CN117291281A - Method for training evaluation model for evaluating brain collateral circulation and related product - Google Patents
Method for training evaluation model for evaluating brain collateral circulation and related product Download PDFInfo
- Publication number
- CN117291281A CN117291281A CN202311102535.5A CN202311102535A CN117291281A CN 117291281 A CN117291281 A CN 117291281A CN 202311102535 A CN202311102535 A CN 202311102535A CN 117291281 A CN117291281 A CN 117291281A
- Authority
- CN
- China
- Prior art keywords
- brain
- model
- collateral circulation
- brain tissue
- regions
- 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.)
- Pending
Links
- 210000004556 brain Anatomy 0.000 title claims abstract description 153
- 206010069729 Collateral circulation Diseases 0.000 title claims abstract description 62
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000012549 training Methods 0.000 title claims abstract description 32
- 238000013210 evaluation model Methods 0.000 title claims abstract description 28
- 210000005013 brain tissue Anatomy 0.000 claims abstract description 81
- 238000010968 computed tomography angiography Methods 0.000 claims abstract description 32
- 238000011156 evaluation Methods 0.000 claims abstract description 26
- 230000011218 segmentation Effects 0.000 claims abstract description 18
- 210000004204 blood vessel Anatomy 0.000 claims description 23
- 230000002490 cerebral effect Effects 0.000 claims description 10
- 230000009466 transformation Effects 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000003066 decision tree Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 3
- 238000007637 random forest analysis Methods 0.000 claims description 3
- 230000001131 transforming effect Effects 0.000 claims description 3
- 230000004087 circulation Effects 0.000 description 19
- 238000010586 diagram Methods 0.000 description 10
- 230000036770 blood supply Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 6
- 210000003657 middle cerebral artery Anatomy 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 210000001519 tissue Anatomy 0.000 description 5
- 230000000302 ischemic effect Effects 0.000 description 4
- 230000002093 peripheral effect Effects 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 230000017531 blood circulation Effects 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000010412 perfusion Effects 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 208000006011 Stroke Diseases 0.000 description 2
- 230000003872 anastomosis Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 230000002792 vascular Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 201000006474 Brain Ischemia Diseases 0.000 description 1
- 208000004434 Calcinosis Diseases 0.000 description 1
- 206010008120 Cerebral ischaemia Diseases 0.000 description 1
- 206010061216 Infarction Diseases 0.000 description 1
- 208000031481 Pathologic Constriction Diseases 0.000 description 1
- 206010053648 Vascular occlusion Diseases 0.000 description 1
- 210000001367 artery Anatomy 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000000988 bone and bone Anatomy 0.000 description 1
- 230000002308 calcification Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 206010008118 cerebral infarction Diseases 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000002872 contrast media Substances 0.000 description 1
- 230000010339 dilation Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000007574 infarction Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000003446 pia mater Anatomy 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000036262 stenosis Effects 0.000 description 1
- 208000037804 stenosis Diseases 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 208000021331 vascular occlusion disease Diseases 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30016—Brain
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The application discloses a method for training an evaluation model for evaluating brain collateral circulation and a related product. The method comprises the following steps: acquiring a CT angiography image aiming at the brain and marking the brain collateral circulation score of the CT angiography image; segmenting the CT angiography image using the segmentation model to obtain a brain tissue region; extracting left and right brain regions and a target layer based on the brain tissue region; calculating a maximum density projection image related to evaluation of brain collateral circulation according to the brain left and right areas and the target layer; and inputting the maximum density projection image and the brain collateral circulation scoring annotation into the main model for training so as to train an evaluation model for evaluating brain collateral circulation. By utilizing the scheme of the application, the evaluation efficiency and the evaluation accuracy can be greatly improved.
Description
Technical Field
The present application relates generally to the field of image processing technology. More specifically, the present application relates to a method for training an assessment model for assessing brain side branch circulation and a method for assessing brain side branch circulation. Further, the present application also relates to a device and a computer readable storage medium for assessing brain side branch circulation.
Background
Side branch refers to a branched vascular structure that connects adjacent large vessels, which are present in most tissues, while the main function of side branch vessels is to alter the blood flow path, providing perfusion of blood flow to the blood supply area of the occluded vessel. The cerebral collateral circulation refers to when the blood supply artery of the brain is severely stenosed or occluded, blood flow reaches the ischemic area through other blood vessels (collateral or newly formed vascular anastomosis), so that the ischemic tissue is compensated for by perfusion to different extents. For example, collateral circulation can be created by anastomosis between an artery-artery or vein-vein, thereby compensating for different degrees of perfusion of ischemic tissue. The collateral circulation determines ischemic penumbra, infarct size, time course and severity of cerebral ischemia, and whether or not stroke occurs after vascular occlusion, and is a major cause of stroke heterogeneity, so understanding and evaluating collateral circulation helps to establish clinical decisions and judge prognosis.
