Disclosure of Invention
In view of the shortcomings of the prior art, the application provides a smoothing method, a smoothing device, electronic equipment and a storage medium of a vessel model, which are applied to the technical field of three-dimensional reconstruction of medical images, and the vessel model collapse problem caused by taubin network smoothing is solved by acquiring vessel voxel initial data VO, and then carrying out self-adaptive filtering on the vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data so as to thicken the tail end of a vessel and the radius of a tiny vessel voxel; in order to display the obtained vascular voxel self-adaptive filtering data on a display screen in a curve form, a voxel isosurface can be extracted according to the vascular voxel self-adaptive filtering data, and finally taubin network smoothing processing is carried out on the voxel isosurface to obtain vascular smooth image information. The method combines voxel self-adaptive filtering and taubin network smoothing for use together, performs smoothing operation on a three-dimensional reconstructed vascular tissue model, ensures the smoothing effect of the model, and simultaneously solves the problems of end collapse of a vascular system and disappearance of tiny blood vessels caused by the existing three-dimensional reconstruction method.
In a first aspect, the present application provides a method of smoothing a vessel model, the method comprising the steps of:
acquiring vessel voxel initial data VO;
performing self-adaptive filtering on the vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data;
extracting a voxel isosurface according to the vessel voxel self-adaptive filtering data;
and carrying out taubin network smoothing processing on the voxel isosurface to obtain vascular smooth image information.
Obtaining vessel voxel initial data VO by the smoothing method of the vessel model, wherein the vessel voxel initial data VO is voxel data of a target vessel; and then, in order to avoid the end collapse of the initial vascular voxel data VO after network smoothing, the initial vascular voxel data VO can be subjected to self-adaptive filtering operation to thicken the end of a blood vessel and the radius of a tiny blood vessel so as to obtain self-adaptive filtering data of the filtered vascular voxel, then, a voxel isosurface is extracted according to the self-adaptive filtering data of the vascular voxel, a model of a target vascular is displayed on a display, at the moment, the surface of the model is rough, and the model can be subjected to smoothing treatment by adopting taubin network smoothing, so that vascular smooth image information can be obtained. Therefore, the application adopts a mode of combining self-adaptive filtering and network smoothing to reconstruct the vascular tissue in three dimensions, and has the beneficial effects of ensuring the smoothing effect of the model and avoiding the collapse of the tail end of the model.
Preferably, in the smoothing method of a vessel model provided by the present application, the step of adaptively filtering the vessel voxel initial data VO to obtain vessel voxel adaptive filtering data includes:
carrying out normalized mean filtering on the vessel voxel initial data VO to obtain vessel voxel mean filtering data VM;
traversing the position of a voxel i in the vessel voxel initial data VO to obtain a mean value filtering data value of the same position in the vessel voxel mean value filtering data VM;
and carrying out self-adaptive filtering on the vessel voxel initial data VO according to the mean value filtering data value to obtain vessel voxel self-adaptive filtering data.
Through the smoothing method of the vessel model, the method for adaptively filtering the vessel voxel initial data VO carries out twice filtering on the vessel voxel initial data VO, namely, firstly carrying out normalized mean filtering on the vessel voxel initial data VO, and obtaining first filtered data: vessel voxel mean filter data VM; then by traversing the position of voxel i in VO, the mean value filtering data value corresponding to the vessel voxel mean value filtering data VM can be obtainedThe method comprises the steps of carrying out a first treatment on the surface of the By passing throughJudging whether the condition is satisfied>And carrying out self-adaptive filtering on the vessel voxel initial data VO of the position of the voxel i meeting the condition at the specific position of the corresponding voxel i, so as to obtain the self-adaptive filtering data of the vessel voxels of the enhanced vessel end and the tiny vessel.
