CN115131487A

CN115131487A - Medical image processing method, system, computer device and storage medium

Info

Publication number: CN115131487A
Application number: CN202210871038.0A
Authority: CN
Inventors: 不公告发明人
Original assignee: Suzhou Xiaowei Changxing Robot Co ltd
Current assignee: Suzhou Xiaowei Changxing Robot Co ltd
Priority date: 2022-07-22
Filing date: 2022-07-22
Publication date: 2022-09-30

Abstract

The application relates to a medical image processing method, a medical image processing system, a computer device and a storage medium. The method comprises the following steps: acquiring a first medical image of a target part, wherein the first medical image is a three-dimensional medical image, and determining at least three key points corresponding to the target part from the first medical image; the reference coordinate system can be determined based on at least three key points, and then an image coordinate system is constructed according to the shooting angle of the first medical image. Performing rotation calibration on the first medical image to enable the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target part to meet a preset condition; and generating a second medical image corresponding to the target part according to the calibrated first medical image. By adopting the method, the three-dimensional medical image can be converted into other medical images of different image types, so that the problem that some medical images are difficult to shoot under specific conditions is solved, and the applicability of the medical images is improved.

Description

Medical image processing method, system, computer device and storage medium

Technical Field

The present application relates to the field of image processing technologies, and in particular, to a medical image processing method, system, computer device, storage medium, and computer program product.

Background

In traditional surgery, medical staff are used to perform surgical planning by using traditional X-ray images, and the X-ray machine has the advantages of low cost, high speed, small radiation dose, capability of displaying characteristic pathological structures and the like. However, there are limitations to using conventional X-ray models, such as the inability to correct the position of the patient's anatomy in the anatomical image. Moreover, in the computer-aided navigation operation, the traditional X-ray machine is influenced by the acquisition mode, X-ray images under certain specific scenes cannot be acquired, the specific business requirements cannot be met, and the problem of low applicability exists.

Disclosure of Invention

In view of the above, it is necessary to provide a medical image processing method, a system, a computer device, a computer readable storage medium and a computer program product capable of improving applicability in view of the above technical problems.

In a first aspect, the present application provides a medical image processing method. The method comprises the following steps:

acquiring a first medical image of a target part, and determining at least three key points corresponding to the target part from the first medical image, wherein the first medical image is a three-dimensional medical image;

constructing an image coordinate system, and determining a reference coordinate system corresponding to the target part based on at least three key points;

performing rotation calibration on the first medical image to enable the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target part to meet a preset condition;

and generating a second medical image corresponding to the target part according to the calibrated first medical image, wherein the second medical image is a two-dimensional medical image.

In one embodiment, acquiring a first medical image of a target site includes:

acquiring a CT three-dimensional image of a target part;

according to the type of the part corresponding to the target part, adjusting the window width and the window level of the CT three-dimensional image;

resampling the CT three-dimensional image with the window width and the window level adjusted, and adjusting the resolution of the CT three-dimensional image according to a preset resolution;

carrying out image enhancement and normalization processing on the CT three-dimensional image with the resolution adjusted;

and adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain a first medical image.

In one embodiment, determining at least three keypoints corresponding to the target site from the first medical image comprises:

inputting the first medical image into the trained neural network model to obtain a thermodynamic diagram corresponding to the first medical image;

determining at least three key areas from the first medical image, and determining at least three key areas in the thermodynamic diagram according to the mapping relation between the first medical image and the thermodynamic diagram;

and respectively acquiring pixel points meeting preset brightness adjustment in each key area from the thermodynamic diagram as at least three key points.

In one embodiment, the performing rotation calibration on the first medical image so that the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target portion satisfies a preset condition includes:

extracting two-dimensional images of each layer of each direction surface in the first medical image;

determining a rotation parameter corresponding to at least one direction surface in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; at least one of the orientation planes is coronal, sagittal, or transverse;

rotating each layer of two-dimensional image of each direction surface according to the rotation parameters corresponding to each direction surface to obtain each layer of process two-dimensional image of each direction surface;

and obtaining a calibrated first medical image based on the two-dimensional image of each layer process of each direction surface.

In one embodiment, the second medical image includes an X-ray image, and the second medical image corresponding to the target portion is generated from the calibrated first medical image, including:

acquiring volume data corresponding to the calibrated first medical image according to the spatial information of the calibrated first medical image;

acquiring a ray source position and a ray direction corresponding to a target part, and determining a radiation field area corresponding to volume data according to the ray source position and the ray direction;

and generating an X-ray image corresponding to the radiation field region by a digital reconstruction radiographic mode.

In one embodiment, generating an X-ray image corresponding to the portal region by digitally reconstructing a radiographic image includes:

processing the volume data corresponding to the field area through a ray projection algorithm to generate an X-ray image corresponding to the field area;

or processing the volume data corresponding to the portal area through a superposition density projection algorithm to generate an X-ray image corresponding to the portal area;

or processing the volume data corresponding to the portal area through a light field reconstruction algorithm to generate an X-ray image corresponding to the portal area.

In one embodiment, generating a second medical image corresponding to the target region from the calibrated first medical image includes:

acquiring a prosthesis three-dimensional image corresponding to a target prosthesis, and adjusting the posture information of the prosthesis three-dimensional image;

fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fused medical image;

and generating a second medical image corresponding to the target part based on the prosthesis fusion medical image, wherein the second medical image is an X-ray image, and the second medical image comprises the X-ray image of the target prosthesis.

determining at least one target sub-part from the calibrated first medical image; the target sub-part is obtained by dividing the target part;

in the calibrated first medical image, adjusting the posture information of at least one target sub-part to obtain a posture adjustment medical image;

and adjusting the medical image according to the posture to generate a second medical image corresponding to the target part.

In a second aspect, the present application further provides a medical image processing system. The system comprises:

the system comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for acquiring a first medical image of a target part and determining at least three key points corresponding to the target part from the first medical image, and the first medical image is a three-dimensional medical image;

the positioning module is used for constructing an image coordinate system and determining a reference coordinate system corresponding to the target part based on at least three key points;

the conversion module is used for performing rotation calibration on the first medical image so that the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target part meets a preset condition;

and the generation module is used for generating a second medical image corresponding to the target part according to the calibrated first medical image, and the second medical image is a two-dimensional medical image.