Currently, the collateral circulation is generally evaluated by acquiring a CT Angiography (CT Angiography)' CTA ") image of a subject, and manually calculating the vessel area ratio of the left and right sides of the CT image in the CTA image, thereby determining a Tan score according to the ratio of the vessel areas. However, the side branch circulation is evaluated based on a manual calculation mode, so that the evaluation efficiency is low, and the evaluation accuracy is low.
In view of this, it is highly desirable to provide a solution for training an assessment model for assessing brain collateral circulation in order to assess brain collateral circulation in an artificial intelligence based manner, greatly improving the assessment efficiency and the accuracy of the assessment.
Disclosure of Invention
In order to solve at least one or more of the technical problems mentioned above, the present application proposes, in various aspects, a solution for training an assessment model for assessing a brain side branch circulation.
In a first aspect, the present application provides a method for training an assessment model for assessing brain side branch circulation, wherein the assessment model comprises a segmentation model and a main model, and the method comprises: acquiring a CT angiography image aiming at the brain and marking the brain collateral circulation score of the CT angiography image; segmenting the CT angiography image using the segmentation model to obtain a brain tissue region; extracting left and right brain regions and a target layer based on the brain tissue region; calculating a maximum density projection image related to evaluation of brain collateral circulation according to the brain left and right areas and the target layer; and inputting the maximum density projection image and the brain collateral circulation scoring annotation into the main model for training so as to train an evaluation model for evaluating brain collateral circulation.
In one embodiment, wherein extracting the left and right brain regions and the target layer based on the brain tissue region comprises: registering the brain tissue region with the brain tissue map to obtain a registered brain tissue region; and extracting the left and right brain regions and the target layer according to the registered brain tissue regions.
In another embodiment, wherein registering the brain tissue region with the brain tissue map to obtain a registered brain tissue region comprises: calculating a transformation matrix from the brain tissue region to the brain tissue map; and transforming the brain tissue region according to the transformation matrix to obtain the registered brain tissue region.
In yet another embodiment, wherein extracting the left and right brain regions and the target layer from the registered brain tissue regions comprises: and performing position matching on the registered brain tissue regions based on the position information in the brain tissue map so as to extract the left and right brain regions and the target layer.
In yet another embodiment, the method further comprises: gray scale layering is carried out on the brain blood vessels in the maximum density projection image so as to obtain gray scale characteristics of the brain blood vessels; and inputting the gray scale characteristics and the brain collateral circulation scoring labels into the main model for training so as to train an evaluation model for evaluating brain collateral circulation.
In yet another embodiment, wherein gray scale layering of the brain blood vessel in the maximum intensity projection image to obtain gray scale characteristics of the brain blood vessel comprises: setting different gray threshold values according to the contrast ratio of the left and right brain regions in the maximum density projection image; and carrying out gray scale layering on the cerebral blood vessels based on the different gray scale thresholds so as to obtain gray scale characteristics of the cerebral blood vessels of different levels of the left and right regions of the brain in the target layer.
In yet another embodiment, wherein the master model comprises a decision tree model, a random forest model, or a neural network model.
In a second aspect, the present application provides a method for assessing brain collateral circulation, comprising: acquiring a CT angiography image of a brain to be evaluated; and inputting the CT angiographic image of the brain to be evaluated into an evaluation model trained according to the plurality of embodiments in the first aspect to evaluate the side branch circulation of the brain, so as to obtain an evaluation result of evaluating the side branch circulation of the brain.
In a third aspect, the present application provides an apparatus for assessing brain collateral circulation, comprising: a processor; and a memory having stored therein program instructions for training an assessment model for assessing a brain side branch cycle, which when executed by the processor, cause the apparatus to implement the plurality of embodiments of the aforementioned first aspect; or program instructions for assessing a brain side branch cycle, which when executed by the processor, cause the apparatus to implement an embodiment of the aforementioned second aspect.
In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon computer-readable instructions for training an assessment model for assessing a brain side branch cycle and for assessing a brain side branch cycle, which when executed by one or more processors, implement the various embodiments of the foregoing first aspect and the embodiments of the foregoing second aspect.
