Disclosure of Invention
The embodiment of the invention aims to provide a spinal surgery preoperative planning method, system and storage medium based on artificial intelligence, which are used for solving the problem that an optimal implantation path and an optimal screw model of a screw cannot be obtained in the prior art.
To achieve the above object, an embodiment of the present invention provides a spinal surgery preoperative planning method based on artificial intelligence, the method specifically including:
Preprocessing the imported spine CT data to obtain a first preprocessed image;
inputting the first preprocessing image into a trained neural network model to obtain a screw feeding point and a screw feeding direction of a screw, calculating the diameter and the length of the screw based on the screw feeding point and the screw feeding direction, and determining the type of the adapted screw in a prosthesis database based on the diameter and the length of the screw.
Based on the technical scheme, the invention can also be improved as follows:
Further, the preprocessing the imported spine CT data to obtain a first preprocessed image includes:
Generating a two-dimensional cross-section image view, a two-dimensional coronal image view, a two-dimensional sagittal image view and a three-dimensional image view which are consistent in three-dimensional coordinate system based on the spine CT data;
Prior to prosthesis matching, surgical planning information for the CT data is acquired in response to an acquisition instruction, wherein the surgical planning information includes disease type information, surgical type information, and surgical segment information.
Further, the inputting the first preprocessed image into the trained neural network model to obtain the nailing point and the nailing direction of the screw includes:
Constructing a neural network model, wherein the neural network comprises a convolution layer, a pooling layer, a replication layer of the convolution layer and a feature layer corresponding to the replication layer;
The corresponding characteristic layers are obtained through up-sampling addition of the replication layer and the corresponding layers in the convolution layer, a large characteristic layer is obtained after the characteristic layers are overlapped, a thermodynamic diagram representing the probability of key points is generated through 1x1 convolution, the point with the maximum probability value in the thermodynamic diagram is taken as the characteristic point, and the position of the characteristic point is the predicted nailing point position;
acquiring spine CT data, and preprocessing the spine CT data to obtain a second preprocessed image;
Dividing the second preprocessed image into a training set, a testing set and a verification set;
Training the neural network model based on the training set;
Performing performance verification on the Hourg l ass neural network model based on the verification set, and storing the neural network model meeting performance conditions;
and evaluating the calculation result of the nailing point and the nailing direction of the neural network model based on the test set.
Further, the pre-operative planning method for spinal surgery further comprises:
calculating the width of the U-shaped structure at the tail of the screw, fitting a titanium rod curve, calculating the optimal titanium rod width based on the width of the U-shaped structure at the tail of the screw, and determining the type of the adaptive titanium rod in a prosthesis database based on the optimal titanium rod width;
Fitting the key points of the titanium rod through cubic spline interpolation to obtain a fitted titanium rod curve.
Further, the pre-operative planning method for spinal surgery further comprises:
and calculating the required transverse connection length based on the titanium rod fitting curve, and determining the adapted transverse connection model in a prosthesis database based on the required transverse connection length.
Further, the calculating the required cross-link length based on the titanium rod fitting curve, and determining the adapted cross-link model based on the required cross-link length in a prosthesis database comprises:
A first punctuation and a second punctuation are obtained at certain length positions of the left titanium rod and the right titanium rod, and the first punctuation and the second punctuation are connected through a straight line;
The straight line is translated up and down within the length range of the titanium rod, and in the translation process, the straight line is always connected with the left titanium rod and the right titanium rod;
The length interval of the straight line is the length range of the transverse connection, the maximum value of the length is m, the calculated value a=the transverse connection model-m, and when a is more than or equal to 0 and the value a is minimum, the optimal transverse connection model is obtained.
Further, the pre-operative planning method for spinal surgery further comprises:
The length and width of the two adjacent spinal interspaces are calculated, and the model of the adaptive fusion device is determined in the prosthesis database based on the length and width of the two adjacent spinal interspaces.
Further, the calculating the length and width of the two adjacent spinal interspaces, determining an adapted cage model in a prosthesis database based on the two adjacent spinal interspaces length and width, comprising:
preprocessing the spine CT data to obtain a third preprocessed image, wherein the third preprocessed image only keeps a bone area and a background area;
counting the area of a bone region in each third preprocessed image, and searching the third preprocessed image with the largest area;
Searching for connected areas in the third preprocessing image, and acquiring edge points based on edges in each connected area;
and calculating the length and the width of the intervertebral space between two adjacent vertebrae based on the edge points.