Preferably, in the present application, there is provided a smoothing method of a vessel model, the step of performing normalized mean filtering on vessel voxel initial data VO to obtain vessel voxel mean filtering data VM includes:
obtaining a filter radius N of normalized mean value filtering;
and carrying out normalized mean filtering on the vessel voxel initial data VO according to the filter radius N to obtain vessel voxel mean filtering data VM.
Through the smoothing method of the vessel model, when the normalized mean value filtering is carried out on the vessel voxel initial data VO, the filtering radius N of the normalized mean value filtering needs to be obtained first, the selection of the filtering radius N needs to be determined according to the size of a target vessel, and the larger the target vessel is, the larger the value of N is, but under the normal condition, the filtering requirement of the application can be met by taking the value of N as 3.
Preferably, the present application provides a method of smoothing a vessel model, filtering data values based on an averageThe step of adaptively filtering the vessel voxel initial data VO to obtain vessel voxel adaptive filtering data comprises the following steps:
determining mean filtered data valuesIs of a size of (2);
obtaining the mean value filtering data value meeting the conditionCorresponding voxel->;
Voxel-to-voxel with filter radius NAnd carrying out self-adaptive filtering on the vessel voxel initial data VO in the range to obtain vessel voxel self-adaptive filtering data.
By the above smoothing method of a vessel model, since the voxels i respectively correspond to an initial data value before filtering and a mean value filtered data value after filteringTherefore, by judging that +.>Acquiring the corresponding meeting conditionVoxel->Then according to the voxels satisfying the condition +>Determining an initial data value before filtering meeting the condition and vessel voxel initial data VO corresponding to the initial data value, and performing filtering radius N on the voxel +.>The in-range vessel voxel initial data VO is adaptively filtered to thicken the vessel ends and the radius of the tiny vessels.
Preferably, the present application provides a method of smoothing a vessel model to filter a pair of voxels with a radius NThe self-adaptive filtering is carried out on the initial data VO of the vascular voxels in the range, and the step of obtaining self-adaptive filtering data of the vascular voxels comprises the following steps:
filtering data values according to the mean value satisfying the conditionCalculating voxel->Filtering probability of the vessel voxel initial data VO in the range;
vessel voxel adaptive filter data is calculated based on the filter probabilities.
Preferably, the present application provides a smoothing method of a vessel model, filtering data values according to an average value satisfying a conditionCalculating voxel->The specific calculation formula of the filtering probability of the vessel voxel initial data VO in the range is as follows:wherein->Mean value filtered data value for satisfying the condition +.>,/>Is the filtering probability.
Preferably, the present application provides a smoothing method of a vessel model, and a specific calculation formula for calculating self-adaptive filtering data of vessel voxels according to filtering probability is as follows:,wherein->For voxel->The filtering result of the in-range vessel voxel initial data VO,representing an arithmetic function rounding the filtering result,/->And (5) rounding the result to obtain the vessel voxel self-adaptive filtering data.
In a second aspect, the present application provides a smoothing device for a vascular model, the device comprising:
the acquisition module is used for: the method comprises the steps of acquiring vessel voxel initial data VO;
and a filtering module: the method comprises the steps of performing self-adaptive filtering on vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data;
and an extraction module: the voxel isosurface is extracted according to the vessel voxel self-adaptive filtering data;
and a smoothing module: and the method is used for carrying out taubin network smoothing on the voxel isosurface to obtain vessel smoothing image information.
In a third aspect, the application provides an electronic device comprising a processor and a memory storing computer readable instructions which, when executed by the processor, perform the steps of the method as provided in the first aspect above.
In a fourth aspect, the present application provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the method as provided in the first aspect above.
The beneficial effects are that: according to the smoothing method, the smoothing device, the electronic equipment and the storage medium of the vessel model, the vessel voxel initial data VO is obtained through the smoothing method of the vessel model, and the vessel voxel initial data VO is voxel data of a target vessel; and then, in order to avoid the end collapse of the initial vascular voxel data VO after network smoothing, the initial vascular voxel data VO can be subjected to self-adaptive filtering operation to thicken the end of a blood vessel and the radius of a tiny blood vessel so as to obtain self-adaptive filtering data of the filtered vascular voxel, then, a voxel isosurface is extracted according to the self-adaptive filtering data of the vascular voxel, a model of a target vascular is displayed on a display, at the moment, the surface of the model is rough, and the model can be subjected to smoothing treatment by adopting taubin network smoothing, so that vascular smooth image information can be obtained. Therefore, the application adopts a mode of combining self-adaptive filtering and network smoothing to reconstruct the vascular tissue in three dimensions, and has the beneficial effects of ensuring the smoothing effect of the model and avoiding the collapse of the tail end of the model.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the application, as presented in the figures, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by a person skilled in the art without making any inventive effort, are intended to be within the scope of the present application.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures. Meanwhile, in the description of the present application, the terms "first, second", etc. are used only for distinguishing the description, and are not to be construed as indicating or implying relative importance.
The following disclosure provides many different embodiments or examples for accomplishing the objectives of the present application and solving the problems of the prior art. In the current medical image three-dimensional reconstruction model technology, in order to smooth the surface of a three-dimensional model, a Laplacian smoothing method is generally adopted to smooth the reconstructed model, and the method can effectively smooth large tissues, but for vascular systems, a large number of vascular ends can be thinned or even disappear in the smoothing after the smoothing, and the phenomenon can cause wrong judgment of the radius of a blood vessel or the branch of the blood vessel during preoperative planning, so that the problem of inaccurate operation planning is caused. In order to solve the problem, the application provides a smoothing method, a smoothing device, an electronic device and a storage medium of a vessel model, which specifically comprises the following steps:
referring to fig. 1, an embodiment of the present application provides a smoothing method for a vascular model, which is applied to the technical field of three-dimensional reconstruction of medical images, and is beneficial to a more accurate operation planning before an operation by performing adaptive filtering on acquired vascular voxel initial data VO, performing voxel isosurface extraction modeling on the data, and performing smoothing operation on a model surface by adopting taubin network smoothing to obtain a vascular system model which can maintain a smoothing effect and has no collapse at the tail end.
The smoothing method of the vascular model of the embodiment of the application comprises the following steps:
a1: acquiring vessel voxel initial data VO;
a2: performing self-adaptive filtering on the vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data;
a3: extracting a voxel isosurface according to the vessel voxel self-adaptive filtering data;
a4: and carrying out taubin network smoothing processing on the voxel isosurface to obtain vascular smooth image information.
Wherein, the vessel voxel initial data VO actually means: voxel data obtained by scanning a target vessel needing modeling through a CT or MRI instrument. In some preferred embodiments, it is desirable to complete an automatic segmentation of the vasculature prior to step A1, which may include:
acquiring a preset number of liver venous period data;
labeling the data of a target tissue in the liver venous period data, and preparing a training set;
training the training set by adopting nnunet to obtain a training model;
inputting the target vessel to be modeled into the training model to complete the segmentation of the target vessel.
The preset number is set manually, and can be 100 sets, the target tissue in the data is marked to form a training set, the training set is input into nnunet to train the training set, wherein nnunet is a self-adaptive framework based on 2D and 3D original U-Nets, the framework can automatically adjust all super parameters according to the attribute of the given data set, and the whole process does not need manual intervention, in other words, the training set of the target tissue is input into a training model formed by training in nnunet, and specific data of the target tissue in the current vascular data input into the training model can be automatically deduced, so that the segmentation of the target tissue is completed. In practical application, by adopting the method, the initial data VO of the vascular voxels can be acquired more quickly, and the efficiency of three-dimensional modeling is improved.