By the method for training the evaluation model for evaluating the brain collateral circulation provided above, the embodiment of the application firstly uses the segmentation model to segment the brain tissue region from the CT angiography image marked with the brain collateral circulation score in an artificial intelligence mode. And then, extracting the left and right brain regions and the target layer according to the brain tissue region, calculating a maximum density projection image, inputting the maximum density projection image and the brain collateral circulation scoring mark into a main model autonomous learning optimization evaluation model, and directly outputting an evaluation result by using the trained evaluation model. Therefore, the evaluation efficiency and the accuracy of the evaluation result are greatly improved. Furthermore, in the embodiment of the application, the brain tissue region and the brain tissue map are registered, namely, the segmented brain tissue region is corrected, so that the positions of the left and right brain regions and the target layer are ensured to be accurate, the accuracy of the calculated maximum density projection image is improved, and the accuracy of the evaluation model is improved. In addition, the embodiment of the application also carries out gray level layering on the cerebral blood vessels, and avoids the influence on the reliability of the evaluation model caused by missing important pixel points.
Drawings
The above, as well as additional purposes, features, and advantages of exemplary embodiments of the present application will become readily apparent from the following detailed description when read in conjunction with the accompanying drawings. Several embodiments of the present application are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar or corresponding parts and in which:
FIG. 1 is an exemplary flow diagram illustrating a method for training an assessment model for assessing brain side branch circulation in accordance with an embodiment of the present application;
FIG. 2 is an exemplary schematic diagram illustrating the location of a middle cerebral artery blood supply region in accordance with embodiments of the present application;
FIG. 3 is an exemplary diagram illustrating an ensemble for training an assessment model for assessing brain side branch circulation in accordance with an embodiment of the present application;
FIG. 4 is an exemplary flow diagram illustrating a method for assessing brain side branch circulation in accordance with an embodiment of the present application; and
fig. 5 is an exemplary block diagram illustrating an apparatus for assessing brain collateral circulation according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the embodiments described in this specification are only some embodiments provided herein for the purpose of facilitating a clear understanding of the solution and meeting legal requirements, and not all embodiments of the application may be implemented. All other embodiments, which can be made by those skilled in the art without the exercise of inventive faculty, are intended to be within the scope of the present application, based on the embodiments disclosed in this specification.
Fig. 1 is an exemplary flow diagram illustrating a method 100 for training an assessment model for assessing brain side branch circulation in accordance with an embodiment of the present application. The evaluation model of the embodiment of the application comprises a segmentation model and a main model. As shown in fig. 1, at step S101, a CT angiography image for a brain is acquired and a brain collateral circulation score labeling is performed on the CT angiography image. In one implementation scenario, the aforementioned CT angiographic image may be, for example, a single phase CTA (sCTA) image, or may be, for example, a multi-phase CTA (mCTA) image, as the application is not limited in this respect. Preferably, embodiments of the present application employ sCTA images. In one implementation scenario, based on the acquired CTA image for the brain, it may be labeled with a brain collateral circulation score employing, for example, a pial blood vessel collateral circulation score (also referred to as the Tan score). It will be appreciated that the foregoing Tan scoring is to score the filling number of the cerebral blood supply area pia mater vessels of the middle cerebral artery ("MCA") using maximum intensity projection (Maximal Intensity Projection, "MIP") techniques. A score of 0 indicates that the MCA region is completely free of blood supply to the leptomeningeal side branch vessels; a score 1 indicates 0 to 50% of the leptomeningeal side branch vessel filling and supplying MCA occlusion region, a score 2 indicates 50% to 100% of the leptomeningeal side branch vessel filling and supplying MCA occlusion region, and a score 3 indicates 100% of the leptomeningeal side branch vessel filling and supplying occlusion region. In some embodiments, the brain side branch circulation score labeling can be performed on the CTA image through scoring by calculating the blood vessel occupation area values of the left side and the right side in the MIP map to correspond to the scores (such as 0,1,2 or 3).
Next, at step S102, the CT angiography image is segmented using a segmentation model to obtain a brain tissue region. In one implementation scenario, the segmentation model in the evaluation model of the embodiments of the present application may be, for example, a convolutional neural network, and by inputting a CTA image into the segmentation model, an accurate brain tissue region, including a bone or other tissue, may be extracted. Based on the obtained brain tissue region, at step S103, the brain left and right regions and the target layer are extracted based on the brain tissue region. In one embodiment, the target layer may be, for example, an intermediate layer of a middle cerebral artery blood supply region. It should be understood that when CTA images of a subject are acquired, there may be a situation that the head posture of the subject is askew, resulting in inaccurate positions of the left and right brain regions and the target layer, so that when the left and right brain regions and the target layer are extracted, for example, an image registration method may be adopted to "align" the brain tissue regions, and then the aligned left and right brain regions and the target layer are extracted.