An artificial intelligence based pre-operative planning system for spinal surgery, comprising:
The importing verification module is used for preprocessing imported spine CT data to obtain a first preprocessed image;
The screw planning module is used for inputting the first preprocessing image into a trained neural network model to obtain a screw feeding point and a screw feeding direction of a screw, calculating the screw diameter and the screw length based on the screw feeding point and the screw feeding direction, and determining the adaptive screw model in a prosthesis database based on the screw diameter and the screw length.
A non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method.
The embodiment of the invention has the following advantages:
According to the artificial intelligence-based spine operation pre-operation planning method, the imported spine CT data are preprocessed to obtain a first preprocessed image; inputting the first preprocessing image into a trained neural network model to obtain a nailing point and a nailing direction of a screw; the method comprises the steps of obtaining a corresponding characteristic layer through up-sampling addition of a replication layer of a neural network model and a corresponding layer in a convolution layer, obtaining a large characteristic layer after superposition of the characteristic layer, generating a thermodynamic diagram representing the probability of key points through 1x1 convolution, taking the maximum probability value point in the thermodynamic diagram as a characteristic point, taking the position of the characteristic point as the predicted nailing point position, intelligently calculating the optimal nailing point and nailing direction of a screw by combining with a medical theory, displaying the calculated nailing point and nailing direction in four views of a vertebra, supporting a user to finely adjust according to personal operation habits, calculating the diameter and the length of the screw based on the nailing point and the nailing direction, and determining the matched screw model in a prosthesis database based on the diameter and the length of the screw. The problem that an optimal implantation path and an optimal screw model of a screw cannot be obtained in the prior art is solved.
Detailed Description
Other advantages and advantages of the present invention will become apparent to those skilled in the art from the following detailed description, which, by way of illustration, is to be read in connection with certain specific embodiments, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Examples
Fig. 1 is a flowchart of an embodiment of a spinal surgery pre-operation planning method based on artificial intelligence, and as shown in fig. 1, the spinal surgery pre-operation planning method based on artificial intelligence provided by the embodiment of the invention comprises the following steps:
s101, preprocessing imported spine CT data to obtain a first preprocessed image;
Specifically, the pre-operation planning system of the spine is installed on a computer, and spine CT data is imported after the pre-operation planning system of the spine runs, wherein the spine CT data is a DICOM file, DICOM (DIGITAL IMAGING AND Communications IN MEDICINE), namely medical digital imaging and communication, and is an international standard (ISO 12052) of medical images and related information. It defines a medical image format that can be used for data exchange with quality meeting clinical needs. After the three-axis linkage effect is achieved, namely, a planner performs clicking operation in any view, the spine preoperative planning system automatically positions a focus to a clicking position, and focuses of the three other views are also positioned to the same position.
Prior to prosthesis matching, surgical planning information for the CT data is acquired in response to an acquisition instruction, wherein the surgical planning information includes disease type information, surgical type information, and surgical segment information.
Preferably, the pre-operation planning system for spine surgery automatically divides one spine into a plurality of independent spines based on an algorithm, and provides a case information editing function to support a user to edit disease type information, operation type information and operation segment information of imported spine CT data. In a subsequent surgical planning module, the pre-operative spinal planning system provides the user with a corresponding prosthesis database based on the disease type information and the surgical type information, defining a planning range based on the surgical segment information.
S102, inputting a first preprocessing image into a trained neural network model to obtain a screw feeding point and a screw feeding direction of a screw, calculating the diameter and the length of the screw based on the screw feeding point and the screw feeding direction, and determining an adaptive screw model in a prosthesis database based on the diameter and the length of the screw;
Alternatively, a neural network model Hourglass is described below as an example. As shown in fig. 14, constructing Hourglass a neural network model, wherein the Hourglass neural network structure is an hourglass structure and can output pixel-level predictions, and the Hourglass neural network comprises convolution layers (C1-C7), pooling layers, replication layers (C1 a-C4 a) of the convolution layers and feature layers corresponding to the replication layers; the method comprises the steps of obtaining a characteristic layer (C1 a-C4 a) with a copy layer as a middle part, obtaining new characteristic information by up-sampling and adding the copy layer and a corresponding layer in a convolution layer, obtaining characteristic fusion effect, namely, obtaining a C1b-C4b part in the graph, wherein the whole Hourglass is symmetrical, each network layer is corresponding to the corresponding network layer in the process of obtaining low-resolution characteristics, obtaining a large characteristic layer (C1 b) by superposing the characteristic layers in the up-sampling process, not only retaining information of all layers, but also being equal to the input original image in size, generating a thermodynamic diagram (heatmap) representing the probability of key points by 1x1 convolution, taking the maximum probability value point in the thermodynamic diagram as a characteristic point, and taking the position of the characteristic point as a predicted nail-in point position.
The training process of the Hourglass neural network model is as follows: acquiring spine CT data, and preprocessing the spine CT data to obtain a second preprocessed image; specifically, spine CT data is obtained, the spine CT data is a DICOM file, a nail-in point to be identified is marked to generate MASK, and the DICOM file and the corresponding MASK are input into a Hourglass network for iterative training. Dividing the second preprocessed image into a training set, a testing set and a verification set;
Training the Hourglass neural network model based on the training set; the network was trained using Tensorflow 1.15.15 frames and NVIDIA RTX 2070, setting the number of iterations 100000, taking 15 hours to complete model training.
In the model training process, the trained batch_size is 4, the initial learning rate is set to be 1e-4, a learning rate attenuation strategy is added, the learning rate is attenuated to be 0.95 at intervals of 5000 item, the optimizer uses an Adam optimizer, the loss function used is DICE loss, the built network is used for training a segmented training set, 1000 iterations are set, and one time of verification is performed on the training set and the verification set. And judging the stopping time of the network training by a premature stopping method to obtain a final model.
The verification process of Hourglass neural network model is: performing performance verification on the Hourglass neural network model based on the verification set, and storing the Hourglass neural network model meeting performance conditions;
The test process of Hourglass neural network model is: and evaluating the nailing point and nailing direction calculation result of the Hourglass neural network model based on the test set.
Preferably, as shown in fig. 8-10, after the introduction confirmation module 10 completes the introduction confirmation, the screw planning module 20 is entered, the vertebral characteristic points are automatically identified based on Hourglass neural network model, and the optimal screw insertion point and screw insertion direction are calculated by combining with medical theory intelligence. The calculated nailing point and nailing direction are displayed in four views of the spine, so that the user is supported to finely adjust according to personal operation habits.
The diameter and the cone depth of the pedicle of the vertebra are automatically calculated, the optimal prosthesis model is intelligently matched with a prosthesis database by combining a medical theory, and the effect of prosthesis implantation is simulated. The situation that after the prosthesis is implanted, each layer of two-dimensional slice is checked by a user layer by layer is supported. After the planning of a certain section is completed, the system is switched to the next section, and the system automatically copies the screw planning data to the next section so as to support the user to adjust on the basis of the original planning scheme.
In a specific example, taking T2 (the second thoracic segment) as an example, the medical principle of calculating the insertion point, insertion direction, and matching the prosthesis type number is taught:
spinal characteristics as shown in fig. 11, left access point (point a) as shown in fig. 12: the inferior articular process midpoint was identified, and a vertical line was drawn 3mm outside of this point. The transverse process substrate is identified and a horizontal line is drawn 1/3 of the way across the substrate. Taking the intersection point of the two lines. Right side staple in point (point b): the transverse process midpoint is identified, and a horizontal line is drawn through the midpoint. The superior articular process outer edge is identified, and a vertical line is drawn through the superior articular process outer edge. Taking the intersection point of the two lines. The nailing direction: in cross section, the left and right screws are each inclined inward by 35 °. In the sagittal plane, the left and right side screws are parallel to the upper and lower endplates.
As shown in fig. 13, screw diameter: perpendicular to the direction of insertion, the pedicle was segmented into n facets. The pedicle diameter for each facet was calculated. When the minimum diameter of the pedicles in all the surfaces is less than or equal to 4.0mm, the diameter of the screw is 3.5mm. When the pedicle minimum diameter in all facets is greater than 4.0mm, the screw diameter is chosen to be 4.0mm.
Screw length: starting from the nailing point, a straight line is drawn along the nailing direction until the cortical bone at the front edge of the vertebral body is broken through. The distance from the nailing point to the breakthrough point was calculated, and 80% of this distance was the nailing depth of the screw. The calculated value a = screw length-depth of penetration 80%. a > 0 and the value of a is the smallest, the length of the screw is the best.
According to the artificial intelligence-based spine operation pre-operation planning method, the imported spine CT data are preprocessed to obtain a first preprocessed image; inputting the first preprocessing image into a trained Hourg l ass neural network model to obtain a nailing point and a nailing direction of a screw; the method comprises the steps of obtaining a corresponding characteristic layer through up-sampling addition of a replication layer of Hourg l ass neural network models and a corresponding layer in a convolution layer, obtaining a large characteristic layer after superposition of the characteristic layer, generating a thermodynamic diagram representing the probability of key points through 1x1 convolution, taking the maximum probability value point in the thermodynamic diagram as a characteristic point, taking the position of the characteristic point as the predicted nailing point position, intelligently calculating the optimal nailing point and nailing direction of a screw by combining with a medical theory, displaying the calculated nailing point and nailing direction in four views of a vertebra, supporting a user to finely adjust according to personal operation habits, calculating the diameter and the length of the screw based on the nailing point and the nailing direction, and determining the matched screw model in a prosthesis database based on the diameter and the length of the screw. The problem that an optimal implantation path and an optimal screw model of a screw cannot be obtained in the prior art is solved.
As shown in fig. 2, the spinal surgery pre-operation planning method based on artificial intelligence of the invention further comprises:
s103, calculating the width of the U-shaped structure at the tail part of the screw, fitting a titanium rod curve, calculating the optimal titanium rod width based on the width of the U-shaped structure at the tail part of the screw, and determining the type of the adaptive titanium rod in a prosthesis database based on the optimal titanium rod width;
specifically, as shown in fig. 16, when the screw planning module 20 completes the screw planning, the titanium rod planning module 30 may be entered. The system automatically fits a titanium rod shape curve according to the U-shaped structure position of the tail part of the screw. And the screw position is manually adjusted by a user, and the fitting curve of the titanium rod is updated in real time after the adjustment.
Preferably, fitting the titanium rod key points through cubic spline interpolation to obtain a fitted titanium rod curve; the cubic Spline interpolation (Cubic Spline Interpolation) is simply Spline interpolation, which is a process of obtaining a curve function set by solving a three-bending moment equation set mathematically through a smooth curve of a series of shape value points, and the lower order interpolation has convergence and stability in view of the characteristics of non-convergence and instability of the higher order interpolation, so that the lower order interpolation has higher practical value, but the smoothness of the lower order interpolation is poorer, such as piecewise linear interpolation polynomials have continuity only in interpolation intervals, edges are formed at interpolation nodes, and the first derivative does not exist; the piecewise cubic Hermite interpolation polynomial has only a first derivative, i.e., first order smoothness, but no second order smoothness in the interpolation interval. Interpolation is similar, each known point must pass through, but the phenomenon of Dragon-Gerdostane appears in higher order, so that piecewise cubic spline interpolation is adopted.
S104, calculating the required transverse connection length based on the titanium rod fitting curve, and determining an adapted transverse connection model in a prosthesis database based on the required transverse connection length;
Specifically, as shown in fig. 17, the pre-operative planning system for spinal surgery automatically calculates the required transverse connection length according to the planning result of the titanium rod, and matches the optimal transverse connection model from the transverse connection prosthesis database. Simulating the effect of the implanted cross connection.
A first punctuation and a second punctuation are obtained at certain length positions of the left titanium rod and the right titanium rod, and the first punctuation and the second punctuation are connected through a straight line;
The straight line is translated up and down within the length range of the titanium rod, and in the translation process, the straight line is always connected with the left titanium rod and the right titanium rod;
The length interval of the straight line is the length range of the transverse connection, the maximum value of the length is m, the numerical value a=the transverse connection type (namely the length of the transverse connection) -m is calculated, and when a is more than or equal to 0 and the numerical value a is minimum, the optimal transverse connection type is obtained.
S105, calculating the length and the width of the two adjacent spinal interspaces, and determining the type of the adaptive fusion device in the prosthesis database based on the length and the width of the two adjacent spinal interspaces.
Specifically, as shown in fig. 15, the length and width of the intervertebral space between two adjacent vertebrae are automatically calculated, and the optimal fusion device model is intelligently matched by combining the medical theory and a prosthesis database. Simulating the effect after implantation of the fusion device.
Preprocessing the spine CT data to obtain a third preprocessed image, wherein the third preprocessed image only keeps a bone area and a background area; the DICOM data is converted into PNG image data. Traversing each PNG image, and binarizing each PNG image, wherein only bone regions and background regions exist in the PNG image.
Counting the area (namely the number of pixel points) of a skeleton region in each third preprocessing Image, and searching the third preprocessing Image (image_max) with the largest area; searching a communication Area in the third preprocessed image, namely using findContours functions of an OpenCV library, wherein the found Area is Area1, area2, … and AreaN, the Area is the minimum circumscribed rectangle of the communication Area, and the largest Area (pelvic Area) is removed, and only the vertebra Area is left at the moment; and the remaining regions are ordered by Y value.
Detecting edges in each region by using a Canny operator of an OpenCV library, and acquiring edge points based on the edges in each connected region; and calculating the length and the width of the intervertebral space between two adjacent vertebrae based on the edge points. The minimum point of the adjacent Area, namely the minimum distance between the identified points in the two areas; the maximum point of the adjacent Area, i.e. the distance between the two areas and the point near the center line.
The principle of calculating the length and the width of the intervertebral space and matching the optimal model is as follows:
length: identifying the front edge and the rear edge of the vertebral body, wherein the distance between the two points is the length of the intervertebral space;
Width: identifying the front edge and the rear edge of the vertebral body, and connecting two points by a straight line. At 1/2 of the straight line, a vertical line is drawn. Two points were taken where the vertical line intersects the vertebral body. The distance between the two points is the width of the intervertebral space.
Matching the best model: and calculating the value a=the length of the intervertebral space and the length of the fusion device, wherein when a is more than 0 and the value a is minimum, the optimal fusion device length is obtained. The calculated value b=intervertebral space width/2-cage width, and when b > 0 and the value b is minimum, the optimal cage width is obtained.
According to the artificial intelligence-based spine operation pre-operation planning method, the imported spine CT data are preprocessed to obtain a first preprocessed image; inputting the first preprocessing image into a trained Hourg l ass neural network model to obtain a screw feeding point and a screw feeding direction of a screw, calculating the diameter and the length of the screw based on the screw feeding point and the screw feeding direction, and determining an adaptive screw model in a prosthesis database based on the diameter and the length of the screw; calculating the width of a U-shaped structure at the tail of a screw, fitting a titanium rod curve, calculating the optimal titanium rod width based on the width of the U-shaped structure at the tail of the screw, and determining the type of the adaptive titanium rod in a prosthesis database based on the optimal titanium rod width; calculating the length and the width of the intervertebral space of two adjacent vertebrae, and determining the type of the adaptive fusion device in a prosthesis database based on the length and the width of the intervertebral space of the two adjacent vertebrae; calculating a required transverse connection length based on the titanium rod fitting curve, and determining an adapted transverse connection model in a prosthesis database based on the required transverse connection length; the problem of doctor is according to CT image among the prior art, planning precision is not high, planning inefficiency and unable suitable prosthesis model of matching is solved.
FIG. 3 is a flow chart of an embodiment of an artificial intelligence based pre-operative planning system for spinal surgery in accordance with the present invention; as shown in fig. 3, the spinal surgery preoperative planning system based on artificial intelligence provided by the embodiment of the invention comprises the following steps:
an importing verification module 10, configured to preprocess imported spine CT data to obtain a first preprocessed image;
Also used for: generating a two-dimensional cross-section image view, a two-dimensional coronal image view, a two-dimensional sagittal image view and a three-dimensional image view which are consistent in three-dimensional coordinate system based on the spine CT data; and acquiring surgical planning information of the CT data in response to the acquisition instruction, wherein the surgical planning information comprises disease type information, surgical type information and surgical segment information.
The screw planning module 20 is configured to input the first preprocessed image into a trained Hourglass neural network model, obtain a screw insertion point and a screw insertion direction of a screw, calculate a screw diameter and a screw length based on the screw insertion point and the screw insertion direction, and determine an adapted screw model in a prosthesis database based on the screw diameter and the screw length;
the screw planning module 20 is also for: constructing Hourg l ass a neural network model, wherein the Hourg l ass neural network comprises a convolution layer, a pooling layer, a replication layer of the convolution layer and a feature layer corresponding to the replication layer;
And adding up samples of the corresponding layers in the replication layer and the convolution layer to obtain the corresponding feature layer, overlapping the feature layers to obtain a large feature layer, generating a thermodynamic diagram representing the probability of the key point through 1x1 convolution, and taking the point with the maximum probability value in the thermodynamic diagram as a feature point, wherein the position of the feature point is the predicted nailing point position.
The titanium rod planning module 30 is used for calculating the width of the tail U-shaped structure of the screw, calculating the optimal titanium rod width based on the width of the tail U-shaped structure of the screw, and determining the adaptive titanium rod model in the prosthesis database based on the optimal titanium rod width; fitting the key points of the titanium rod through cubic spline interpolation to obtain a fitted titanium rod curve.
And the transverse connection planning module 40 is used for calculating the required transverse connection length based on the titanium rod fitting curve and determining the adaptive transverse connection model in the prosthesis database based on the required transverse connection length.
The cross-over planning module 40 is also configured to: a first punctuation and a second punctuation are obtained at certain length positions of the left titanium rod and the right titanium rod, and the first punctuation and the second punctuation are connected through a straight line; the straight line is translated up and down within the length range of the titanium rod, and in the translation process, the straight line is always connected with the left titanium rod and the right titanium rod; the length interval of the straight line is the length range of the transverse connection, the maximum value of the length is m, the numerical value a=the transverse connection type (namely the length of the transverse connection) -m is calculated, and when a is more than or equal to 0 and the numerical value a is minimum, the optimal transverse connection type is obtained.
A cage planning module 50 for calculating the length and width of the two adjacent spinal interspaces, determining an adapted cage model in the prosthesis database based on the length and width of the two adjacent spinal interspaces;
the cage planning module 50 is also configured to: preprocessing the spine CT data to obtain a third preprocessed image, wherein the third preprocessed image only keeps a bone area and a background area; counting the area of a bone region in each third preprocessed image, and searching the third preprocessed image with the largest area; searching for connected areas in the third preprocessing image, and acquiring edge points based on edges in each connected area; and calculating the length and the width of the intervertebral space between two adjacent vertebrae based on the edge points.
According to the artificial intelligence-based spinal surgery preoperative planning system, the imported spinal CT data is preprocessed through the importing confirmation module 10, and a first preprocessed image is obtained; inputting the first preprocessing image into a trained Hourglass neural network model through a screw planning module 20 to obtain a screw feeding point and a screw feeding direction of a screw, calculating the screw diameter and the screw length based on the screw feeding point and the screw feeding direction, and determining an adapted screw model in a prosthesis database based on the screw diameter and the screw length; calculating the width of a tail U-shaped structure of a screw through a titanium rod planning module 30, calculating the optimal titanium rod width based on the width of the tail U-shaped structure of the screw, and determining an adaptive titanium rod model in a prosthesis database based on the optimal titanium rod width; calculating a required cross-connection length based on the titanium rod fitting curve by a cross-connection planning module 40, and determining an adapted cross-connection model in a prosthesis database based on the required cross-connection length; the fusion cage planning module 50 calculates two adjacent spinal disc space lengths and widths, and determines an adapted fusion cage model based on the two adjacent spinal disc space lengths and widths in the prosthesis database. The problem of doctor is according to CT image among the prior art, planning precision is not high, planning inefficiency and unable suitable prosthesis model of matching is solved.
The present embodiment provides a non-transitory computer readable storage medium storing computer instructions that cause a computer to perform the methods provided by the above-described method embodiments, for example, including: preprocessing the imported spine CT data to obtain a first preprocessed image; inputting the first preprocessing image into a trained Hourglass neural network model to obtain a screw feeding point and a screw feeding direction of a screw, calculating the diameter and the length of the screw based on the screw feeding point and the screw feeding direction, and determining the type of the adapted screw in a prosthesis database based on the diameter and the length of the screw.
Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware associated with program instructions, where the foregoing program may be stored in a computer readable storage medium, and when executed, the program performs steps including the above method embodiments; and the aforementioned storage medium includes: various storage media such as ROM, RAM, magnetic or optical disks may store program code.
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 determined 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.
While the invention has been described in detail in the foregoing general description and specific examples, it will be apparent to those skilled in the art that modifications and improvements can be made thereto. Accordingly, such modifications or improvements may be made without departing from the spirit of the invention and are intended to be within the scope of the invention as claimed.