Automatic screening method and system for nuclear magnetic resonance images

Abstract

The invention relates to the technical field of medical image processing, in particular to a nuclear magnetic resonance image automatic screening method and system. The method comprises the following steps: acquiring nuclear magnetic resonance image data, and preprocessing the image data; dividing the region of the image into a plurality of sub-regions, identifying the position features of each sub-region, filling the position regions according to the position features, and outputting position data; dividing the position area into a plurality of mark areas, identifying mark points of each mark area, and integrating mark point data; performing pathological screening on the mark point data, and outputting a screening result; the system comprises an image data preprocessing module, a position feature recognition module, a mark point recognition module and a pathology screening module. Through the mode, after the nuclear magnetic resonance image is shot, pathological screening is carried out on the image, so that an attending physician is prompted about the pathological condition of a patient, and the attending physician is assisted to carry out judgment.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to a method and system for automatically screening nuclear magnetic resonance images. Background Art

[0002] Magnetic resonance imaging (MRI) is a commonly used medical imaging technology that uses magnetic fields and radio frequency pulses to make human tissue generate signals, which are then processed and analyzed by computers to generate images of the internal structure of the human body.

[0003] At present, when analyzing MRI images, a dedicated person or the patient needs to hand them over to the attending physician, who then determines the pathological conditions in the MRI images. In the above method, the attending physician is mainly responsible for the on-site judgment, and the patient may also hand over the previous MRI images to the attending physician. When the attending physician makes a judgment on the patient's pathological condition, the key points of the judgment are unclear, which affects the judgment time.

[0004] Therefore, it is very necessary to propose an automatic screening method and system for nuclear magnetic resonance images, which can perform pathological screening on the images after they are taken, so as to prompt the attending physician of the patient's pathological condition, thereby assisting the attending physician in making judgments. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for automatic screening of nuclear magnetic resonance images, aiming to realize pathological screening of the images after the nuclear magnetic resonance images are taken, so as to prompt the attending physician of the patient's pathological condition, thereby assisting the attending physician in making a judgment.

[0006] To achieve the above object, the present invention adopts an automatic screening method of nuclear magnetic resonance images, comprising the following steps: Acquire nuclear magnetic resonance image data and preprocess the image data; Divide the image area into multiple sub-areas, identify the position features of each sub-area, fill the position area according to the position features, and output the position data; Divide the location area into multiple landmark areas, identify landmark points of each landmark area, and integrate landmark point data; Perform pathological screening on the landmark point data and output the screening results.

[0007] Among them, in the step of acquiring nuclear magnetic resonance image data and preprocessing the image data: After receiving the magnetic resonance imaging data, image format conversion is performed, and anterior commissure and posterior commissure correction is performed; The resolution of the image is unified, the non-uniform tissue intensity in the image is corrected, and the pre-processed image data is output.

[0008] Among them, in the steps of dividing the region of the image into a plurality of sub-regions, identifying the position features of each sub-region, filling the position region according to the position features, and outputting the position data: According to the positioning data, the image area is divided into multiple sub-areas; Match the position features in the sub-region to obtain the shape, size, and edge information of the position features; Match the edge features in the sub-region, and fill the position region according to the edge features to obtain the position data.

[0009] After the step of dividing the image region into a plurality of sub-regions according to the positioning data: Defines the association information between multiple sub-areas.

[0010] Wherein, after the step of matching the edge features in the sub-region and filling the position region according to the edge features to obtain the position data: Match the position data with the patient's examination position and output matching information.

[0011] Among them, in the steps of dividing the location area into a plurality of landmark areas, identifying landmark points of each landmark area, and integrating landmark point data: Dividing the location area into multiple marker areas, and defining association information between the multiple marker areas; Features related to the pathological location are extracted in the landmark area as landmark points.

[0012] Among them, after the step of extracting features related to the pathological position in the landmark area as landmark points: Locate each landmark point, integrate each landmark point, and obtain the landmark point positioning data.

[0013] Among them, in the step of performing pathological screening on the landmark data and outputting the screening results: According to the different weighted images of the image data, the morphological characteristics of the landmark point data are recognized and the screening results are output.

[0014] The present invention also provides an automatic screening system for nuclear magnetic resonance images, comprising an image data preprocessing module, a position feature recognition module, a landmark recognition module and a pathology screening module; wherein: The image data preprocessing module is used to obtain nuclear magnetic resonance image data and preprocess the image data; The position feature recognition module is used to divide the image area into multiple sub-areas, identify the position feature of each sub-area, fill the position area according to the position feature, and output the position data; The landmark point recognition module is used to divide the location area into multiple landmark areas, identify the landmark points of each landmark area, and integrate the landmark point data; The pathology screening module is used to perform pathology screening on the landmark point data and output the screening results.

[0015] The present invention provides an automatic screening method and system for nuclear magnetic resonance images, which respectively use the image data preprocessing module, the position feature recognition module, the landmark point recognition module and the pathology screening module to perform the following operations: acquiring nuclear magnetic resonance image data and preprocessing the image data; dividing the image area into multiple sub-areas, identifying the position feature of each sub-area, filling the position area according to the position feature, and outputting the position data; dividing the position area into multiple landmark areas, identifying the landmark point of each landmark area, and integrating the landmark point data; performing pathology screening on the landmark point data and outputting the screening result; and realizing pathology screening of the nuclear magnetic resonance image after the image is taken, so as to prompt the attending physician of the patient's pathological condition, thereby assisting the attending physician in making a judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is a schematic diagram of the process of the automatic screening method of nuclear magnetic resonance images of the present invention.

[0018] Figure 2 It is a flow chart of the steps of the automatic screening method of nuclear magnetic resonance images of the present invention.

[0019] Figure 3 It is a step flow chart of S100 of the present invention.

[0020] Figure 4 It is a step flow chart of S200 of the present invention.

[0021] Figure 5 It is a step flow chart of S300 of the present invention.

[0022] Figure 6 It is a structural principle diagram of the automatic screening system for nuclear magnetic resonance images of the present invention.

[0023] Figure 7 It is a structural principle diagram of the electronic equipment of the present invention.

[0024] 501 - image data preprocessing module, 502 - position feature recognition module, 503 - landmark recognition module, 504 - pathology screening module. DETAILED DESCRIPTION

[0025] Here, exemplary embodiments will be described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0026] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0027] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0028] See also Figure 1~Figure 5 The present invention provides a method for automatically screening nuclear magnetic resonance images, comprising the following steps: S100: Acquire nuclear magnetic resonance image data and preprocess the image data; In this embodiment, the nuclear magnetic resonance image data is first acquired and preprocessed; the specific process is as follows: S101: after receiving the nuclear magnetic resonance image data, image format conversion is performed, and anterior commissure and posterior commissure correction is performed; S102: unify the resolution of the image, correct the non-uniform tissue intensity in the image, and output the pre-processed image data.

[0029] The raw data of MRI images are usually stored in DICOM format, which is a standard format commonly used in medical imaging. In order to facilitate subsequent processing, it is usually necessary to convert the DICOM format to NIfTI format (.nii extension). Use software such as MIPAV to perform anterior commissure (AC) and posterior commissure (PC) correction to ensure that the image is anatomically correctly aligned. Resample the image to a standard 256×256×256 mode to unify the resolution of the image. Use the N3 algorithm or other methods to correct the non-uniform tissue intensity in the image to improve image quality.

[0030] S200: Divide the image area into multiple sub-areas, identify the position features of each sub-area, fill the position area according to the position features, and output position data; In this embodiment, the region of the preprocessed image is divided into multiple sub-regions, the position features of each sub-region are identified, and the position region is filled according to the position features, and the position data is output; the specific process is as follows: S201: Divide the image area into multiple sub-areas according to the positioning data, and define association information between the multiple sub-areas; S202: Matching the position features in the sub-areas to obtain the shape, size, and edge information of the position features; S203: Match edge features in the sub-region according to the association information, and fill the position region according to the edge features to obtain position data; S204: Match the position data with the patient's examination position and output matching information.

[0031] Usually, there is a positioning map in the lower right corner of the MRI image, which can determine which lumbar intervertebral disc or other bone structure is being imaged. This positioning map can be used as positioning data. Based on the positioning data, the image area is divided into multiple sub-areas, and the association information between the multiple sub-areas is defined. The association information includes the adjacent sub-areas of a sub-area. Then, the position features in the sub-areas are matched, which are bone features. In MRI images, the signal intensity of bones is usually low, appearing as dark areas. This is because the proton density of bones is low, and the T1 and T2 relaxation times are relatively short in different weighted images, resulting in weak signals in MRI images. Bones present specific morphological structures in MRI images, such as the distinction between cortical bone and cancellous bone. Cortical bone is thicker and has a darker signal, while cancellous bone is relatively thinner and has a slightly brighter signal. The above process can train a machine learning or deep learning model, and the model is used to obtain the shape, size, and edge information of the position feature. According to the associated information, the edge features in the sub-area are matched, and the edge features are skin. According to the edge features, the corresponding sub-area is filled to obtain the position area and position data. The position data is matched with the patient's examination position. The position data can be matched with the data in the hospital database to determine whether there is a difference between the position data and the patient's examination position. If there is a difference, an alarm is issued.

[0032] S300: Divide the location area into a plurality of landmark areas, identify landmark points of each landmark area, and integrate landmark point data; In this embodiment, the location area is divided into a plurality of marker areas, the marker points of each marker area are identified, and the marker point data are integrated; the specific process is as follows: S301: Divide the location area to obtain multiple marker areas, and define association information between the multiple marker areas; S302: extracting features related to the pathological position in the landmark area as landmark points; S303: Locate each landmark point, and integrate each landmark point to obtain the positioning data of the landmark point.

[0033] First, the location area is divided to obtain a plurality of marker areas, and association information between the plurality of marker areas is defined; the association information includes adjacent marker areas of a certain marker area.

[0034] Using image processing algorithms and machine learning techniques, features related to the pathological location are extracted from the MRI images. These features may include texture features, shape features, color information, etc., which can describe the appearance of the pathological location in the MRI image. The extracted features are input into a pre-trained machine learning or deep learning model. These models may include support vector machines (SVM), random forests, convolutional neural networks (CNN), etc., which can locate the pathological location in the MRI image based on the learned features. Features related to the pathological location are extracted in the landmark area as landmark points; each landmark point is located, and each landmark point is integrated to obtain the landmark point positioning data.

[0035] S400: Perform pathological screening on the landmark data and output the screening results; In this embodiment, the signal characteristics of the pathological position are analyzed based on different weighted images of the nuclear magnetic resonance image. For example, on T1WI, some lesions may appear as low signals, while on T2WI, they may appear as high signals. The morphological characteristics of the pathological position are determined, such as size, shape, contour, etc. These morphological characteristics can provide clues about the nature of the lesion. For example, a malignant tumor may appear as a lump with irregular shape and blurred edges, while a benign tumor may appear as a nodule with regular shape and clear edges. After pathological screening, the screening results are output to prompt the attending physician of the patient's pathological condition, thereby assisting the attending physician in making a judgment.

[0036] Corresponding to the aforementioned embodiment of the method for automatic screening of nuclear magnetic resonance images, the present application also provides an embodiment of a system for automatic screening of nuclear magnetic resonance images.

[0037] Figure 6 FIG. 1 is a block diagram of an automatic screening system for nuclear magnetic resonance images according to an exemplary embodiment. Figure 6 The system may include: an image data preprocessing module 501, a position feature recognition module 502, a landmark recognition module 503 and a pathology screening module 504; wherein: The image data preprocessing module 501 is used to obtain nuclear magnetic resonance image data and preprocess the image data; The position feature recognition module 502 is used to divide the image area into multiple sub-areas, recognize the position feature of each sub-area, fill the position area according to the position feature, and output the position data; The landmark point recognition module 503 is used to divide the location area into multiple landmark areas, recognize the landmark points of each landmark area, and integrate the landmark point data; The pathology screening module 504 is used to perform pathology screening on the landmark data and output the screening result.

[0038] In this embodiment, the image data preprocessing module 501 acquires nuclear magnetic resonance image data and preprocesses the image data; the position feature recognition module 502 divides the image area into multiple sub-areas, identifies the position feature of each sub-area, fills the position area according to the position feature, and outputs the position data; the landmark point recognition module 503 divides the position area into multiple landmark areas, identifies the landmark points of each landmark area, and integrates the landmark point data; the pathology screening module 504 performs pathology screening on the landmark point data and outputs the screening results; after the nuclear magnetic resonance image is taken, the image is pathologically screened so as to prompt the attending physician of the patient's pathological condition, thereby assisting the attending physician in making a judgment.

[0039] Regarding the system in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0040] For the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application. A person of ordinary skill in the art can understand and implement it without creative work.

[0041] Accordingly, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned automatic screening method for nuclear magnetic resonance images. Figure 7As shown in FIG. 1 , a hardware structure diagram of any device with data processing capability in which a nuclear magnetic resonance image automatic screening system provided by an embodiment of the present invention is located, except that Figure 7 In addition to the processor, memory and network interface shown, any device with data processing capability in which the apparatus in the embodiment is located may also include other hardware, generally based on the actual functions of the device with data processing capability, which will not be described in detail.

[0042] Accordingly, the present application also provides a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed by the processor, the automatic screening method of nuclear magnetic resonance images as described above is implemented. The computer-readable storage medium can be an internal storage unit of any device with data processing capabilities described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device. Further, the computer-readable storage medium can also include both an internal storage unit of any device with data processing capabilities and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capabilities, and can also be used to temporarily store data that has been output or is to be output.

[0043] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the contents disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not disclosed in the present application.

[0044] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. A method for automatically screening nuclear magnetic resonance images, characterized in that: The steps include: Acquire nuclear magnetic resonance image data and preprocess the image data; Divide the image area into multiple sub-areas, identify the position features of each sub-area, fill the position area according to the position features, and output the position data; Divide the location area into multiple landmark areas, identify landmark points of each landmark area, and integrate landmark point data; Perform pathological screening on the landmark point data and output the screening results.

2. The method for automatically screening nuclear magnetic resonance images according to claim 1, characterized in that: In the step of acquiring nuclear magnetic resonance image data and preprocessing the image data: After receiving the magnetic resonance imaging data, image format conversion is performed, and anterior commissure and posterior commissure correction is performed; The resolution of the image is unified, the non-uniform tissue intensity in the image is corrected, and the pre-processed image data is output.

3. The automatic screening method for nuclear magnetic resonance images according to claim 1, characterized in that: In the steps of dividing the region of the image into a plurality of sub-regions, identifying the position features of each sub-region, filling the position region according to the position features, and outputting the position data: According to the positioning data, the image area is divided into multiple sub-areas; Match the position features in the sub-region to obtain the shape, size, and edge information of the position features; Match the edge features in the sub-region, and fill the position region according to the edge features to obtain the position data.

4. The method for automatically screening nuclear magnetic resonance images according to claim 3, characterized in that: After the step of dividing the image area into multiple sub-areas according to the positioning data: Defines the association information between multiple sub-areas.

5. The method for automatically screening nuclear magnetic resonance images according to claim 3, characterized in that: After the steps of matching edge features in the sub-region and filling the position region according to the edge features and obtaining the position data: Match the position data with the patient's examination position and output matching information.

6. The method for automatically screening nuclear magnetic resonance images according to claim 1, characterized in that: In the steps of dividing the location area into a plurality of landmark areas, identifying landmark points of each landmark area, and integrating landmark point data: Dividing the location area into multiple marker areas, and defining association information between the multiple marker areas; Features related to the pathological location are extracted in the landmark area as landmark points.

7. The method for automatically screening nuclear magnetic resonance images according to claim 6, characterized in that: After the step of extracting features related to the pathological location as landmark points in the landmark area: Locate each landmark point, integrate each landmark point, and obtain the landmark point positioning data.

8. The method for automatically screening nuclear magnetic resonance images according to claim 1, characterized in that: In the step of performing pathological screening on the landmark data and outputting the screening results: According to the different weighted images of the image data, the morphological characteristics of the landmark point data are recognized and the screening results are output.

9. A nuclear magnetic resonance image automatic screening system, applied to the nuclear magnetic resonance image automatic screening method as claimed in claim 1, characterized in that: It includes an image data preprocessing module, a position feature recognition module, a landmark recognition module and a pathology screening module; wherein: The image data preprocessing module is used to obtain nuclear magnetic resonance image data and preprocess the image data; The position feature recognition module is used to divide the image area into multiple sub-areas, identify the position feature of each sub-area, fill the position area according to the position feature, and output the position data; The landmark point recognition module is used to divide the location area into multiple landmark areas, identify the landmark points of each landmark area, and integrate the landmark point data; The pathology screening module is used to perform pathology screening on the landmark point data and output the screening results.

Images