Specifically, in one embodiment, a brain tissue region is registered with a brain tissue map to obtain a registered brain tissue region, and then left and right brain regions and a target layer are extracted from the registered brain tissue region. In one implementation scenario, the registered brain tissue region may be obtained by calculating a transformation matrix of the brain tissue region to the brain tissue map, and further transforming the brain tissue region according to the transformation matrix. That is, the brain tissue region is aligned ("aligned") according to its spatial transformation relationship (i.e., transformation matrix) with the brain tissue map using the brain tissue map as a reference frame, thereby obtaining a registered brain tissue region. In one embodiment, the registered brain tissue regions may be matched in position based on the positional information in the brain tissue map to extract the brain left and right regions and the target layer. By matching the registered brain tissue region with the brain tissue map, the target layer and the central axis thereof can be determined, so that the left and right brain regions can be determined.
After the above-described brain left and right regions and target layers are extracted, a maximum intensity projection ("MIP") image associated with evaluating the brain collateral circulation is calculated from the brain left and right regions and target layers at step S104. In one embodiment, a MIP image of a layer thickness may be obtained from the target layer. As an example, a target layer having a layer thickness of 25mm to 45mm may be provided, and a pixel having the greatest density is retained in the target layer having a certain layer thickness, so that a MIP image can be obtained. It will be appreciated that MIP can reflect the X-ray attenuation values of the corresponding pixels, and that smaller density variations can be displayed on the MIP image, which can well reveal stenosis, dilation, filling defects of the vessel, and distinguish calcifications on the vessel wall from contrast agent in the vessel lumen. Thus, the condition of filling defect of the blood vessel can be better reflected by calculating the MIP image, thereby being beneficial to evaluating side branch circulation.
Based on the obtained MIP image, at step S105, the maximum intensity projection image and the brain collateral circulation score label are input to the main model for training to train the evaluation model for evaluating the brain collateral circulation. In one implementation scenario, the primary model in the assessment model of an embodiment of the present application may include, but is not limited to, a decision tree model, a random forest model, or a neural network model. The training of the assessment model for assessing the brain collateral circulation is achieved by inputting the MIP images and the brain collateral circulation scoring labels into, for example, a decision tree model for autonomous learning optimization. The trained evaluation model can directly output the Tan score to obtain the evaluation result of the brain collateral circulation. In one embodiment, considering different contrast and tissue conditions of left and right brain regions, the application further proposes to perform gray scale layering on brain blood vessels in the maximum density projection image so as to obtain gray scale characteristics of the brain blood vessels, and further input gray scale characteristics and brain collateral circulation scoring labels into a main model for training so as to train an assessment model for assessing brain collateral circulation.
Specifically, in one implementation scenario, gray scale layering is performed on the cerebral blood vessels based on different gray scale thresholds by setting different gray scale thresholds according to the contrast of the left and right regions of the brain in the maximum density projection image, so as to obtain gray scale characteristics of the cerebral blood vessels of different levels of the left and right regions of the brain in the target layer. As an example, the gradation threshold value may be set to [100,150,200,240], and based on the gradation threshold value, gradation characteristics of the brain blood vessels of different gradation, which are the number of pixels satisfying each gradation threshold value, may be extracted. Based on this, omission of important pixel points can be avoided, thereby improving reliability of the evaluation model.
As can be seen from the above description, in the embodiment of the present application, the brain tissue region is segmented by using the segmentation model, and the left and right brain regions and the target layer are extracted according to the brain tissue region, and the maximum density projection image is calculated, so that the maximum density projection image and the brain collateral circulation score label are input into the main model autonomous learning optimization assessment model. Based on the method, the brain collateral circulation is evaluated in an artificial intelligence mode, so that the evaluation efficiency and the accuracy of the evaluation result are greatly improved. Further, according to the embodiment of the application, the brain tissue region and the brain tissue map are registered, so that the extracted positions of the left and right brain regions and the target layer are accurate, the accuracy of the calculated maximum density projection image is improved, and the accuracy of the evaluation model is improved. In addition, the embodiment of the application also carries out gray level layering on the cerebral blood vessels, so that important pixel points can be prevented from being omitted, and the reliability of the evaluation model is improved.
Fig. 2 is an exemplary schematic diagram illustrating the location of a middle cerebral artery blood supply region in accordance with an embodiment of the present application. As exemplarily shown in fig. 2, three layers are shown with R1, R2, and R3, respectively, for multiple layers containing a middle cerebral artery blood supply region extracted for a brain tissue region segmented out of a CTA image using a segmentation model (e.g., a convolutional neural network). In an implementation scenario, one of the target layers may be selected as referred to in the context of the present application, e.g. the R3 layer may be selected as the target layer. Then, a certain layer thickness (for example, 25mm to 45 mm) can be set on the basis of the target layer, the MIP image is obtained by reserving the pixel with the maximum density in the target layer with the certain layer thickness, and the MIP image and the brain collateral circulation score mark are input into a main model (for example, a decision tree model) for autonomous learning training, so that the evaluation result of the brain collateral circulation is obtained.
Fig. 3 is an exemplary diagram illustrating an ensemble for training an assessment model for assessing brain side branch circulation according to an embodiment of the present application. As shown in fig. 3, the evaluation model of an embodiment of the present application comprises a segmentation model 301 and a main model 302, wherein the segmentation model 301 may be, for example, a convolutional neural network and the main model 302 may be, for example, a decision tree model. In one implementation scenario, the brain tissue region 304 is obtained by inputting a CTA image 303 for the brain into the segmentation model 301 for segmentation. From the foregoing, it can be seen that when a CTA image of a subject is acquired, there may be a case that the head pose of the subject is skewed, so that the embodiment of the present application obtains a registered brain tissue region 306 by registering 311 the brain tissue region 304 with the brain tissue map 305. Specifically, the brain tissue region is transformed according to the transformation matrix by calculating the transformation matrix from the brain tissue region to the brain tissue map to obtain the registered brain tissue region 306.
Further, by performing, for example, position matching on the registered brain tissue region 306 based on the positional information in the brain tissue map, the brain left and right regions and the target layer 307 can be extracted. Next, a MIP map 308 can be calculated from the left and right brain regions and the target layer 307. Based on the method, the positions of the left and right brain regions and the target layer can be accurately determined, so that an accurate MIP map is obtained, and the accuracy of the evaluation model is improved. In some embodiments, in order to avoid missing important pixels, the embodiments of the present application further perform gray scale layering on the MIP map 308 to extract gray scale features 309 of different levels, and input the gray scale features 309 of different levels to the main model 302 for training, so as to train an evaluation model for evaluating brain collateral circulation. The evaluation model trained by the foregoing embodiments may output a Tan score 310, from which an evaluation result of the brain collateral circulation is obtained. For example, when the score is 3, it indicates that 100% of the leptomeningeal side branch vessels are filled to supply MCA occlusion area, i.e. brain side branch circulation is most effective.
Fig. 4 is an exemplary flow diagram illustrating a method 400 for assessing brain side branch circulation in accordance with an embodiment of the present application. As shown in fig. 4, at step S401, a CT angiographic image of the brain to be evaluated is acquired. In one embodiment, the CT angiography image may be, for example, a sCTA image or a mCTA image. Next, at step S402, the CT angiographic image of the brain to be evaluated is input into the trained evaluation model for evaluation of the brain collateral circulation, so as to obtain an evaluation result of evaluating the brain collateral circulation. Similar to the training process, when the trained evaluation model is adopted to perform brain collateral circulation evaluation, the brain collateral circulation evaluation method also firstly performs segmentation on a CTA image of a brain to be evaluated through a segmentation model to obtain a brain tissue region. Then, the left and right brain regions and the target layer are extracted by registering the brain tissue regions, and MIP images are calculated, and the MIP images are input into the main model for evaluation, and a Tan score is output, so that an evaluation result of the brain collateral circulation is obtained. Based on the above, through the scheme of the embodiment of the application, the evaluation efficiency and the accuracy of the evaluation result can be greatly improved.
Fig. 5 is an exemplary block diagram illustrating an apparatus 500 for assessing brain collateral circulation in accordance with an embodiment of the present application. It is to be appreciated that the device implementing aspects of the subject application may be a single device (e.g., a computing device) or a multi-function device including various peripheral devices.
As shown in fig. 5, the apparatus of the present application may include a central processing unit or central processing unit ("CPU") 511, which may be a general purpose CPU, a special purpose CPU, or other information processing and program running execution unit. Further, device 500 may also include a mass memory 512 and a read only memory ("ROM") 513, wherein mass memory 512 may be configured to store various types of data including various CT angiographic images of the brain, brain tissue regions, target layers, maximum density projection images, algorithm data, intermediate results, and various programs required to operate device 500. ROM 513 may be configured to store data and instructions necessary to power-on self-test for device 500, initialization of functional modules in the system, drivers for basic input/output of the system, and boot the operating system.