In one embodiment, the obtaining module comprises:

the acquisition unit is used for acquiring a CT three-dimensional image of a target part;

the adjusting unit is used for adjusting the window width and the window level of the CT three-dimensional image according to the part type corresponding to the target part;

the resampling unit is used for resampling the CT three-dimensional image after the window width and the window level are adjusted, and adjusting the resolution of the CT three-dimensional image according to the preset resolution;

the enhancing unit is used for enhancing the image of the CT three-dimensional image with the adjusted resolution;

the normalization unit is used for performing normalization processing on the CT three-dimensional image after image enhancement;

and the self-adaptive unit is used for adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain a first medical image.

In one embodiment, the generating module is further configured to process the volume data corresponding to the portal area through a ray-casting algorithm to generate an X-ray image corresponding to the portal area; or the generating module is further used for processing the volume data corresponding to the portal area through a superposition density projection algorithm to generate an X-ray image corresponding to the portal area; or the generating module is further configured to process the volume data corresponding to the portal region through a light field reconstruction algorithm, and generate an X-ray image corresponding to the portal region.

In a third aspect, the present application also provides a computer device. The computer device comprises a memory storing a computer program and a processor implementing the following steps when executing the computer program:

In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of:

In a fifth aspect, the present application further provides a computer program product. The computer program product comprising a computer program which when executed by a processor performs the steps of:

The medical image processing method, the medical image processing system, the computer equipment, the storage medium and the computer program product are used for acquiring a first medical image of a target part, wherein the first medical image is a three-dimensional medical image, and at least three key points corresponding to the target part are determined from the first medical image; therefore, the reference coordinate system corresponding to the target part can be determined based on at least three key points, and then the image coordinate system is constructed according to the shooting angle of the first medical image. Performing rotation calibration on the first medical image to enable the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target part to meet a preset condition; thus, the shooting angle for generating the second medical image can be ensured to meet medical requirements. And generating a second medical image corresponding to the target part according to the calibrated first medical image, wherein the second medical image is a two-dimensional medical image. The three-dimensional medical image can be converted into other two-dimensional medical images, so that the problem that some medical images are difficult to shoot under specific conditions is solved, a doctor can observe special anatomical landmark points which cannot be observed in the three-dimensional image conveniently, and the applicability of the medical images is improved.

Drawings

FIG. 1 is a flow diagram illustrating a method for medical image processing according to an embodiment;

FIG. 2 is a schematic flow diagram illustrating data preprocessing in one embodiment;

FIG. 3 is a schematic diagram of the structure of the SHVNet neural network in one embodiment;

FIG. 4 is a schematic diagram of the structure of each Hourglass Vnet in the SHVNet according to an embodiment;

FIG. 5 is a schematic flow chart illustrating key point acquisition in one embodiment;

FIG. 6 is a diagram illustrating an exemplary key point application;

FIG. 7 is a schematic diagram of a coordinate system for rotational calibration in one embodiment;

FIG. 8 is a schematic diagram illustrating a rotation of a two-dimensional image of each orientation plane in one embodiment;

FIG. 9 is a schematic flow chart illustrating the generation of an X-ray image in one embodiment;

FIG. 10 is a schematic illustration of an embodiment for generating an X-ray image;

FIG. 11 is a schematic diagram of an embodiment of a ray casting algorithm;

FIG. 12 is a schematic diagram of the calculation of attenuation by the ray casting algorithm in one embodiment;

FIG. 13 is a schematic diagram of an overlay density projection algorithm in accordance with one embodiment;

FIG. 14 is a schematic diagram of a light field reconstruction algorithm in one embodiment;

FIG. 15 is a graph illustrating the effect of a second medical image including a target prosthesis according to one embodiment;

FIG. 16 is a graph illustrating the effect of a second medical image according to an embodiment;

FIG. 17 is a block diagram of a medical image processing system in accordance with an embodiment;

FIG. 18 is a diagram of an internal structure of a computer device in one embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

In an embodiment, as shown in fig. 1, a medical image processing method is provided, and this embodiment is illustrated by applying the method to a computer device, and it is understood that the computer device may specifically be a terminal or a server. The terminal can be but not limited to various personal computers, notebook computers, smart phones, tablet computers, internet of things equipment and portable wearable equipment, and the internet of things equipment can be intelligent sound boxes, intelligent televisions, intelligent air conditioners, intelligent medical equipment and the like. The portable wearable device can be a smart watch, a smart bracelet, a head-mounted device, and the like. The server may be implemented as a stand-alone server or as a server cluster consisting of a plurality of servers. In this embodiment, the method includes the steps of:

step 102, obtaining a first medical image of a target part, and determining at least three key points corresponding to the target part from the first medical image, wherein the first medical image is a three-dimensional medical image.

The target site refers to an organ, body part or body region that needs medical examination, including but not limited to brain, heart, bone, blood vessel, liver, kidney, gallbladder, pancreas, thyroid, urinary system, uterus and adnexa, teeth, etc. of human or animal. The first medical image refers to a three-dimensional medical image, including but not limited to a three-dimensional CT (Computed Tomography) image.

Optionally, the computer device receives a three-dimensional CT image of the target portion, and performs image preprocessing on the three-dimensional CT image to obtain a first medical image, where the image preprocessing includes, but is not limited to, adjusting a window width and a window level, adjusting a resolution, enhancing the image, normalizing the image, adjusting an image size, and the like. Three key points of the target part are determined in the first medical image and are used for positioning the whole target part in the image.

And 104, constructing an image coordinate system, and determining a reference coordinate system corresponding to the target part based on at least three key points.

Optionally, the computer device first constructs an image coordinate system according to a shooting angle when the first medical image is shot, and then establishes a reference coordinate system in a three-dimensional space according to the three key points of the target portion and the physiological structure of the target portion in the anatomy.

Step 106, performing rotation calibration on the first medical image, so that the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target portion meets a preset condition.

Optionally, the computer device rotates the first medical image from the image coordinate system to the reference coordinate system, and the first medical image synchronously rotates along with the coordinate system until an included angle between the image coordinate system and the reference coordinate system meets an actual requirement.

