should be noted that the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
The foregoing description of the preferred embodiments of the present invention is not intended to be limiting, but rather, any modifications, equivalents, improvements, etc. that fall within the principles of the present invention are intended to be included within the scope of the present invention.
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411825006.2A CN119295493B (en) | 2024-12-12 | 2024-12-12 | Tumor medical image processing method and system of tumor ablation treatment system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411825006.2A CN119295493B (en) | 2024-12-12 | 2024-12-12 | Tumor medical image processing method and system of tumor ablation treatment system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119295493A CN119295493A (en) | 2025-01-10 |
| CN119295493B true CN119295493B (en) | 2025-02-25 |
Family
ID=94155199
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411825006.2A Active CN119295493B (en) | 2024-12-12 | 2024-12-12 | Tumor medical image processing method and system of tumor ablation treatment system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119295493B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109493343A (en) * | 2018-12-29 | 2019-03-19 | 上海鹰瞳医疗科技有限公司 | Medical image abnormal area dividing method and equipment |
| CN116309649A (en) * | 2023-05-17 | 2023-06-23 | 青岛市妇女儿童医院（青岛市妇幼保健院、青岛市残疾儿童医疗康复中心、青岛市新生儿疾病筛查中心） | Throat lesion abnormal region segmentation system based on CT image |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8577104B2 (en) * | 2010-06-25 | 2013-11-05 | Carestream Health, Inc. | Liver lesion segmentation |
| TWI496112B (en) * | 2013-09-13 | 2015-08-11 | Univ Nat Cheng Kung | Cell image segmentation method and a nuclear-to-cytoplasmic ratio evaluation method using the same |
| US10492723B2 (en) * | 2017-02-27 | 2019-12-03 | Case Western Reserve University | Predicting immunotherapy response in non-small cell lung cancer patients with quantitative vessel tortuosity |
| EP4038572A1 (en) * | 2019-10-02 | 2022-08-10 | Memorial Sloan Kettering Cancer Center | Deep multi-magnification networks for multi-class image segmentation |
| CN118521793B (en) * | 2024-07-22 | 2024-11-01 | 高州市人民医院 | Feature extraction-based tumor segmentation method, system and storage medium |
-
2024
- 2024-12-12 CN CN202411825006.2A patent/CN119295493B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109493343A (en) * | 2018-12-29 | 2019-03-19 | 上海鹰瞳医疗科技有限公司 | Medical image abnormal area dividing method and equipment |
| CN116309649A (en) * | 2023-05-17 | 2023-06-23 | 青岛市妇女儿童医院（青岛市妇幼保健院、青岛市残疾儿童医疗康复中心、青岛市新生儿疾病筛查中心） | Throat lesion abnormal region segmentation system based on CT image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119295493A (en) | 2025-01-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kandhway et al. | A novel reformed histogram equalization based medical image contrast enhancement using krill herd optimization | |
| EP0757544B1 (en) | Computerized detection of masses and parenchymal distortions | |
| CN110163809A (en) | Confrontation network DSA imaging method and device are generated based on U-net | |
| TW201903708A (en) | Method and system for analyzing digital subtraction angiography images | |
| CN113066061A (en) | Aneurysm detection method, system, terminal and medium based on MRA | |
| CN116993628B (en) | A CT image enhancement system for tumor radiofrequency ablation guidance | |
| CN111340825A (en) | Method and system for generating mediastinal lymph node segmentation model | |
| CN118229538B (en) | Intelligent enhancement method for bone quality CT image | |
| CN117745722B (en) | Medical health physical examination big data optimization enhancement method | |
| CN110288698B (en) | Meniscus three-dimensional reconstruction system based on MRI | |
| CN117495818A (en) | Orthopedics auxiliary examination method and system based on image processing | |
| Kesuma et al. | Improved Chest X-Ray Image Quality Using Median and Gaussian Filter Methods | |
| CN109816665B (en) | A method and device for fast segmentation of optical coherence tomography images | |
| Mirajkar et al. | Acute ischemic stroke detection using wavelet based fusion of CT and MRI images | |
| CN118379313B (en) | A medical image segmentation method for intracranial arteriosclerotic plaques in patients with atrial fibrillation | |
| CN119295493B (en) | Tumor medical image processing method and system of tumor ablation treatment system | |
| CN116309647B (en) | Method for constructing craniocerebral lesion image segmentation model, image segmentation method and device | |
| EP2178046A2 (en) | Multi-image correlation | |
| Aprea et al. | Neural machine registration for motion correction in breast DCE-MRI | |
| CN118781140B (en) | Artificial intelligence-based heart ultrasonic image segmentation method and system | |
| CN118644509B (en) | A CT image segmentation method for surgical operations | |
| Kaftan et al. | Locally adaptive fuzzy pulmonary vessel segmentation in contrast enhanced CT data | |
| Koompairojn et al. | Semi-automatic segmentation and volume determination of brain mass-like lesion | |
| CN118799328B (en) | Contrast image analysis method and device | |
| US20230136320A1 (en) | System and method for control of motion in medical images using aggregation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |