CN115100189B - Three-dimensional medical image cutting method, electronic device and storage medium - Google Patents

Three-dimensional medical image cutting method, electronic device and storage medium Download PDF

Info

Publication number
CN115100189B
CN115100189B CN202210896276.7A CN202210896276A CN115100189B CN 115100189 B CN115100189 B CN 115100189B CN 202210896276 A CN202210896276 A CN 202210896276A CN 115100189 B CN115100189 B CN 115100189B
Authority
CN
China
Prior art keywords
cutting
dimensional medical
medical image
dimensional
historical
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
Application number
CN202210896276.7A
Other languages
Chinese (zh)
Other versions
CN115100189A (en
Inventor
邱峰
邱俊达
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Pinke Information Technology Co ltd
Original Assignee
Shanghai Pinke Information Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Pinke Information Technology Co ltd filed Critical Shanghai Pinke Information Technology Co ltd
Priority to CN202210896276.7A priority Critical patent/CN115100189B/en
Publication of CN115100189A publication Critical patent/CN115100189A/en
Application granted granted Critical
Publication of CN115100189B publication Critical patent/CN115100189B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • 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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • 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
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T19/00—Manipulating 3D models or images for computer graphics
    • G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/761—Proximity, similarity or dissimilarity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2200/00—Indexing scheme for image data processing or generation, in general
    • G06T2200/04—Indexing scheme for image data processing or generation, in general involving 3D image data
    • 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
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • 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
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Biomedical Technology (AREA)
  • Public Health (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Primary Health Care (AREA)
  • Epidemiology (AREA)
  • General Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Pathology (AREA)
  • Mathematical Physics (AREA)
  • Molecular Biology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Hardware Design (AREA)
  • Quality & Reliability (AREA)
  • Architecture (AREA)
  • Computer Graphics (AREA)
  • Apparatus For Radiation Diagnosis (AREA)

Abstract

The application relates to the technical field of medical image processing and provides a three-dimensional medical image cutting method, electronic equipment and a storage medium, wherein the three-dimensional medical image cutting method comprises the steps of obtaining similar cases corresponding to target cases and cutting information corresponding to the similar cases, obtaining historical cutting information corresponding to historical three-dimensional medical images of the target cases, and determining first cutting guide based on the similar cutting information and the historical cutting information; processing the current three-dimensional medical image of the target case based on a pre-trained neural network to obtain a second cutting guide; and cutting the current three-dimensional medical image of the target case along an orthogonal cutting direction or along a cutting direction of any angle based on the first cutting guide and the second cutting guide respectively to obtain a cutting plan with medical diagnosis significance. The accuracy and the efficiency of cutting the medical image are improved, and the medical diagnosis is assisted.