In a practical real-time manner, the computer device rotates the first medical image from the image coordinate system to the reference coordinate system, and the first medical image synchronously rotates along with the coordinate system until the image coordinate system and the reference coordinate system coincide, that is, the included angle between the two coordinate systems is 0.

And 108, generating a second medical image corresponding to the target part according to the calibrated first medical image, wherein the second medical image is a two-dimensional medical image.

The second medical image is a two-dimensional or three-dimensional medical image, including but not limited to an X-ray image.

Optionally, in a case that the first medical image belongs to a three-dimensional CT image, the computer device calculates the calibrated first medical image by using a related reconstruction algorithm of a Digital Reconstruction Radiological (DRR) image, and generates a simulated reconstructed X-ray image. DDR reconstruction algorithms include algorithms not limited to ray, lightfield, etc.

In the medical image processing method, a first medical image of a target part is obtained, the first medical image is a three-dimensional medical image, and at least three key points corresponding to the target part are determined from the first medical image; therefore, the reference coordinate system corresponding to the target part can be determined based on at least three key points, and then the image coordinate system is constructed according to the shooting angle of the first medical image. Performing rotation calibration on the first medical image to enable the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target part to meet a preset condition; thus, the shooting angle for generating the second medical image can be ensured to meet medical requirements. And generating a second medical image corresponding to the target part according to the calibrated first medical image, wherein the second medical image is a two-dimensional medical image. The three-dimensional medical image can be converted into other two-dimensional medical images, so that the problem that some medical images are difficult to shoot under specific conditions is solved, a doctor can observe special anatomical landmark points which cannot be observed in the three-dimensional image conveniently, and the applicability of the medical images is improved.

In one embodiment, acquiring a first medical image of a target site includes: acquiring a CT three-dimensional image of a target part; according to the type of the part corresponding to the target part, adjusting the window width and the window level of the CT three-dimensional image; resampling the CT three-dimensional image with the window width and the window level adjusted, and adjusting the resolution of the CT three-dimensional image according to a preset resolution; carrying out image enhancement and normalization processing on the CT three-dimensional image with the resolution adjusted; and adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain a first medical image.

Optionally, after the computer device acquires the CT three-dimensional image of the target region, data preprocessing is performed on the CT three-dimensional image, so that the requirement of the neural network model on the input data is met in the process of processing the image by using the neural network model to obtain the key points. As shown in fig. 2, a specific data preprocessing process is performed by setting a specific window width and window level for an input CT three-dimensional image, compressing an HU value range according to a human tissue type (site type) corresponding to a target site, implementing image filtering, and facilitating better learning by a network. And then resampling is carried out, and different resolutions of the CT three-dimensional images obtained every time are unified. And then, data enhancement is carried out, wherein the purpose of the data enhancement is to expand the diversity of data, so that the generalization and the robustness of the neural network model are improved. Further, data normalization processing is needed, data distribution is unified, and convergence of the neural network model is accelerated. And finally, adaptively adjusting the image size of the CT three-dimensional image according to the input requirement of the neural network model, and meeting the requirement of the neural network model on the input size.

In the embodiment, a CT three-dimensional image of a target part is obtained; according to the type of the part corresponding to the target part, adjusting the window width and the window level of the CT three-dimensional image; resampling the CT three-dimensional image with the window width and the window level adjusted, and adjusting the resolution of the CT three-dimensional image according to a preset resolution; carrying out image enhancement and normalization processing on the CT three-dimensional image with the resolution adjusted; and adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain a first medical image. The preprocessed first medical image can achieve the best image expression effect of the target part, and the input requirement of a subsequent neural network model is met.

In one embodiment, determining at least three keypoints corresponding to the target site from the first medical image comprises: inputting the first medical image into the trained neural network model to obtain a thermodynamic diagram corresponding to the first medical image; determining at least three key areas from the first medical image, and determining at least three key areas in the thermodynamic diagram according to the mapping relation between the first medical image and the thermodynamic diagram; and respectively acquiring pixel points meeting preset brightness adjustment in each key area from the thermodynamic diagram as at least three key points.

The trained neural network model is obtained by training based on a medical image training set, the medical image training set comprises a plurality of training samples, each training sample comprises a sample medical image and at least one label, and the labels are used for representing anatomical key points in the sample medical image.

In one possible embodiment, a convolutional neural network of SHVNet (Stacked Hourglass Vnet, Stacked Hourglass network structure) is used for training, resulting in a trained neural network model. The first medical image is input into a trained neural network model for forward calculation. The structure of the key point detection network is shown in fig. 3 and consists of network stacking of a Vnet structure n times. The general structure of each Hourglass Vnet in SHVNet is shown in FIG. 4. Where a convolution block may consist of some different basic structure. The downsampling may be implemented by using a pooling layer, or may be implemented by a convolution step (stride ═ 2). The upsampling may be implemented by interpolation or deconvolution.

Optionally, the computer device determines at least three key points from the first medical image by using a trained neural network model, where the trained neural network model may be but is not limited to an image segmentation model, and the first medical image is processed by the image segmentation model, so that a visualization volume and a neural network thermodynamic diagram can be obtained. As shown in fig. 5, a medical image for which a key point needs to be determined is preprocessed to obtain a first medical image, the first medical image is made to meet the input requirement of the neural network model, then the first medical image is input into the neural network model to obtain a thermodynamic diagram output by the neural network model, and the key point coordinates of the target portion are determined according to the brightest pixel point in the thermodynamic diagram. Three key areas can be defined in the first medical image in advance, and then the brightest pixel point in each key area is taken as a key point coordinate of the target part.

Specifically, for example, the target site is the hip joint as shown in fig. 6, and in the case of demarcating three key regions, three anatomical joint points of the hip joint, two anterior superior iliac spines and the pubic symphysis, respectively, can be identified using a trained neural network. Meanwhile, more key areas can be defined, so that more anatomical key points such as a femoral head central point, greater trochanter of femur, lesser trochanter of femur and the like can be obtained. According to the anatomical key points, some human anatomical features such as anteversion angles, abduction angles and the like can be automatically calculated.