Optionally, the device 500 may also include other hardware platforms or components, such as a tensor processing unit ("TPU") 514, a graphics processing unit ("GPU") 515, a field programmable gate array ("FPGA") 516, and a machine learning unit ("MLU") 517, as shown. It will be appreciated that while various hardware platforms or components are shown in device 500, this is by way of example only and not limitation, and that one of skill in the art may add or remove corresponding hardware as desired. For example, device 500 may include only a CPU, associated memory devices, and interface devices to implement the methods of the present application for training an assessment model for assessing a brain side branch cycle and methods for assessing a brain side branch cycle.
In some embodiments, to facilitate the transfer and interaction of data with external networks, the device 500 of the present application further comprises a communication interface 518, such that it may be connected to a local area network/wireless local area network ("LAN/WLAN") 505 via the communication interface 518, and further to a local server 506 or to the Internet ("Internet") 507 via the LAN/WLAN. Alternatively or additionally, the device 500 of the present application may also be directly connected to the internet or cellular network via the communication interface 518 based on wireless communication technology, such as wireless communication technology based on generation 3 ("3G"), generation 4 ("4G"), or generation 5 ("5G"). In some application scenarios, the device 500 of the present application may also access the server 508 and database 509 of the external network as needed to obtain various known algorithms, data and modules, and may store various data remotely, such as various types of data or instructions for presenting, for example, CT angiographic images of the brain, brain tissue regions, target layers, maximum-density projection images, etc.
The peripheral devices of the apparatus 500 may include a display device 502, an input device 503, and a data transmission interface 504. In one embodiment, display device 502 may include, for example, one or more speakers and/or one or more visual displays configured for voice prompts and/or visual display of an assessment model of the present application for training to assess brain side branch circulation and for assessing brain side branch circulation. The input device 503 may include other input buttons or controls, such as a keyboard, mouse, microphone, gesture-capturing camera, etc., configured to receive input of audio data and/or user instructions. The data transfer interface 504 may include, for example, a serial interface, a parallel interface, or a universal serial bus interface ("USB"), a small computer system interface ("SCSI"), serial ATA, fireWire ("FireWire"), PCI Express, and high definition multimedia interface ("HDMI"), etc., configured for data transfer and interaction with other devices or systems. According to aspects of the subject application, the data transmission interface 504 may receive CT angiography images of the brain acquired from the image acquisition device and transmit CT angiography images including the brain or various other types of data or results to the device 500.
The above-described CPU 511, mass memory 512, ROM 513, TPU 514, GPU 515, FPGA 516, MLU 517, and communication interface 518 of the device 500 of the present application may be interconnected by a bus 519 and data interaction with peripheral devices may be accomplished by the bus. In one embodiment, CPU 511 may control other hardware components in device 500 and its peripherals through this bus 519.
An apparatus for assessing brain collateral circulation that may be used to perform the present application is described above in connection with fig. 5. It is to be understood that the device structure or architecture herein is merely exemplary, and that the implementation and implementation entities of the present application are not limited thereto, but that changes may be made without departing from the spirit of the present application.
Those skilled in the art will also appreciate from the foregoing description, taken in conjunction with the accompanying drawings, that embodiments of the present application may also be implemented in software programs. The present application thus also provides a computer readable storage medium having stored thereon computer readable instructions for training an assessment model for assessing a brain side branch cycle and for assessing a brain side branch cycle, which when executed by one or more processors, may be used to implement the method for training an assessment model for assessing a brain side branch cycle and the method for assessing a brain side branch cycle described herein in connection with fig. 1 and 4.
It should be noted that although the operations of the methods of the present application are depicted in the drawings in a particular order, this does not require or imply that the operations must be performed in that particular order or that all of the illustrated operations be performed in order to achieve desirable results. Rather, the steps depicted in the flowcharts may change the order of execution. Additionally or alternatively, certain steps may be omitted, multiple steps combined into one step to perform, and/or one step decomposed into multiple steps to perform.
It should be understood that when the terms "first," "second," "third," and "fourth," etc. are used in the claims, the specification and the drawings of this application, they are used merely to distinguish between different objects and not to describe a particular sequence. The terms "comprises" and "comprising," when used in the specification and claims of this application, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the present application is for the purpose of describing particular embodiments only, and is not intended to be limiting of the application. As used in the specification and claims of this application, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should be further understood that the term "and/or" as used in the present specification and claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
Although the embodiments of the present application are described above, the content is only an example adopted for understanding the present application, and is not intended to limit the scope and application scenario of the present application. Any person skilled in the art can make any modifications and variations in form and detail without departing from the spirit and scope of the disclosure, but the scope of the disclosure is still subject to the scope of the claims.