Description

Three-dimensional medical image cutting method, electronic device and storage medium
Technical Field
The present disclosure relates to the field of medical image processing technologies, and in particular, to a method for cutting a three-dimensional medical image, an electronic device, and a storage medium.
Background
With the development of medical technology, it is becoming more and more common to perform disease diagnosis through three-dimensional medical images, but how to quickly and accurately acquire corresponding case information based on three-dimensional medical images is becoming more and more important and urgent.
In the related art, a three-dimensional image is cut through medical experience of professional personnel to obtain a cutting plane graph with medical diagnosis significance, and further medical diagnosis is carried out, but the cutting efficiency and the cutting accuracy of the medical image are lower due to the fact that the cutting experience of related personnel is needed.
Disclosure of Invention
The present application aims to solve at least one of the technical problems existing in the prior art.
Therefore, the application provides a method for cutting a three-dimensional medical image.
The application also provides electronic equipment.
The present application also proposes a non-transitory computer readable storage medium.
The present application also proposes a computer program product.
According to an embodiment of the first aspect of the present application, a method for cutting a three-dimensional medical image includes: acquiring similar cases corresponding to the target cases and similar cutting information corresponding to similar medical images of the similar cases; acquiring historical cutting information corresponding to the historical three-dimensional medical image of the target case; a first cutting guideline determined based on the similar cutting information and the historical cutting information; processing the current three-dimensional medical image of the target case based on a pre-trained neural network to obtain a second cutting guide, wherein the neural network is obtained based on different types of medical data training, and the neural network is trained based on different network models and training parameters aiming at different types of medical data; and cutting the current three-dimensional medical image of the target case along an orthogonal cutting direction or along a cutting direction of any angle based on the first cutting guide and the second cutting guide respectively to obtain a cutting plan with medical diagnosis significance.
According to one embodiment of the present application, the method further comprises: judging whether similar cases corresponding to the target cases exist or not, and when the similar cases corresponding to the target cases exist, processing the three-dimensional medical images of the target cases by using a neural network trained based on case information corresponding to the similar cases to obtain similar cutting guide, wherein the similar cutting guide comprises similar cutting information; judging whether a historical three-dimensional medical image exists in the target case, and when the historical three-dimensional medical image corresponding to the target case exists, processing the three-dimensional medical image of the target case by using a neural network obtained by training cutting information based on different disease states of the same case to obtain historical cutting guide, wherein the historical cutting guide comprises historical cutting information; and determining a first cutting guide based on the similar cutting information and the historical cutting information.
According to an embodiment of the present application, the method for cutting a three-dimensional medical image includes a module for identifying an image type, where the image type identifying module is configured to detect and identify a three-dimensional medical image of the target case, so as to obtain an image type corresponding to the three-dimensional medical image; judging whether similar cases corresponding to the target case exist or not and whether historical three-dimensional medical images exist in the target case or not based on the identified image types; when a similar case corresponding to the target case exists, starting a neural network obtained by training based on case information corresponding to the similar case to process a three-dimensional medical image of the target case to obtain similar cutting guide, and when the similar case corresponding to the target case does not exist, closing the neural network obtained by training based on case information corresponding to the similar case to process the three-dimensional medical image of the target case; and when the historical cutting information corresponding to the target case exists, starting a neural network obtained by training based on the cutting information of different illness states of the same case to process the three-dimensional medical image of the target case to obtain historical cutting guidance, and when the historical cutting information corresponding to the target case does not exist, closing the step of processing the three-dimensional medical image of the target case by using the neural network obtained by training based on the cutting information of different illness states of the same case.
According to one embodiment of the application, the first and second cutting directions are used for respectively cutting the current three-dimensional medical image of the target case along an orthogonal cutting direction or along any angleCutting to obtain a cutting plane diagram with medical diagnosis significance, wherein the cutting plane diagram comprises the following steps: determining a reference point in the current three-dimensional medical image based on the first cutting guide and the second cutting guide, determining a plane perpendicular to a three-dimensional coordinate axis based on the reference point, and forming three mutually orthogonal cross-sectional images with diagnostic significance, wherein the cross-sectional images at least comprise a cross section, a coronal plane and a sagittal plane; or determining a plane with any azimuth in the current three-dimensional medical image based on the first cutting guide and the second cutting guide, determining a normal vector and an interior point corresponding to the plane, and performing plane cutting with any angle on the three-dimensional medical image based on the normal vector and the interior point to obtain image information of a cutting plane with diagnostic significance, wherein the coordinates of the interior point are (x 0 ，y 0 ，z 0 ) The normal vector of the plane is n (a 1 ，a 2 ，a 3 ) Satisfy a 1 (x-x 0 )+a 2 (y-y 0 )+a 3 (z-z 0 )。
According to one embodiment of the application, the normal vector is determined based on the direction angle of the plane, the direction angle comprises alpha, beta, gamma, and 0.ltoreq.alpha.ltoreq.pi, and 0.ltoreq.beta.ltoreq.pi, and 0.ltoreq.gamma.ltoreq.pi, if the three-dimensional origin of coordinates is O, the direction cosine of the vector OP is cos alpha, cos beta, cos gamma, and the normal vector n= (a) 1 ，a 2 ，a 3 ) Then
According to one embodiment of the present application, the method further comprises: moving the initial plane along the direction of the normal vector, moving the initial plane by a preset moving distance b to obtain a moved cutting plane, and translating m '(x) corresponding to the distance b for any point m (x', y ', z') 0 ，y 0 ，z 0 ) X ' =x+b×cos α, y ' =y+b×cos β, z ' =z+b×cos γ.
According to one embodiment of the present application, the obtaining similar cases corresponding to the target case and the cutting information corresponding to the similar cases, obtaining historical cutting information corresponding to the historical three-dimensional medical image of the target case, and determining the first cutting guide based on the similar cutting information and the historical cutting information, includes: acquiring medical data of the target case, wherein the medical data comprises case information, historical three-dimensional medical images and current three-dimensional medical images of the target case; acquiring similar medical data similar to the medical data of the target case from a medical database; and obtaining a first cutting guide based on the historical cutting information of the similar medical data and the historical cutting information of the target case.
According to one embodiment of the application, the three-dimensional medical image is constructed based on two-dimensional medical image slice data; the training step of the neural network comprises the following steps: acquiring three-dimensional medical images corresponding to different parts of a human body, three-dimensional medical images corresponding to different sampling parameters at the same part, and historical cutting information corresponding to each three-dimensional medical image; training to obtain neural networks corresponding to different parts of the human body based on the three-dimensional medical images and the historical cutting information corresponding to the three-dimensional medical images, and selecting the neural networks of the corresponding types to process when the three-dimensional medical images corresponding to the different parts of the human body are required to be cut so as to obtain corresponding second cutting guide.
According to one embodiment of the application, the three-dimensional medical image is constructed based on two-dimensional medical image slice data; the processing the current three-dimensional medical image of the target case based on the pre-trained neural network to obtain a second cutting guide comprises: performing model training based on the acquired two-dimensional medical image and the history cutting information corresponding to the two-dimensional medical image to obtain a two-dimensional neural network model; acquiring the three-dimensional medical images, and processing the two-dimensional medical images in the three-dimensional medical images based on a trained two-dimensional neural network model to obtain second cutting guide corresponding to each two-dimensional medical image; and determining the second cutting guide corresponding to the three-dimensional medical image based on the second cutting guide corresponding to each two-dimensional medical image.