In the embodiment, the first medical image is input into the trained neural network model to obtain a thermodynamic diagram corresponding to the first medical image; determining at least three key areas from the first medical image, and determining at least three key areas in the thermodynamic diagram according to the mapping relation between the first medical image and the thermodynamic diagram; and respectively acquiring pixel points meeting preset brightness adjustment in each key area from the thermodynamic diagram, and taking the pixel points as at least three key points. The method can automatically acquire the anatomical key points of the target part of the patient, reduce the diagnosis workload and improve the diagnosis efficiency.

In one embodiment, performing rotation calibration on the first medical image so that a degree of coincidence between an image coordinate system of the calibrated first medical image and a reference coordinate system corresponding to the target portion satisfies a preset condition includes: extracting two-dimensional images of each layer of each direction plane in the first medical image; determining a rotation parameter corresponding to at least one direction surface in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; at least one of the orientation planes is coronal, sagittal, or transverse; rotating each layer of two-dimensional image of each direction surface according to the rotation parameters corresponding to each direction surface to obtain each layer of process two-dimensional image of each direction surface; and obtaining a calibrated first medical image based on the two-dimensional image of each layer process of each direction surface.

Alternatively, as shown in fig. 7, the three origins are three key points determined from the first medical image. The computer device firstly constructs an image coordinate system (X 'Y' Z ') according to the shooting angle when the first medical image is shot, then establishes a reference coordinate system (XYZ) in a three-dimensional space according to three key points of the target part and the physiological structure of the target part in the anatomy, and then enables the image coordinate system (X' Y 'Z') to coincide with the established reference coordinate system (XYZ) through rotation transformation.

Further, according to the rotation matrix between the two coordinate systems, the computer device determines the corresponding rotation angles (i.e. rotation parameters) for the coronal plane, the sagittal plane, and the transverse plane of the first medical image respectively. And then, carrying out rotary interpolation of a corresponding rotation angle on the two-dimensional image of each layer on each direction surface to obtain a new two-dimensional image (namely a process two-dimensional image). For example, as shown in fig. 8, if the coronal plane of the one-layer two-dimensional image is located at a coronal plane, and the rotation angle corresponding to the coronal plane is 15 degrees, the two-dimensional image shown in fig. a is rotated by 15 degrees to obtain the procedural two-dimensional image shown in fig. b.

In this embodiment, each layer of two-dimensional images of each direction plane in the first medical image is extracted; determining a rotation parameter corresponding to at least one direction surface in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; at least one of the orientation planes is coronal, sagittal, or transverse; rotating each layer of two-dimensional image of each direction surface according to the rotation parameters corresponding to each direction surface to obtain each layer of process two-dimensional image of each direction surface; and obtaining a calibrated first medical image based on the two-dimensional image of each layer process of each direction surface. The method can automatically and accurately adjust and calibrate the position of the target part in the first medical image according to the key point of the target part, and is convenient for obtaining the second medical image with the correct position of the target part subsequently.

In one embodiment, the second medical image comprises an X-ray image, and the second medical image corresponding to the target portion is generated from the calibrated first medical image, comprising: acquiring volume data corresponding to the calibrated first medical image according to the spatial information of the calibrated first medical image; acquiring a ray source position and a ray direction corresponding to a target part, and determining a radiation field area corresponding to volume data according to the ray source position and the ray direction; and generating an X-ray image corresponding to the radiation field region by a digital reconstruction radiographic mode.

Alternatively, in the case that the first medical image belongs to a three-dimensional CT image and the second medical image belongs to a two-dimensional X-ray image, as shown in fig. 9. After the computer device obtains the first medical image after the position calibration, Volume data is established for the first medical image, and the Volume data can be obtained according to spatial information in CT, as shown in fig. 10, three-dimensional CT image data (CT Scan Images) is converted into Volume data (Volume Date), a basic unit forming the Volume data is a Voxel (Voxel), and each Voxel includes both a value corresponding to a pixel in the three-dimensional CT image data and spatial data (position and interval, Slice & Beam) of the pixel. Then determining an irradiation range required by the second medical image, wherein the irradiation range can be a complete target part or a part of interested region in the target part; the position of the simulated ray source (ray source position) and the ray projection direction (ray direction) are set according to the irradiation range, and after the parameters are set, the region of all rays passing through the target part position body data is the field region. Since the CT volume data has already been calibrated by position, the source position can be directly assigned to the midpoint position of the coronal plane. And finally, processing the field area by adopting a DDR (double data rate) reconstruction algorithm to generate an X-ray image corresponding to the field area.

In some embodiments, the DDR reconstruction algorithm may employ a ray-cast algorithm (ray casting algorithm) to generate DRR images by computing the attenuation of each ray at different locations (e.g., bone, soft tissue, water) in the human body. As shown in fig. 11, the DRR is obtained by a set of cone beam projections, the source point of the set of rays is a simulated x-ray source, and the destination points are corresponding pixel points in the DRR image. Each pixel value is a value after attenuation of the corresponding ray in the CT volume data. The ray casting algorithm calculates the attenuation as shown in fig. 12, and when each ray passes through the CT volume data, it has a value I as shown in the following equation:

wherein, I ₀ Is the initial value of the ray, i is the voxel number through which the ray passed, u _i Is the attenuation coefficient, d, corresponding to the voxel i _i Is the distance traversed by the ray in voxel I, n is the total number of voxels traversed, and I is the attenuation value after the ray has traversed all the intersecting voxels. Wherein the attenuation coefficient can be calculated by CT value:

wherein u is _w Hu is the linear attenuation coefficient of water and the CT value of the corresponding voxel.

In some embodiments, the DDR reconstruction algorithm may employ the Ray-sum algorithm (the overlay density projection algorithm), which is an assumption that X-rays are emitted from each pixel source to the volume data and finally imaged on the screen, with the goal of multiplying all Ray lengths through all voxels in the CT volume data by the voxel density to obtain the radiation path. As shown in fig. 13, each pixel in the final imaged DRR image corresponds to a ray, the rays are parallel, and the average of the sum of the values of all the intersecting voxels in the volume data traversed by each ray is the corresponding pixel value in the final image. The Ray-sum algorithm is one of forward projection, i.e. parallel beam forward projection, and can be assumed that the X-Ray source of the Ray casting algorithm is infinitely far from the CT volume data, so its radiation path is defined as:

u _i is the attenuation coefficient corresponding to this voxel I, the value I represents the attenuation value of the radiation path after it has traversed the volume data, the other parameters are as described above, the distance parameter is ignored since the distance across each voxel is the same.

In some embodiments, the ray casting algorithm is time consuming to compute, and the DDR reconstruction algorithm may employ a light field reconstruction algorithm to reconstruct the DRRs. As shown in fig. 14a, regarding plane (u, v) as a plane on which the light source is located, any point (u, v) on the plane represents a point light source, and regarding plane (s, t) as a projection plane, any point (s, t) on the plane is a pixel value of the DRRs image. The volume data is placed between two planes, and a ray from plane (u, v) passes through the volume data and is attenuated before reaching plane (s, t). For any point source (u, v), the light is interpolated from the known neighboring light, as shown in fig. 14 b. Thus, the calculation time of DRR reconstruction can be greatly reduced after the preprocessing.

In this embodiment, volume data corresponding to the calibrated first medical image is acquired according to the spatial information of the calibrated first medical image; acquiring a ray source position and a ray direction corresponding to a target part, and determining a radiation field area corresponding to volume data according to the ray source position and the ray direction; and generating an X-ray image corresponding to the field area through a DDR (double data rate) reconstruction algorithm. The problem that an X-ray machine is poor in applicability and X-ray images are difficult to shoot under specific conditions can be solved on the basis of traditional CT.

In one embodiment, generating a second medical image corresponding to the target region according to the calibrated first medical image includes: obtaining a prosthesis three-dimensional image corresponding to the target prosthesis and adjusting the posture information of the prosthesis three-dimensional image; fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fused medical image; a second medical image corresponding to the target site is generated based on the prosthesis fusion medical image.

Optionally, under the condition that the first medical image belongs to a three-dimensional CT image, if there is a need to install a prosthesis at a target portion, a prosthesis three-dimensional image of the same type as the first medical image may be constructed according to the target prosthesis that needs to be used, then posture information such as an angle, a height, a placement position and the like of the prosthesis three-dimensional image is adjusted according to a human anatomy structure, the prosthesis three-dimensional image and the calibrated first medical image are superposed and fused to obtain a prosthesis fusion medical image, and then a DDR reconstruction algorithm is adopted to generate a second medical image corresponding to the target portion based on the prosthesis fusion medical image. The second medical image may include an image of the target prosthesis, allowing the physician to confirm the specific anatomical landmark points after installation of the target prosthesis prior to surgery. For example, as shown in fig. 15, a DRR image (corresponding to a second medical image) generated by volume data synthesized by a prosthesis (corresponding to a three-dimensional image of the prosthesis, a white portion at the bottom right in the figure) and a hip joint CT (corresponding to a first medical image) includes an X-ray image of the prosthesis, i.e., a doctor can see information such as a position and an angle after the prosthesis is implanted before an operation, and can adjust the segmented femur by using an established coordinate system, for example, to a position of a maximum inclination angle, and project an X-ray image of a relevant angle, which is not available in conventional X-ray, so as to facilitate preoperative planning for the doctor.

In the embodiment, a prosthesis three-dimensional image corresponding to a target prosthesis is obtained, and posture information of the prosthesis three-dimensional image is adjusted; fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fused medical image; a second medical image corresponding to the target site is generated based on the prosthesis fusion medical image. The first medical image can be adjusted, the target prosthesis is added in the first medical image, and the prosthesis fusion medical image containing the target prosthesis is obtained, so that the second medical image containing the target prosthesis is generated, and prosthesis transplantation planning is facilitated.

In one embodiment, generating a second medical image corresponding to the target region from the calibrated first medical image comprises: determining at least one target sub-part from the calibrated first medical image; the target sub-part is obtained by dividing the target part; in the calibrated first medical image, adjusting the posture information of at least one target sub-part to obtain a posture adjustment medical image; and generating a second medical image corresponding to the target part according to the posture adjustment medical image.

Optionally, under the condition that the first medical image belongs to the three-dimensional CT image, if the target portion has a requirement for adjusting the posture of a partial structure, the target portion may be divided into a plurality of target sub-portions according to human anatomy, and then the prosthesis three-dimensional image and the calibrated first medical image are superimposed and fused on posture information of each target sub-portion of the first medical image, such as an angle, a height, a placement position, and the like, to obtain a posture-adjusted medical image; and then generating a second medical image corresponding to the target part based on the posture adjustment medical image by adopting a DDR (double data rate) reconstruction algorithm. For example, as shown in fig. 16, posture parameters of a femur (target sub-part) in a hip joint (target part) CT (corresponding to a first medical image) are adjusted, and then a DRR image (corresponding to a second medical image) generated by the hip joint CT is projected, which can measure a physiological structure like a conventional image, and can observe the effect after the posture adjustment of each bone, as shown in the figure, the lower right marked line is a femoral neck angle, and can measure a CE angle, an anteversion angle, an abduction angle, and the like. The T-shaped marked line in the middle area is a coordinate system established by anatomical key points identified by an AI network, so that key structures such as teardrops and the like can be clearly seen from a DRR image generated by CT after the coordinate system is aligned, and preoperative planning is facilitated.

In this embodiment, at least one target sub-portion is determined from the calibrated first medical image; the target sub-part is obtained by dividing the target part; in the calibrated first medical image, adjusting the posture information of at least one target sub-part to obtain a posture adjustment medical image; and generating a second medical image corresponding to the target part according to the posture adjustment medical image. The first medical image can be adjusted, the posture of the target sub-part in the target part is adjusted in the first medical image, the posture adjustment medical image after the posture of the target sub-part is adjusted is obtained, and therefore the second medical image after the posture of the target sub-part is adjusted is generated, and preoperative planning related to the target part is facilitated.

In one embodiment, a medical image processing method includes:

acquiring a CT three-dimensional image of a target part; according to the type of the part corresponding to the target part, adjusting the window width and the window level of the CT three-dimensional image; resampling the CT three-dimensional image with the window width and the window level adjusted, and adjusting the resolution of the CT three-dimensional image according to a preset resolution; carrying out image enhancement and normalization processing on the CT three-dimensional image with the resolution adjusted; and adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain a first medical image. Wherein, the first medical image is a three-dimensional medical image.

Inputting the first medical image into the trained neural network model to obtain a thermodynamic diagram corresponding to the first medical image; determining at least three key areas from the first medical image, and determining at least three key areas in the thermodynamic diagram according to the mapping relation between the first medical image and the thermodynamic diagram; and respectively acquiring pixel points meeting preset brightness adjustment in each key area from the thermodynamic diagram as at least three key points.

And constructing an image coordinate system, and determining a reference coordinate system corresponding to the target part based on at least three key points.

Extracting two-dimensional images of each layer of each direction surface in the first medical image; determining a rotation parameter corresponding to at least one direction surface in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; at least one of the orientation planes is coronal, sagittal, or transverse; rotating each layer of two-dimensional image of each direction surface according to the rotation parameters corresponding to each direction surface to obtain each layer of process two-dimensional image of each direction surface; and obtaining a calibrated first medical image based on the two-dimensional image of each layer process of each direction surface.

Acquiring volume data corresponding to the calibrated first medical image according to the spatial information of the calibrated first medical image; acquiring a ray source position and a ray direction corresponding to a target part, and determining a radiation field area corresponding to volume data according to the ray source position and the ray direction; and generating an X-ray image corresponding to the radiation field region by a digital reconstruction radiographic mode.

Or acquiring a prosthesis three-dimensional image corresponding to the target prosthesis and adjusting the posture information of the prosthesis three-dimensional image; fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fused medical image; a second medical image corresponding to the target site is generated based on the prosthesis fusion medical image.

Or, determining at least one target sub-part from the calibrated first medical image; the target sub-part is obtained by dividing the target part; in the calibrated first medical image, adjusting the posture information of at least one target sub-part to obtain a posture adjustment medical image; and adjusting the medical image according to the posture to generate a second medical image corresponding to the target part.

The second medical image is a two-dimensional medical image.

It should be understood that, although the steps in the flowcharts related to the embodiments described above are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a part of the steps in the flowcharts related to the embodiments described above may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the execution order of the steps or stages is not necessarily sequential, but may be rotated or alternated with other steps or at least a part of the steps or stages in other steps.

Based on the same inventive concept, the embodiment of the present application further provides a medical image processing system for implementing the medical image processing method mentioned above. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme described in the above method, so specific limitations in one or more embodiments of the medical image processing system provided below can be referred to the limitations of the medical image processing method in the above description, and details are not repeated here.

In one embodiment, as shown in fig. 17, there is provided a medical image processing system 1700 comprising: an obtaining module 1701, a positioning module 1702, a converting module 1703, and a generating module 1704, wherein:

an obtaining module 1701, configured to obtain a first medical image of a target portion, and determine at least three key points corresponding to the target portion from the first medical image, where the first medical image is a three-dimensional medical image;

a positioning module 1702, configured to construct an image coordinate system, and determine a reference coordinate system corresponding to the target portion based on at least three key points;

a conversion module 1703, configured to perform rotation calibration on the first medical image, so that a coincidence degree between an image coordinate system of the calibrated first medical image and a reference coordinate system corresponding to the target portion meets a preset condition;

a generating module 1704, configured to generate a second medical image corresponding to the target portion according to the calibrated first medical image, where the second medical image is a two-dimensional medical image.

In one embodiment, the acquisition module 1701 further includes:

In one embodiment, the obtaining module 1701 is further configured to input the first medical image into the trained neural network model to obtain a thermodynamic diagram corresponding to the first medical image; determining at least three key areas from the first medical image, and determining at least three key areas in the thermodynamic diagram according to the mapping relation between the first medical image and the thermodynamic diagram; and respectively acquiring pixel points meeting preset brightness adjustment in each key area from the thermodynamic diagram as at least three key points.

In one embodiment, the conversion module 1703 is further configured to extract a two-dimensional image of each layer of each orientation plane in the first medical image; determining a rotation parameter corresponding to at least one direction plane in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; at least one of the orientation planes is coronal, sagittal, or transverse; rotating each layer of two-dimensional image of each direction surface according to the rotation parameters corresponding to each direction surface to obtain each layer of process two-dimensional image of each direction surface; and obtaining a calibrated first medical image based on the two-dimensional image of each layer of process of each direction surface.

In one embodiment, the generation module 1704 is further configured to obtain volume data corresponding to the calibrated first medical image according to the spatial information of the calibrated first medical image; acquiring a ray source position and a ray direction corresponding to a target part, and determining a radiation field area corresponding to volume data according to the ray source position and the ray direction; and generating an X-ray image corresponding to the radiation field region by a digital reconstruction radiographic mode.

In one embodiment, the generation module 1704 is further configured to obtain a three-dimensional image of the prosthesis corresponding to the target prosthesis, and adjust pose information of the three-dimensional image of the prosthesis; fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fusion medical image; a second medical image corresponding to the target site is generated based on the prosthesis fusion medical image.

In one embodiment, the generation module 1704 is further configured to determine at least one target sub-site from the calibrated first medical image; the target sub-part is obtained by dividing the target part; in the calibrated first medical image, adjusting the posture information of at least one target sub-part to obtain a posture adjustment medical image; and generating a second medical image corresponding to the target part according to the posture adjustment medical image.

In one embodiment, the generating module 1704 is further configured to process the volume data corresponding to the portal area by using a ray-casting algorithm to generate an X-ray image corresponding to the portal area; or, the generating module 1701 is further configured to process the volume data corresponding to the portal area through a superposition density projection algorithm, and generate an X-ray image corresponding to the portal area; alternatively, the generating module 1704 is further configured to process the volume data corresponding to the portal region through a light field reconstruction algorithm to generate an X-ray image corresponding to the portal region.

The modules in the medical image processing system can be wholly or partially implemented by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.

In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as shown in fig. 18. The computer apparatus includes a processor, a memory, an input/output interface, a communication interface, a display unit, and an input device. The processor, the memory and the input/output interface are connected by a system bus, and the communication interface, the display unit and the input device are connected by the input/output interface to the system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The input/output interface of the computer device is used for exchanging information between the processor and an external device. The communication interface of the computer device is used for communicating with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. The computer program is executed by a processor to implement a medical image processing method. The display unit of the computer equipment is used for forming a visual and visible picture, and can be a display screen, a projection device or a virtual reality imaging device, the display screen can be a liquid crystal display screen or an electronic ink display screen, the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.