Claims (10)
1. A method for training an assessment model for assessing brain collateral circulation, wherein the assessment model comprises a segmentation model and a main model, and the method comprises:
acquiring a CT angiography image aiming at the brain and marking the brain collateral circulation score of the CT angiography image;
segmenting the CT angiography image using the segmentation model to obtain a brain tissue region;
extracting left and right brain regions and a target layer based on the brain tissue region;
calculating a maximum density projection image related to evaluation of brain collateral circulation according to the brain left and right areas and the target layer; and
inputting the maximum density projection image and the brain collateral circulation scoring label into the main model for training so as to train an evaluation model for evaluating brain collateral circulation.
2. The method of claim 1, wherein extracting left and right brain regions and a target layer based on the brain tissue region comprises:
registering the brain tissue region with the brain tissue map to obtain a registered brain tissue region; and
and extracting the left and right brain regions and the target layer according to the registered brain tissue regions.
3. The method of claim 2, wherein registering the brain tissue region with a brain tissue map to obtain a registered brain tissue region comprises:
calculating a transformation matrix from the brain tissue region to the brain tissue map; and
transforming the brain tissue region according to the transformation matrix to obtain the registered brain tissue region.
4. The method of claim 3, wherein extracting the left and right brain regions and the target layer from the registered brain tissue regions comprises:
and performing position matching on the registered brain tissue regions based on the position information in the brain tissue map so as to extract the left and right brain regions and the target layer.
5. The method of claim 1, further comprising:
gray scale layering is carried out on the brain blood vessels in the maximum density projection image so as to obtain gray scale characteristics of the brain blood vessels; and
and inputting the gray scale characteristics and the brain collateral circulation scoring labels into the main model for training so as to train an evaluation model for evaluating brain collateral circulation.
6. The method of claim 5, wherein grayscaling brain blood vessels in the maximum intensity projection image to obtain gray scale features of brain blood vessels comprises:
setting different gray threshold values according to the contrast ratio of the left and right brain regions in the maximum density projection image; and
and carrying out gray level layering on the cerebral blood vessels based on the different gray level thresholds so as to obtain gray level characteristics of the cerebral blood vessels of different levels of the left and right regions of the brain in the target layer.
7. The method of any of claims 1-6, wherein the master model comprises a decision tree model, a random forest model, or a neural network model.
8. A method for assessing brain collateral circulation, comprising:
acquiring a CT angiography image of a brain to be evaluated; and
inputting the CT angiographic image of the brain to be evaluated into an evaluation model trained according to the method of any one of claims 1-7 for evaluation of the brain collateral circulation, so as to obtain an evaluation result for evaluating the brain collateral circulation.
9. An apparatus for assessing brain collateral circulation, comprising:
a processor; and
a memory having stored therein program instructions for training an assessment model for assessing a brain side branch cycle, which when executed by the processor, cause the apparatus to carry out the method according to any one of claims 1-7; or program instructions for assessing brain collateral circulation, which when executed by the processor, cause the device to carry out the method according to claim 8.