According to a second aspect of the present application, a three-dimensional medical image cutting device includes: the first cutting guide determining module is used for acquiring similar cases corresponding to the target cases and cutting information corresponding to the similar cases, acquiring historical cutting information corresponding to historical three-dimensional medical images of the target cases and determining first cutting guide based on the similar cutting information and the historical cutting information; the second cutting guide determining module is used for processing the current three-dimensional medical image of the target case based on a pre-trained neural network to obtain a second cutting guide, wherein the neural network is obtained by training based on a large number of different types of medical data, and the training is performed on the different types of medical data based on different network models and training parameters; and the cutting module is used for respectively cutting the current three-dimensional medical image of the target case along an orthogonal cutting direction or along a cutting direction of any angle based on the first cutting guide and the second cutting guide to obtain a cutting plan with medical diagnosis significance.
An electronic device according to an embodiment of the third aspect of the present application includes a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements a method for cutting a three-dimensional medical image as described above when the computer program is executed by the processor.
A non-transitory computer readable storage medium according to an embodiment of the fourth aspect of the present application, on which a computer program is stored, which when executed by a processor implements a method of cutting a three-dimensional medical image as described above.
A computer program product according to an embodiment of the fifth aspect of the present application comprises a computer program which, when executed by a processor, implements a method for cutting a three-dimensional medical image as described above.
The above technical solutions in the embodiments of the present application have at least one of the following technical effects:
the method comprises the steps of determining a first cutting guide based on medical data of a case similar to a target case and historical medical data of a target patient, determining a second cutting guide based on a network model obtained by big data training, and finally cutting a three-dimensional medical image based on the first cutting guide and the second cutting guide, so that historical medical information is fully utilized, current medical cutting is guided through historical experience, accuracy and efficiency of cutting the medical image are improved, and medical diagnosis is assisted. Further, the method and the device also comprise the step of cutting the three-dimensional medical image based on personal interests of doctors or diagnosis customization, so that the degree of freedom of cutting is further improved.
Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a method for cutting a three-dimensional medical image according to one embodiment of the present disclosure;
FIG. 2 is a flowchart of a process including an image recognition module according to an embodiment of the present application;
FIG. 3 is a schematic view of a cutting interface provided by an embodiment of the present application;
fig. 4 is a schematic diagram of a three-dimensional medical image cutting device according to an embodiment of the present disclosure;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
Embodiments of the present application are described in further detail below with reference to the accompanying drawings and examples. The following examples are illustrative of the present application but are not intended to limit the scope of the present application.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the embodiments of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
The execution main body of the three-dimensional medical image cutting method can be a central server or can also be a terminal of a user, including but not limited to a mobile phone, a tablet computer, a pc terminal and the like.
In some embodiments, as shown in fig. 1, the method for cutting a three-dimensional medical image includes:
step 110, obtaining similar cases corresponding to the target cases and similar cutting information corresponding to similar medical images of the similar cases.
Similar cases refer to cases that have the same or similar condition as the current target case. The diagnosis guidance can be carried out on the illness state of the current user based on the history diagnosis information of the similar cases, so that more experience data of the current doctor can be provided, the accuracy of medical diagnosis is improved, and the medical diagnosis is better carried out.
The medical database stores a large number of medical data of patients, and the medical data includes case information, two-dimensional medical image information, and three-dimensional medical image information corresponding to each patient. Specifically, the case information includes the condition of the patient and the diagnosis information of the doctor on the patient. The two-dimensional medical image information comprises a body part corresponding to the two-dimensional medical image, acquisition parameters corresponding to the two-dimensional medical image, a patient position on the two-dimensional medical image and cutting information aiming at the two-dimensional medical image. The three-dimensional medical image information comprises a body part corresponding to the three-dimensional medical image, acquisition parameters corresponding to the two-dimensional medical image for synthesizing the three-dimensional medical image, a patient position on the three-dimensional medical image and cutting information aiming at the three-dimensional medical image. And the two-dimensional medical image corresponding to the section with medical diagnosis function in the three-dimensional medical image.
And 120, acquiring historical cutting information corresponding to the historical three-dimensional medical image of the target case.
Each patient corresponds to medical data of a plurality of disease stages, the medical data of different disease stages corresponds to different patient time, and the medical data also comprises corresponding information of each patient in different patient time stages, so that diagnosis information of different time periods can be obtained based on case information corresponding to the patient in different disease time periods, and further, optimal diagnosis guidance can be given to the current target case based on the diagnosis information of different time periods. And the disease development of the target case can be predicted through case information of a plurality of stages of the historical case, so that the current diagnosis of the current target case can be effectively guided based on the historical diagnosis and treatment data of the historical case.
For example, a lung image of a current patient is acquired, medical data of the corresponding patient whose lung image is acquired is matched in a history database, similar medical data similar to the acquired lung image of the current patient is matched in the medical data, history diagnosis information of the similar medical data is acquired, for example, history cutting information for the similar medical data is acquired, and then cutting guidance of the current patient is given.
When the medical data of a plurality of stages is included in the matched historical medical data, the most similar medical image is further matched in the medical data of the plurality of stages, and diagnosis information, such as cutting information, for the medical image is acquired. In some embodiments, the matched diagnosis information or cutting information can be output and displayed. Or, the coordinate information corresponding to the cutting information can be output, so that the doctor can adaptively modify the coordinate information based on the output coordinate information to obtain the cutting information for the current patient.
Historical medical diagnostic information for the patient may also be obtained, such that current diagnosis may also be directed based on the historical diagnostic information for the patient. Because the current symptoms of the same patient are closely related to the historical symptoms of the same patient, the historical illness state data of the target patient can be timely obtained by calling the historical symptom information of the target patient, so that the current diagnosis is guided, and the accuracy and the effectiveness of medical diagnosis are improved.
And 130, determining a first cutting guide based on the similar cutting information and the historical cutting information.
In some embodiments, the acquiring similar cases corresponding to the target case and the cutting information corresponding to the similar cases, acquiring historical cutting information corresponding to the historical three-dimensional medical image of the target case, and determining the first cutting guide based on the similar cutting information and the historical cutting information, includes: acquiring medical data of the target case, wherein the medical data comprises case information, historical three-dimensional medical images and current three-dimensional medical images of the target case; acquiring similar medical data similar to the medical data of the target case from a medical database; and obtaining a first cutting guide based on the historical cutting information of the similar medical data and the historical cutting information of the target case.
In this way, based on the acquired medical data of the similar case and the historical medical diagnosis data of the patient, even the stage medical data of the similar case, which is matched with the current state of illness of the current patient, the first cutting guide is determined by integrating the related medical data under various conditions, so that the most effective and most accurate diagnosis information can be acquired to the greatest extent when the current patient is subjected to cutting diagnosis, and the accuracy and the effectiveness of the medical diagnosis are improved.
In some embodiments, the method for cutting a three-dimensional medical image includes a module for image type identification.
As shown in fig. 2, a process flow diagram including an image recognition module is provided in some embodiments.
Step 210, the image type recognition module is configured to detect and recognize a three-dimensional medical image of the target case, so as to obtain an image type corresponding to the three-dimensional medical image.
In the method for cutting and processing the three-dimensional medical image, the image type recognition module is included, whether a similar case similar to the illness state of a target case exists or not is recognized through the image type recognition module, or whether a historical three-dimensional medical image exists in the target case or not is recognized through the image type recognition module, so that the processing can be performed through the image type recognition module without performing multiple processing through two or more models, the redundancy of data processing is reduced, the processing can be performed through one recognition model, the occupied resources of a computer are reduced, and the processing efficiency of the model is improved.
And 220, judging whether similar cases corresponding to the target cases exist or not based on the identified image types.
And step 221, when a similar case corresponding to the target case exists, starting a neural network trained based on case information corresponding to the similar case to process the three-dimensional medical image of the target case, and obtaining a similar cutting guide.
And step 222, closing the step of processing the three-dimensional medical image of the target case by the neural network trained based on the case information corresponding to the similar case when the similar case corresponding to the target case does not exist.
Step 230, judging whether the historical three-dimensional medical image exists in the target case or not based on the identified image type.
And step 231, when the historical cutting information corresponding to the target case exists, starting a neural network trained based on the cutting information of different illness states of the same case to process the three-dimensional medical image of the target case, and obtaining the historical cutting guide.
And step 232, closing the neural network trained based on the cutting information of different disease stages of the same case to process the three-dimensional medical image of the target case when the historical cutting information corresponding to the target case does not exist.
In the embodiment, the method can be realized through one identification model, thereby saving computer resources, being beneficial to side end deployment and realizing end-to-end processing.
In some embodiments, the method further comprises: judging whether similar cases corresponding to the target cases exist or not, and when the similar cases corresponding to the target cases exist, processing the three-dimensional medical images of the target cases by using a neural network trained based on case information corresponding to the similar cases to obtain similar cutting guide, wherein the similar cutting guide comprises similar cutting information.
In some embodiments, the method further comprises: judging whether a historical three-dimensional medical image exists in the target case, and when the historical three-dimensional medical image corresponding to the target case exists, processing the three-dimensional medical image of the target case by using a neural network trained based on cutting information of different disease stages of the same case to obtain historical cutting guide, wherein the historical cutting guide comprises historical cutting information.
And determining a first cutting guide based on the similar cutting information and the historical cutting information.
And 140, processing the current three-dimensional medical image of the target case based on the pre-trained neural network to obtain a second cutting guide. The neural network is trained based on a large number of different types of medical data, and is trained based on different network models and training parameters for the different types of medical data.
In one embodiment, two-dimensional medical images corresponding to different parts of a human body are acquired, three-dimensional medical images are generated based on the two-dimensional medical images, and a network model is trained based on the three-dimensional medical images at the different parts and segmentation information for the three-dimensional medical images so as to obtain a trained neural network for the different parts of the human body. In another embodiment, two-dimensional medical images of different acquisition parameters for the same body part are acquired, three-dimensional medical images are generated based on the two-dimensional medical images, and neural networks for the different acquisition parameters are generated based on the same body part training.
In this way, the neural network model of the corresponding type can be matched based on the body part corresponding to the medical data of the current patient and the acquisition parameters, and then the cutting guide information for the medical data is determined based on the matched neural network model.
The corresponding neural network models are respectively trained through at least two dimensions, namely the body part dimension and the acquisition parameter dimension, so that each type of neural network model can be accurately adapted to the data processing of the corresponding type of three-dimensional medical image, and the accuracy of the prediction of the segmentation part in the three-dimensional medical image can be further improved.
As shown in fig. 2, the device comprises different parts such as limbs, brain, head, heart and the like, and each part corresponds to a plurality of medical images. In fig. 2, when the head is selected, a medical image corresponding to the head is displayed, and any one of the medical images can be selected for cutting.
It can be understood that the cutting information of the three-dimensional medical image can be information saved by a doctor in the actual operation process of the history, or can be manually indexed for acquiring the labeling data. Therefore, the network model is trained based on the historical medical big data, so that a model with a prediction function is obtained, and a second cutting guide can be given based on the prediction function of the model when a doctor performs medical diagnosis, so that the accuracy of the doctor in performing medical diagnosis is further improved. And the function can be used in teaching assistance, so that more cutting prompts and directions can be given to an operator to acquire historical related medical data, which is a great progress in teaching assistance.
The first cutting guide and the second cutting guide are determined based on historical medical data, historical medical information is fully utilized, and current medical cutting is guided through historical experience. Also included in this application are custom cutting three-dimensional medical images based on the personal interests or diagnosis of the physician.
For example, after receiving the first cutting guide and the second cutting guide, the doctor may also perform custom cutting on the three-dimensional medical image based on the received cutting guide and the personal diagnosis, so as to obtain a cutting plan with a diagnosis section.
In some embodiments, the method further includes automatically identifying a lesion area in the medical image and providing a cutting guide for the lesion information, so that the user can cut the medical image based on the cutting guide, and the user can directly confirm the cutting guide to automatically obtain a cut image.
In some embodiments, the three-dimensional medical image may be identified and predicted based on a pre-trained neural network, e.g., neural network model training may be performed directly on the three-dimensional medical image to obtain a trained network model that may be used to identify the lesion area.
The three-dimensional medical image is a three-dimensional image, and model training can be performed based on the three-dimensional image to obtain a trained network model. However, the network model obtained based on the three-dimensional medical image training needs to consume a large amount of computer resources, and on the other hand, has the problem of inaccurate training.
In some embodiments, the three-dimensional medical image is constructed based on two-dimensional medical image slice data; the processing the current three-dimensional medical image of the target case based on the pre-trained neural network to obtain a second cutting guide comprises: performing model training based on the acquired two-dimensional medical image and the history cutting information corresponding to the two-dimensional medical image to obtain a two-dimensional neural network model; acquiring the three-dimensional medical images, and processing the two-dimensional medical images in the three-dimensional medical images based on a trained two-dimensional neural network model to obtain second cutting guide corresponding to each two-dimensional medical image; and determining the second cutting guide corresponding to the three-dimensional medical image based on the second cutting guide corresponding to each two-dimensional medical image.
For example, the two-dimensional neural network is used for processing one or more two-dimensional medical images in the three-dimensional medical images to obtain second cutting directions for the one or more two-dimensional medical images, so that the cutting directions corresponding to the three-dimensional medical images can be obtained based on the second cutting directions of all the two-dimensional medical images.
In the training process of the two-dimensional neural network, the two-dimensional medical image model is utilized for training, and only two-dimensional labeling is needed at the moment, and three-dimensional labeling is not needed, so that the labeling cost is greatly reduced.
In the above embodiment, the three-dimensional medical image is formed by stacking a plurality of two-dimensional medical images, so that the model training can be performed based on the two-dimensional images to obtain the two-dimensional recognition model, and thus, when the focus of the three-dimensional medical image is recognized, the recognition result of the three-dimensional medical image can be obtained based on the recognition of a plurality of two-dimensional medical images in the three-dimensional medical image. Therefore, by training the two-dimensional network model, the dimension of training data is lower, training parameters are fewer, training efficiency and accuracy are greatly improved, and meanwhile, prediction of the three-dimensional medical image is realized.
In addition, in the process of identifying the two-dimensional medical images, a large amount of medical information of the two-dimensional medical images can be obtained, so that the two-dimensional medical images containing the most or most key medical information can be extracted from the large amount of two-dimensional medical images, and the two-dimensional medical images containing more information can be sent to a user terminal for display, so that more medical instructions are given to the user. Therefore, in the process of processing the three-dimensional medical image, the medical experience of a professional doctor and the observation capability of the professional doctor are not only relied on, but also data analysis can be carried out based on big data or more scientific means, so that more medical information can be mined from the three-dimensional medical image, medical diagnosis is assisted by means of the scientific means, and the accuracy and the efficiency of medical diagnosis can be further improved.
And when the medical image is processed based on the two-dimensional network model obtained through training, focus information in the two-dimensional medical image can be marked, so that marking information of the three-dimensional medical image can be obtained by marking information of each two-dimensional medical image, and when the three-dimensional medical image is cut, medical staff can be assisted in diagnosis based on identification marking information of the two-dimensional medical image corresponding to the cut section.
Considering that the three-dimensional medical image is obtained based on a plurality of two-dimensional medical images, the two-dimensional medical images can be obtained by medical equipment, when the two-dimensional network model is trained, the two-dimensional medical images acquired by the medical equipment can be used for training, and when the network model is used, the two-dimensional medical images acquired by the medical equipment are processed. And when the user cuts the three-dimensional medical image, the network model can be displayed in the cut section to process the corresponding medical information of the two-dimensional medical image corresponding to the section. In this way, big data calculation processing is introduced in the medical diagnosis process, so that doctors are assisted in medical diagnosis, more medical guidance is provided for the doctors, and the accuracy and efficiency of medical diagnosis are improved.
In addition, because some focuses are typical or damage to patients is relatively large, in some embodiments, model training can be performed based on some focuses to obtain a trained network model, when the trained network processes three-dimensional medical images, the focuses can be directly identified, a professional doctor is prompted and guided at the first time, and even automatic cutting is performed directly based on the identification result, so that the cutting efficiency and accuracy are further improved.
In other embodiments, predictive models may also be trained based on historical case information for different stages of a condition. For example, the training of the prediction model can be performed based on the case information of the same user in different disease states or the case information of different users of the same case in different disease states, so that when the three-dimensional medical model of the current user is processed by using the prediction model, the disease state of the current patient can be predicted, and the disease state development trend of the user can be timely given to a professional doctor, so that the medical diagnosis can be timely performed.
In some embodiments, the three-dimensional medical image is trained based on slice data corresponding to a two-dimensional medical image; the training step of the neural network comprises the following steps: acquiring three-dimensional medical images corresponding to different parts of a human body, three-dimensional medical images corresponding to different sampling parameters at the same part, and historical cutting information corresponding to each three-dimensional medical image; training to obtain neural networks corresponding to different parts of the human body based on the three-dimensional medical images and the historical cutting information corresponding to the three-dimensional medical images, and selecting the neural networks of the corresponding types to process when the three-dimensional medical images corresponding to the different parts of the human body are required to be cut so as to obtain corresponding second cutting guide.
In some embodiments, the method further comprises the step of magnifying the cutting image obtained by cutting based on the cutting guide, so that a clearer cutting image can be obtained.
And step 150, based on the first cutting guide and the second cutting guide, respectively cutting the current three-dimensional medical image of the target case along an orthogonal cutting direction or along a cutting direction of any angle, so as to obtain a cutting plan with medical diagnosis significance.
In some embodiments, the cutting the current three-dimensional medical image of the target case along the orthogonal cutting direction based on the first cutting guide and the second cutting guide to obtain a cutting plan having medical diagnostic significance includes: and determining a reference point in the current three-dimensional medical image based on the first cutting guide and the second cutting guide, and determining a plane perpendicular to a three-dimensional coordinate axis based on the reference point to form three mutually orthogonal cross-sectional images with diagnostic significance, wherein the cross-sectional images at least comprise a cross section, a coronal plane and a sagittal plane.
The orthogonal cutting direction is the direction along the X, Y, Z axis of the Cartesian coordinate system, a point is found to be a reference point, and a plane perpendicular to the three-dimensional coordinate axis is determined to cut the three-dimensional medical image, so that a two-dimensional cross-sectional image with medical diagnosis information can be obtained, and three cross-sectional images which are mutually orthogonal in the cross-sectional image can be obtained. The sectional image includes at least a cross section, a coronal plane, and a sagittal plane. Further, the doctor can also randomly extract and display the orthogonal faces based on the actual requirements.
In another embodiment, it is also possible to extract and display the image information of the cross section at any angle and orientation, not just at the orthogonal plane. For example, a plane which does not limit the orientation is determined in the three-dimensional medical image, and a normal vector and an internal point which are related to the plane are further determined, so that the three-dimensional medical image can be cut in any orientation.
In some embodiments, the cutting the current three-dimensional medical image of the target case along the cutting direction of any angle based on the first cutting guide and the second cutting guide to obtain a cutting plan having medical diagnostic significance includes: determining a plane with any azimuth in the current three-dimensional medical image based on the first cutting guide and the second cutting guide, determining a normal vector and an interior point corresponding to the plane, and performing plane cutting at any angle on the three-dimensional medical image based on the normal vector and the interior point to obtain image information of a cutting plane with diagnostic significance, wherein the coordinates of the interior point are (x 0 ，y 0 ，z 0 ) The normal vector of the plane is n (a 1 ，a 2 ，a 3 ) Satisfy a 1 (x-x 0 )+a 2 (y-y 0 )+a 3 (z-z 0 )。
In some embodiments, the normal vector is determined based on the direction angle of the plane, the direction angle including α, β, γ, and 0.ltoreq.α.ltoreq.pi, and 0.ltoreq.β.ltoreq.pi, and 0.ltoreq.γ.ltoreq.pi, if the three-dimensional origin of coordinates is O, the directional cosine of vector OP is cos α, cos β, cos γ, normal vector n= (a) 1 ，a 2 ，a 3 ) Then
In some embodiments, to obtain other cuts along the normal vectorIt is also possible to move the initially defined plane along the normal vector, for example by a distance b. The method further comprises the steps of: moving the initial plane along the direction of the normal vector, moving the initial plane by a preset moving distance b to obtain a moved cutting plane, and translating m '(x) corresponding to the distance b for any point m (x', y ', z') 0 ，y 0 ，z 0 ) X ' =x+b×cos α, y ' =y+b×cos β, z ' =z+b×cos γ.
Furthermore, the doctor or the learner operator can perform man-machine interaction operation based on the requirement, and can also change the normal vector and the interior point position in the defined plane to cut the three-dimensional medical image in real time and display the three-dimensional medical image in real time.
In some embodiments, the method may further include first preprocessing the obtained three-dimensional medical image, for example, denoising, sharpening, or filling the three-dimensional medical image, and then performing subsequent processing on the processed three-dimensional medical image. Like this, through carrying out preliminary treatment to three-dimensional medical image for follow-up can acquire more accurate case information, make follow-up processing more accurate.
As shown in fig. 4, the present application provides a three-dimensional medical image cutting device, including:
a first cutting guide determining module 310, configured to obtain a similar case corresponding to a target case and cutting information corresponding to the similar case, obtain historical cutting information corresponding to a historical three-dimensional medical image of the target case, and determine a first cutting guide based on the similar cutting information and the historical cutting information;
the second cutting guide determining module 320 is configured to process the current three-dimensional medical image of the target case based on a pre-trained neural network, so as to obtain a second cutting guide, where the neural network is obtained by training based on a large number of different types of medical data, and training is performed on different types of medical data based on different network models and training parameters;
and the cutting module 330 is configured to cut the current three-dimensional medical image of the target case along an orthogonal cutting direction or along a cutting direction with any angle based on the first cutting guide and the second cutting guide, so as to obtain a cutting plan with medical diagnosis significance.
In some embodiments, the cutting module 330 is further to:
Determining a reference point in the current three-dimensional medical image based on the first cutting guide and the second cutting guide, determining a plane perpendicular to a three-dimensional coordinate axis based on the reference point, and forming three mutually orthogonal cross-sectional images with diagnostic significance, wherein the cross-sectional images at least comprise a cross section, a coronal plane and a sagittal plane; or alternatively, the process may be performed,
determining a plane with any azimuth in the current three-dimensional medical image based on the first cutting guide and the second cutting guide, determining a normal vector and an interior point corresponding to the plane, and performing plane cutting at any angle on the three-dimensional medical image based on the normal vector and the interior point to obtain image information of a cutting plane with diagnostic significance, wherein the coordinates of the interior point are (x 0 ，y 0 ，z 0 ) The normal vector of the plane is n (a 1 ，a 2 ，a 3 ) Satisfy a 1 (x-x 0 )+a 2 (y-y 0 )+a 3 (z-z 0 )。
In some embodiments, the normal vector in the cutting device of the three-dimensional medical image is determined based on the direction angle of the plane, the direction angle comprises alpha, beta, gamma, and 0.ltoreq.alpha.ltoreq.pi, 0.ltoreq.beta.ltoreq.pi, and 0.ltoreq.gamma.ltoreq.pi, if the three-dimensional origin of coordinates is O, the direction cosine of the vector OP is cos alpha, cos beta, cos gamma, and the normal vector n= (a) 1 ，a 2 ，a 3 ) Then
In some embodiments, the three-dimensional medical image cutting device is further configured to:
moving the initial plane along the direction of the normal vector, moving the initial plane by a preset moving distance b to obtain a moved cutting plane, and translating m '(x) corresponding to the distance b for any point m (x', y ', z') 0 ，y 0 ，z 0 ) X ' =x+b×cos α, y ' =y+b×cos β, z ' =z+b×cos γ.
In some embodiments, the first cutting direction determination module 310 is further to:
acquiring medical data of the target case, wherein the medical data comprises case information, historical three-dimensional medical images and current three-dimensional medical images of the target case;
acquiring similar medical data similar to the medical data of the target case from a medical database;
and obtaining a first cutting guide based on the historical cutting information of the similar medical data and the historical cutting information of the target case.
In some embodiments, the three-dimensional medical image cutting device is further configured to:
the three-dimensional medical image is constructed based on two-dimensional medical image slice data; the training step of the neural network comprises the following steps:
acquiring three-dimensional medical images corresponding to different parts of a human body, three-dimensional medical images corresponding to different sampling parameters at the same part, and historical cutting information corresponding to each three-dimensional medical image;
Training to obtain neural networks corresponding to different parts of the human body based on the three-dimensional medical images and the historical cutting information corresponding to the three-dimensional medical images, and selecting the neural networks of the corresponding types to process when the three-dimensional medical images corresponding to the different parts of the human body are required to be cut so as to obtain corresponding second cutting guide.
In some embodiments, the cutting area may be highlighted, for example, different color display may be performed, so that the important attention area may be quickly acquired.
In some embodiments, the region of interest of the user and the automatically identified region may be comprehensively processed in the cutting process to obtain the cut region, for example, when the identified cutting direction of the user is consistent or substantially consistent with the direction obtained after the model identification, the cutting direction of the user may be guided based on the identified information, so that the problem of inaccurate cutting of the user may be prevented, and further more accurate cutting information may be obtained. Of course, the user may reject automatically identified cutting information, and thus, active cutting processing may be performed directly based on the user's region of interest.
It can be appreciated that there are multiple functional modules on the interface at the medical image, so that the user can send corresponding processing instructions based on the functional modules to obtain more accurate cutting information.
Because the three-dimensional space regular volume data field has a certain layering sense, in actual observation, internal tissues and organs of the three-dimensional visualized image are usually shielded by external information such as skin, bones and the like, and the tissue structure and related information in the three-dimensional reconstruction body cannot be seen.
Compared with the traditional method for displaying the internal information by simply adjusting the opacity, gray scale and other numerical values of the external surfaces such as skin, bones and the like, the method for cutting the medical image can cut the image information in the three-dimensional reconstruction body, and has stronger instantaneity and interactivity.
Fig. 5 illustrates a physical schematic diagram of an electronic device, as shown in fig. 4, which may include: processor 410, communication interface (Communications Interface) 420, memory 430 and communication bus 440, wherein processor 410, communication interface 420 and memory 430 communicate with each other via communication bus 440. The processor 410 may invoke logic instructions in the memory 430 to perform a method of cutting a three-dimensional medical image, comprising: acquiring similar cases corresponding to a target case and cutting information corresponding to the similar cases, acquiring historical cutting information corresponding to historical three-dimensional medical images of the target case, and determining a first cutting guide based on the similar cutting information and the historical cutting information; processing the current three-dimensional medical image of the target case based on a pre-trained neural network to obtain a second cutting guide, wherein the neural network is obtained based on a large number of different types of medical data training, and training is performed on different types of medical data based on different network models and training parameters; and cutting the current three-dimensional medical image of the target case along an orthogonal cutting direction or along a cutting direction of any angle based on the first cutting guide and the second cutting guide respectively to obtain a cutting plan with medical diagnosis significance.
Further, the logic instructions in the memory 430 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Further, the present application also provides a computer program product, where the computer program product includes a computer program, where the computer program may be stored on a non-transitory computer readable storage medium, and when the computer program is executed by a processor, the computer is capable of executing a method for cutting a three-dimensional medical image provided by the foregoing method embodiments, where the method includes: acquiring similar cases corresponding to a target case and cutting information corresponding to the similar cases, acquiring historical cutting information corresponding to historical three-dimensional medical images of the target case, and determining a first cutting guide based on the similar cutting information and the historical cutting information; processing the current three-dimensional medical image of the target case based on a pre-trained neural network to obtain a second cutting guide, wherein the neural network is obtained based on a large number of different types of medical data training, and training is performed on different types of medical data based on different network models and training parameters; and cutting the current three-dimensional medical image of the target case along an orthogonal cutting direction or along a cutting direction of any angle based on the first cutting guide and the second cutting guide respectively to obtain a cutting plan with medical diagnosis significance.
In another aspect, embodiments of the present application further provide a non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor, is implemented to perform a method for cutting a three-dimensional medical image provided in the above embodiments, including: acquiring similar cases corresponding to a target case and cutting information corresponding to the similar cases, acquiring historical cutting information corresponding to historical three-dimensional medical images of the target case, and determining a first cutting guide based on the similar cutting information and the historical cutting information; processing the current three-dimensional medical image of the target case based on a pre-trained neural network to obtain a second cutting guide, wherein the neural network is obtained based on a large number of different types of medical data training, and training is performed on different types of medical data based on different network models and training parameters; and cutting the current three-dimensional medical image of the target case along an orthogonal cutting direction or along a cutting direction of any angle based on the first cutting guide and the second cutting guide respectively to obtain a cutting plan with medical diagnosis significance.
The apparatus embodiments described above are merely illustrative, wherein elements illustrated as separate elements may or may not be physically separate, and elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solutions may be embodied essentially or in part in the form of a software product, which may be stored in a computer-readable storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the various embodiments or methods of some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting thereof; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.
The above embodiments are merely illustrative of the present application and are not limiting thereof. While the present application has been described in detail with reference to the embodiments, those skilled in the art will understand that various combinations, modifications, or equivalents of the technical solutions of the present application may be made without departing from the spirit and scope of the technical solutions of the present application, and all such modifications are intended to be covered by the claims of the present application.

Claims (5)

1. A method for cutting a three-dimensional medical image, comprising: the image type identification module is used for detecting and identifying the three-dimensional medical image of the target case to obtain an image type corresponding to the three-dimensional medical image;
When judging that similar cases corresponding to the target cases exist based on the image types recognized by the image type recognition module, acquiring the similar cases corresponding to the target cases, starting a neural network obtained by training based on case information corresponding to the similar cases to process three-dimensional medical images of the target cases, obtaining similar cutting information corresponding to the similar medical images of the similar cases, and when the similar cases corresponding to the target cases do not exist, closing the neural network obtained by training based on the case information corresponding to the similar cases to process the three-dimensional medical images of the target cases;
when judging that the historical cutting information corresponding to the target case exists based on the image type recognized by the image recognition module, starting a neural network obtained by training cutting information of different illness states of the same case to process the three-dimensional medical image of the target case, and obtaining the historical cutting information corresponding to the historical three-dimensional medical image of the target case so as to obtain the illness state development trend of the current target case;
when the historical cutting information corresponding to the target case does not exist, closing a neural network obtained by training the cutting information of different disease stages of the same case to process the three-dimensional medical image of the target case; a first cutting guideline determined based on the similar cutting information and the historical cutting information;
Processing the current three-dimensional medical image of the target case based on a pre-trained neural network to obtain a second cutting guide, wherein the neural network is obtained based on different types of medical data training, and the neural network is trained based on different network models and training parameters aiming at different types of medical data; the neural network for obtaining the second cutting guide is determined based on a two-dimensional neural network model obtained by training a two-dimensional medical image corresponding to the three-dimensional medical image, and the two-dimensional neural network is obtained by training based on two-dimensional labeling data; the training method comprises the steps of obtaining two-dimensional medical images corresponding to different parts of a human body based on different network models and training parameters aiming at different types of medical data, generating three-dimensional medical images based on the two-dimensional medical images, training the network models based on the three-dimensional medical images at the different parts and segmentation information aiming at the three-dimensional medical images to obtain trained neural networks aiming at different body parts, obtaining two-dimensional medical images aiming at different acquisition parameters of the same body part, generating the three-dimensional medical images based on the two-dimensional medical images, and training to generate the neural networks aiming at different acquisition parameters based on the same body part; after the three-dimensional medical image is obtained, processing the two-dimensional medical images in the three-dimensional medical image based on a trained two-dimensional neural network model to obtain second cutting directions corresponding to each two-dimensional medical image, and determining the second cutting directions corresponding to the three-dimensional medical image based on the second cutting directions corresponding to each two-dimensional medical image;
Cutting the current three-dimensional medical image of the target case along a cutting direction of any angle based on the first cutting guide and the second cutting guide to obtain a cutting plane graph with medical diagnosis significance, wherein a plane of any azimuth is determined in the current three-dimensional medical image based on the first cutting guide and the second cutting guide, a normal vector and an inner point corresponding to the plane are determined, and the plane of any angle is cut on the three-dimensional medical image based on the normal vector and the inner point to obtain image information of a cutting plane with diagnosis significance, wherein coordinates of the inner point are (x) 0 ，y 0 ，z 0 ) The normal vector of the plane is n (a 1 ，a 2 ，a 3 ) The normal vector is determined based on the direction angle of the plane, the direction angle comprises alpha, beta and gamma, alpha is more than or equal to 0 and less than or equal to pi, beta is more than or equal to 0 and less than or equal to pi, gamma is more than or equal to 0 and less than or equal to pi, if the three-dimensional origin of coordinates is O, the cosine of the direction of the vector OP is cos alpha, cos beta and cos gamma, and the normal vector n= (a) 1 ，a 2 ，a 3 ) Then Wherein P is any point in the plane.
2. The method according to claim 1, wherein the method further comprises:
moving the initial plane along the direction of the normal vector, moving the initial plane by a preset moving distance b to obtain a moved cutting plane, and translating the corresponding m '(x) after the distance b for any point m (x', y ', z') 0 ，y 0 ，z 0 ) X is then 0 ＝x′+b*cosα，y 0 ＝y′+b*cosβ，z 0 ＝z′+b*cosγ。
3. The method of claim 1, wherein obtaining a similar case corresponding to a target case and cutting information corresponding to the similar case, obtaining historical cutting information corresponding to a historical three-dimensional medical image of the target case, determining a first cutting guideline based on the similar cutting information and the historical cutting information, comprises:
acquiring medical data of the target case, wherein the medical data comprises case information, historical three-dimensional medical images and current three-dimensional medical images of the target case;
acquiring similar medical data similar to the medical data of the target case from a medical database;
and obtaining a first cutting guide based on the historical cutting information of the similar medical data and the historical cutting information of the target case.
4. The method of claim 1, wherein the three-dimensional medical image is constructed based on two-dimensional medical image slice data; the training step of the neural network comprises the following steps:
acquiring three-dimensional medical images corresponding to different parts of a human body, three-dimensional medical images corresponding to different sampling parameters at the same part, and historical cutting information corresponding to each three-dimensional medical image;
Training to obtain neural networks corresponding to different parts of the human body based on the three-dimensional medical images and the historical cutting information corresponding to the three-dimensional medical images, and selecting the neural networks of the corresponding types to process when the three-dimensional medical images corresponding to the different parts of the human body are required to be cut so as to obtain corresponding second cutting guide.
5. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the method of cutting a three-dimensional medical image according to any one of claims 1 to 4.
CN202210896276.7A 2022-07-28 2022-07-28 Three-dimensional medical image cutting method, electronic device and storage medium Active CN115100189B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210896276.7A CN115100189B (en) 2022-07-28 2022-07-28 Three-dimensional medical image cutting method, electronic device and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210896276.7A CN115100189B (en) 2022-07-28 2022-07-28 Three-dimensional medical image cutting method, electronic device and storage medium

Publications (2)

Publication Number Publication Date
CN115100189A CN115100189A (en) 2022-09-23
CN115100189B true CN115100189B (en) 2023-07-18

Family

ID=83299964

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210896276.7A Active CN115100189B (en) 2022-07-28 2022-07-28 Three-dimensional medical image cutting method, electronic device and storage medium

Country Status (1)

Country Link
CN (1) CN115100189B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021151338A1 (en) * 2020-09-22 2021-08-05 平安科技（深圳）有限公司 Medical imagery analysis method, apparatus, electronic device and readable storage medium
WO2022037696A1 (en) * 2020-08-21 2022-02-24 张逸凌 Bone segmentation method and system based on deep learning

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105354846A (en) * 2015-11-05 2016-02-24 沈阳东软医疗***有限公司 Method and apparatus for segmenting three-dimensional medical image
CN108417272B (en) * 2018-02-08 2020-10-09 合肥工业大学 Similar case recommendation method and device with time sequence constraint
CN113192588A (en) * 2021-04-29 2021-07-30 广东粤信智能科技有限公司 Diagnosis and treatment assisting method based on data right determination, storage medium and system
CN113362327B (en) * 2021-08-10 2021-11-05 杭州晟视科技有限公司 Region segmentation method, device, terminal and storage medium in chest image

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022037696A1 (en) * 2020-08-21 2022-02-24 张逸凌 Bone segmentation method and system based on deep learning
WO2021151338A1 (en) * 2020-09-22 2021-08-05 平安科技（深圳）有限公司 Medical imagery analysis method, apparatus, electronic device and readable storage medium

Also Published As

Publication number Publication date
CN115100189A (en) 2022-09-23

Similar Documents

Publication Publication Date Title
JP6947759B2 (en) Systems and methods for automatically detecting, locating, and semantic segmenting anatomical objects
KR102013857B1 (en) Method and apparatus for generating learning data based on surgical video
CN110652317B (en) Automatic positioning method for standard tangent plane in prenatal fetal ultrasound volume image
CN112509119B (en) Spatial data processing and positioning method and device for temporal bone and electronic equipment
Nurmaini et al. Accurate detection of septal defects with fetal ultrasonography images using deep learning-based multiclass instance segmentation
CN109712163B (en) Coronary artery extraction method, device, image processing workstation and readable storage medium
JP2022527525A (en) Identification of lesion boundaries in image data
CN115994902A (en) Medical image analysis method, electronic device and storage medium
AU2020340234A1 (en) System and method for identification, labeling, and tracking of a medical instrument
CN111667474A (en) Fracture identification method, apparatus, device and computer readable storage medium
EP2525323A1 (en) Visualizing deformation fields for image transformations
EP3973508A1 (en) Sampling latent variables to generate multiple segmentations of an image
CN110490841B (en) Computer-aided image analysis method, computer device and storage medium
Bucioli et al. Holographic real time 3D heart visualization from coronary tomography for multi-place medical diagnostics
US9123163B2 (en) Medical image display apparatus, method and program
CN112465771B (en) Spine nuclear magnetic resonance image analysis method and device and computer equipment
CN115100189B (en) Three-dimensional medical image cutting method, electronic device and storage medium
CN113012118A (en) Image processing method and image processing apparatus
US20230237657A1 (en) Information processing device, information processing method, program, model generating method, and training data generating method
Ilyasova et al. Development of a Computer System for Automatically Generating a Laser Photocoagulation Plan to Improve the Retinal Coagulation Quality in the Treatment of Diabetic Retinopathy
CN113706644A (en) Image processing method, image processing apparatus, and storage medium
CN117058405B (en) Image-based emotion recognition method, system, storage medium and terminal
CN111476768B (en) Image registration method, image registration device, path planning method, path planning device, path planning system and medium
CN113743414B (en) Method, device and system for identifying focus based on semantic coding
CN116523857B (en) Hearing state prediction device and method based on diffusion tensor image

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