Those skilled in the art will appreciate that the architecture shown in fig. 18 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program: acquiring a first medical image of a target part, and determining at least three key points corresponding to the target part from the first medical image, wherein the first medical image is a three-dimensional medical image; constructing an image coordinate system, and determining a reference coordinate system corresponding to the target part based on at least three key points; performing rotation calibration on the first medical image to enable the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target part to meet a preset condition; and generating a second medical image corresponding to the target part according to the calibrated first medical image, wherein the second medical image is a two-dimensional medical image.

In one embodiment, the processor, when executing the computer program, further performs the steps of: acquiring a CT three-dimensional image of a target part; according to the part type corresponding to the target part, adjusting the window width and the window level of the CT three-dimensional image; resampling the CT three-dimensional image with the window width and the window level adjusted, and adjusting the resolution of the CT three-dimensional image according to a preset resolution; carrying out image enhancement and normalization processing on the CT three-dimensional image with the resolution adjusted; and adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain a first medical image.

In one embodiment, the processor, when executing the computer program, further performs the steps of: inputting the first medical image into the trained neural network model to obtain a thermodynamic diagram corresponding to the first medical image; determining at least three key areas from the first medical image, and determining at least three key areas in the thermodynamic diagram according to the mapping relation between the first medical image and the thermodynamic diagram; and respectively acquiring pixel points meeting preset brightness adjustment in each key area from the thermodynamic diagram as at least three key points.

In one embodiment, the processor, when executing the computer program, further performs the steps of: extracting two-dimensional images of each layer of each direction plane in the first medical image; determining a rotation parameter corresponding to at least one direction surface in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; at least one of the orientation planes is coronal, sagittal, or transverse; rotating each layer of two-dimensional image of each direction surface according to the rotation parameters corresponding to each direction surface to obtain each layer of process two-dimensional image of each direction surface; and obtaining a calibrated first medical image based on the two-dimensional image of each layer process of each direction surface.

In one embodiment, the processor, when executing the computer program, further performs the steps of: acquiring volume data corresponding to the calibrated first medical image according to the spatial information of the calibrated first medical image; acquiring a ray source position and a ray direction corresponding to a target part, and determining a radiation field area corresponding to volume data according to the ray source position and the ray direction; and generating an X-ray image corresponding to the radiation field region by a digital reconstruction radiographic mode.

In one embodiment, the processor, when executing the computer program, further performs the steps of: acquiring a prosthesis three-dimensional image corresponding to a target prosthesis, and adjusting the posture information of the prosthesis three-dimensional image; fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fused medical image; a second medical image corresponding to the target site is generated based on the prosthesis fusion medical image.

In one embodiment, the processor, when executing the computer program, further performs the steps of: determining at least one target sub-part from the calibrated first medical image; the target sub-part is obtained by dividing the target part; in the calibrated first medical image, adjusting the posture information of at least one target sub-part to obtain a posture adjustment medical image; and adjusting the medical image according to the posture to generate a second medical image corresponding to the target part.

In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of: acquiring a first medical image of a target part, and determining at least three key points corresponding to the target part from the first medical image, wherein the first medical image is a three-dimensional medical image; constructing an image coordinate system, and determining a reference coordinate system corresponding to the target part based on at least three key points; performing rotation calibration on the first medical image to enable the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target part to meet a preset condition; and generating a second medical image corresponding to the target part according to the calibrated first medical image, wherein the second medical image is a two-dimensional medical image.

In one embodiment, the computer program when executed by the processor further performs the steps of: acquiring a CT three-dimensional image of a target part; according to the type of the part corresponding to the target part, adjusting the window width and the window level of the CT three-dimensional image; resampling the CT three-dimensional image with the window width and the window level adjusted, and adjusting the resolution of the CT three-dimensional image according to a preset resolution; carrying out image enhancement and normalization processing on the CT three-dimensional image with the resolution adjusted; and adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain a first medical image.

In one embodiment, the computer program when executed by the processor further performs the steps of: inputting the first medical image into the trained neural network model to obtain a thermodynamic diagram corresponding to the first medical image; determining at least three key areas from the first medical image, and determining at least three key areas in the thermodynamic diagram according to the mapping relation between the first medical image and the thermodynamic diagram; and respectively acquiring pixel points meeting preset brightness adjustment in each key area from the thermodynamic diagram as at least three key points.

In one embodiment, the computer program when executed by the processor further performs the steps of: extracting two-dimensional images of each layer of each direction surface in the first medical image; determining a rotation parameter corresponding to at least one direction plane in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; at least one of the orientation planes is coronal, sagittal, or transverse; rotating each layer of two-dimensional image of each direction surface according to the rotation parameters corresponding to each direction surface to obtain each layer of process two-dimensional image of each direction surface; and obtaining a calibrated first medical image based on the two-dimensional image of each layer process of each direction surface.

In one embodiment, the computer program when executed by the processor further performs the steps of: acquiring volume data corresponding to the calibrated first medical image according to the spatial information of the calibrated first medical image; acquiring a ray source position and a ray direction corresponding to a target part, and determining a radiation field area corresponding to volume data according to the ray source position and the ray direction; and generating an X-ray image corresponding to the radiation field region by a digital reconstruction radiographic mode.

In one embodiment, the computer program when executed by the processor further performs the steps of: acquiring a prosthesis three-dimensional image corresponding to a target prosthesis, and adjusting the posture information of the prosthesis three-dimensional image; fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fusion medical image; a second medical image corresponding to the target site is generated based on the prosthesis fusion medical image.

In one embodiment, the computer program when executed by the processor further performs the steps of: determining at least one target sub-part from the calibrated first medical image; the target sub-part is obtained by dividing the target part; in the calibrated first medical image, adjusting the posture information of at least one target sub-part to obtain a posture adjustment medical image; and adjusting the medical image according to the posture to generate a second medical image corresponding to the target part.