10. A computer readable storage medium having stored thereon computer readable instructions for training an assessment model for assessing a brain side branch cycle and for assessing a brain side branch cycle, which when executed by one or more processors, implement the method of any one of claims 1-7 and implement the method of claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311102535.5A CN117291281A (en) | 2023-08-29 | 2023-08-29 | Method for training evaluation model for evaluating brain collateral circulation and related product |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311102535.5A CN117291281A (en) | 2023-08-29 | 2023-08-29 | Method for training evaluation model for evaluating brain collateral circulation and related product |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117291281A true CN117291281A (en) | 2023-12-26 |
Family
ID=89250810
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311102535.5A Pending CN117291281A (en) | 2023-08-29 | 2023-08-29 | Method for training evaluation model for evaluating brain collateral circulation and related product |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117291281A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111445575A (en) * | 2020-04-01 | 2020-07-24 | 沈阳先进医疗设备技术孵化中心有限公司 | Image reconstruction method and device of Wirisi ring, electronic device and storage medium |
| CN112907563A (en) * | 2021-03-17 | 2021-06-04 | 复旦大学附属华山医院 | CTA full-automatic collateral circulation scoring method and system based on deep learning |
| CN113222974A (en) * | 2021-05-31 | 2021-08-06 | 沈阳先进医疗设备技术孵化中心有限公司 | Automatic evaluation method and device for collateral circulation |
| CN114209344A (en) * | 2021-11-05 | 2022-03-22 | 东软医疗系统股份有限公司 | Collateral circulation state evaluation method and device, storage medium and electronic equipment |
| CN116309909A (en) * | 2023-03-07 | 2023-06-23 | 北京羽医甘蓝信息技术有限公司 | Oral cavity scanning image processing method and device |
| CN116491968A (en) * | 2023-03-09 | 2023-07-28 | 珠海中科先进技术研究院有限公司 | Acute ischemic stroke collateral circulation assessment method based on single-phase CTA |
-
2023
- 2023-08-29 CN CN202311102535.5A patent/CN117291281A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111445575A (en) * | 2020-04-01 | 2020-07-24 | 沈阳先进医疗设备技术孵化中心有限公司 | Image reconstruction method and device of Wirisi ring, electronic device and storage medium |
| CN112907563A (en) * | 2021-03-17 | 2021-06-04 | 复旦大学附属华山医院 | CTA full-automatic collateral circulation scoring method and system based on deep learning |
| CN113222974A (en) * | 2021-05-31 | 2021-08-06 | 沈阳先进医疗设备技术孵化中心有限公司 | Automatic evaluation method and device for collateral circulation |
| CN114209344A (en) * | 2021-11-05 | 2022-03-22 | 东软医疗系统股份有限公司 | Collateral circulation state evaluation method and device, storage medium and electronic equipment |
| CN116309909A (en) * | 2023-03-07 | 2023-06-23 | 北京羽医甘蓝信息技术有限公司 | Oral cavity scanning image processing method and device |
| CN116491968A (en) * | 2023-03-09 | 2023-07-28 | 珠海中科先进技术研究院有限公司 | Acute ischemic stroke collateral circulation assessment method based on single-phase CTA |
Non-Patent Citations (1)
| Title |
|---|
| 王静杰: "基于全脑4D-CTA/CTP技术的急性缺血性卒中患者侧支循环评估研究", 《中国博士论文全文数据库 医药卫生科技辑》, 15 February 2023 (2023-02-15), pages 070 - 90 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Torosdagli et al. | Deep geodesic learning for segmentation and anatomical landmarking | |
| CN108648172B (en) | CT (computed tomography) map pulmonary nodule detection system based on 3D-Unet | |
| CN104346821B (en) | Automatic planning for medical imaging | |
| WO2020182036A1 (en) | Image processing method and apparatus, server, and storage medium | |
| Guo et al. | Dual attention enhancement feature fusion network for segmentation and quantitative analysis of paediatric echocardiography | |
| CN110298844B (en) | X-ray radiography image blood vessel segmentation and identification method and device | |
| US20190131012A1 (en) | Image search device, image search method, and image search program | |
| EP3657437B1 (en) | Automatic detection and quantification of the aorta from medical images | |
| CN109509177B (en) | Method and device for recognizing brain image | |
| Jamart et al. | Mini review: Deep learning for atrial segmentation from late gadolinium-enhanced MRIs | |
| CN110880366B (en) | Medical image processing system | |
| US20090257550A1 (en) | Slice image display apparatus, method and recording-medium having stored therein program | |
| CN111899244B (en) | Image segmentation method, network model training method, device and electronic equipment | |
| CN111145160B (en) | Method, device, server and medium for determining coronary artery branches where calcified regions are located | |
| Popescu et al. | Anatomically informed deep learning on contrast-enhanced cardiac magnetic resonance imaging for scar segmentation and clinical feature extraction | |
| CN110751187A (en) | Training method of abnormal area image generation network and related product | |
| CN110522449B (en) | Plaque parting method and device, electronic equipment and readable storage medium | |
| US20220004797A1 (en) | Linear structure extraction device, method, program, and learned model | |
| Yang et al. | Benchmarking the CoW with the TopCoW Challenge: Topology-Aware Anatomical Segmentation of the Circle of Willis for CTA and MRA | |
| CN117291281A (en) | Method for training evaluation model for evaluating brain collateral circulation and related product | |
| US20230115927A1 (en) | Systems and methods for plaque identification, plaque composition analysis, and plaque stability detection | |
| CN111899273A (en) | Image segmentation method, computer device and storage medium | |
| CN116452579A (en) | Chest radiography image-based pulmonary artery high pressure intelligent assessment method and system | |
| CN113393445B (en) | Breast cancer image determination method and system | |
| JPWO2019220871A1 (en) | Chest X-ray image abnormality display control method, abnormality display control program, abnormality display control device, and server device |
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 |