In one embodiment, a computer program product is provided, comprising a computer program which when executed by a processor performs the steps of: acquiring a first medical image of a target part, and determining at least three key points corresponding to the target part from the first medical image, wherein the first medical image is a three-dimensional medical image; constructing an image coordinate system, and determining a reference coordinate system corresponding to the target part based on at least three key points; performing rotation calibration on the first medical image to enable the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target part to meet a preset condition; and generating a second medical image corresponding to the target part according to the calibrated first medical image, wherein the second medical image is a two-dimensional medical image.

In one embodiment, the computer program when executed by the processor further performs the steps of: extracting two-dimensional images of each layer of each direction surface in the first medical image; determining a rotation parameter corresponding to at least one direction surface in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; at least one of the orientation planes is coronal, sagittal, or transverse; rotating each layer of two-dimensional image of each direction surface according to the rotation parameters corresponding to each direction surface to obtain each layer of process two-dimensional image of each direction surface; and obtaining a calibrated first medical image based on the two-dimensional image of each layer process of each direction surface.

In one embodiment, the computer program when executed by the processor further performs the steps of: acquiring a prosthesis three-dimensional image corresponding to a target prosthesis, and adjusting the posture information of the prosthesis three-dimensional image; fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fused medical image; a second medical image corresponding to the target site is generated based on the prosthesis fusion medical image.

In one embodiment, the computer program when executed by the processor further performs the steps of: determining at least one target sub-part from the calibrated first medical image; the target sub-part is obtained by dividing the target part; in the calibrated first medical image, adjusting the posture information of at least one target sub-part to obtain a posture adjustment medical image; and generating a second medical image corresponding to the target part according to the posture adjustment medical image.

It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) referred to in the present application are information and data authorized by the user or sufficiently authorized by each party, and the collection, use and processing of the related data need to comply with the relevant laws and regulations and standards of the relevant country and region.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above may be implemented by hardware instructions of a computer program, which may be stored in a non-volatile computer-readable storage medium, and when executed, may include the processes of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high-density embedded nonvolatile Memory, resistive Random Access Memory (ReRAM), Magnetic Random Access Memory (MRAM), Ferroelectric Random Access Memory (FRAM), Phase Change Memory (PCM), graphene Memory, and the like. Volatile Memory can include Random Access Memory (RAM), external cache Memory, and the like. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others. The databases referred to in various embodiments provided herein may include at least one of relational and non-relational databases. The non-relational database may include, but is not limited to, a block chain based distributed database, and the like. The processors referred to in the embodiments provided herein may be general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing based data processing logic devices, etc., without limitation.

The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method of medical image processing, the method comprising:

constructing an image coordinate system, and determining a reference coordinate system corresponding to the target part based on the at least three key points;

2. The method of claim 1, wherein the obtaining a first medical image of a target site comprises:

acquiring a CT three-dimensional image of a target part;

adjusting the window width and the window level of the CT three-dimensional image according to the part type corresponding to the target part;

resampling the CT three-dimensional image with the window width and the window position adjusted, and adjusting the resolution of the CT three-dimensional image according to a preset resolution;

and adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain the first medical image.

3. The method of claim 1, wherein said determining at least three keypoints corresponding to the target site from the first medical image comprises:

inputting the first medical image into a trained neural network model to obtain a thermodynamic diagram corresponding to the first medical image;

and respectively acquiring pixel points meeting preset brightness adjustment in each key area from the thermodynamic diagram as the at least three key points.

4. The method according to claim 1, wherein the performing the rotation calibration on the first medical image so that the coincidence degree between the image coordinate system of the calibrated first medical image and the reference coordinate system corresponding to the target portion satisfies a predetermined condition comprises:

determining a rotation parameter corresponding to at least one direction surface in the first medical image according to a rotation matrix between the image coordinate system and the reference coordinate system; the at least one orientation plane is a coronal plane, a sagittal plane, or a transverse plane;

5. The method of claim 1, wherein the second medical image comprises an X-ray image, and wherein generating the second medical image corresponding to the target site from the calibrated first medical image comprises:

acquiring a ray source position and a ray direction corresponding to the target part, and determining a radiation field area corresponding to the volume data according to the ray source position and the ray direction;

and generating an X-ray image corresponding to the field area by a digital reconstruction radiographic mode.

6. The method of claim 5, wherein generating the X-ray image corresponding to the portal region by digitally reconstructing the radiographic image comprises:

7. The method according to any one of claims 1 to 6, wherein the generating a second medical image corresponding to the target site from the calibrated first medical image comprises:

fusing the prosthesis three-dimensional image and the calibrated first medical image to obtain a prosthesis fusion medical image;

8. The method according to any one of claims 1 to 6, wherein the generating a second medical image corresponding to the target site from the calibrated first medical image comprises:

determining at least one target sub-site from the calibrated first medical image; the target sub-part is obtained by dividing the target part;

in the calibrated first medical image, adjusting the posture information of the at least one target sub-part to obtain a posture-adjusted medical image;

and generating a second medical image corresponding to the target part according to the posture adjustment medical image.

9. A medical image processing system, the system comprising:

the system comprises an acquisition module, a processing module and a display module, wherein the acquisition module is used for acquiring a first medical image of a target part and determining at least three key points corresponding to the target part from the first medical image, and the first medical image is a three-dimensional medical image;

the positioning module is used for constructing an image coordinate system and determining a reference coordinate system corresponding to the target part based on the at least three key points;

and the generating module is used for generating a second medical image corresponding to the target part according to the calibrated first medical image, wherein the second medical image is a two-dimensional medical image.

10. The system of claim 9, wherein the acquisition module comprises:

the resampling unit is used for resampling the CT three-dimensional image with the window width and the window level adjusted and adjusting the resolution of the CT three-dimensional image according to a preset resolution;

and the self-adaptive unit is used for adjusting the size of the CT three-dimensional image after image enhancement and normalization processing according to the network input parameters corresponding to the trained neural network model to obtain the first medical image.

11. The system of claim 10, wherein the generating module is further configured to process the volume data corresponding to the portal region through a ray-casting algorithm to generate an X-ray image corresponding to the portal region; or the generating module is further configured to process volume data corresponding to the portal area through a superposition density projection algorithm, and generate an X-ray image corresponding to the portal area; or the generating module is further configured to process the volume data corresponding to the portal region through a light field reconstruction algorithm, and generate an X-ray image corresponding to the portal region.

12. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 8.

13. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 8.