In step A2, in practical application, the end of the vasculature is usually finer, and if the initial data VO of the vascular voxels is modeled and smoothed directly, the blood vessels with fine ends are likely to collapse or disappear, which has a great negative effect on preoperative planning of the operator. Thus, the vessel voxel initial data VO may be subjected to a filtering operation before modeling the vessel voxel initial data VO, the purpose of the filtering operation being to thicken the vessel ends and the small vessel voxel radii, so that the vessels at the vessel system ends remain clear after smoothing the vessel system model after modeling, and thus, in some preferred embodiments, the step of adaptively filtering the vessel voxel initial data VO to obtain vessel voxel adaptive filtered data comprises:
carrying out normalized mean filtering on the vessel voxel initial data VO to obtain vessel voxel mean filtering data VM;
traversing the position of voxel i in the vessel voxel initial data VO to obtain the mean value filtering data value of the same position in the vessel voxel mean value filtering data VM;
Filtering data values according to meanPulse alignmentAnd carrying out self-adaptive filtering on the tube voxel initial data VO to obtain vessel voxel self-adaptive filtering data.
Wherein, since the vessel voxel initial data VO is CT or MRI data acquired based on CT or MRI scanning instrument, the data is displayed on a display screen in the form of image, the image forming process is divided into a plurality of cuboids with the same volume as the selected layer, namely, voxels, so that the position of the voxel i represents the relative position of the voxel i in the whole image, and the mean value of the mean value filtering data of the same position in the vessel voxel mean value filtering data VMRepresenting the mean filtered data value of voxel i after filtering the vessel voxel initial data VO +.>。
The method comprises the steps of firstly carrying out normalized mean filtering on vessel voxel initial data VO to carry out initial smoothing on the vessel system so as to blur out the vessel voxel mean filtering data VM in the observation range, and then traversing voxel i in the vessel voxel initial data VO to obtain the initial data value of voxel i, wherein the vessel voxel initial data VO can be normalized mean filtered to obtain the vessel voxel mean filtering data VM in the observation range, and the vessel voxel mean filtering data VM in the observation range can be obtainedWherein the vessel voxel initial data VO can be understood as a data set, so that the voxel i can be understood as a pointer of the data set, a specific data value corresponding to the voxel i can be obtained through the pointer, and the initial data value +.>The method comprises the steps of carrying out a first treatment on the surface of the While the vessel voxel mean filter data VM is the vessel voxel initial data VOThe obtained data set after filtering is that VM has the same pointer as VO, in other words, voxel i can indicate two groups of data, namely, the initial data VO of the vascular voxels before filtering and the mean value filtering data VM of the vascular voxels after filtering, so that after traversing the VO, the mean value filtering data value (I) of the position corresponding to the same position in the VM data set can be obtained through the position of voxel i in the VO>Thus, the relationship between the above parameters is: VO is the data set before voxel i filtering; VM is voxel i filtered dataset, < +.>Data value for each specific voxel in VM, < >>For the data value of each specific voxel in VO, -, is given>And->One-to-one correspondence. Wherein the initial data value before filtering +.>Generally 1, after normalized mean filtering, mean filtered data values +.>Typically between 0 and 1.
In some preferred embodiments, when the normalized mean value filtering is performed on the vessel voxel initial data VO, a specific value of the filtering radius N needs to be determined, and since the mean value filtering can blur data of a fine portion of information to be processed, if the value of the filtering radius N of the mean value filtering is larger, the blurring range will also increase accordingly, and in the present application, that is, as the image of the target vasculature is larger, the value of the filtering radius N should also increase, where the value of the filtering radius N is set in advance by a technician, but in actual operation, in order to avoid the mean value filtering, the filtering radius N is set to 3, so as to meet the filtering requirement of the present application.
Wherein, in some preferred embodiments, the data values are filtered according to an averageThe step of adaptively filtering the vessel voxel initial data VO to obtain vessel voxel adaptive filtering data comprises the following steps:
determining mean filtered data valuesIs of a size of (2);
when (when)At this time, the mean value filter data value satisfying the condition is acquired +.>Corresponding voxel->;
Voxel-to-voxel with filter radius NAnd carrying out self-adaptive filtering on the vessel voxel initial data VO in the range to obtain vessel voxel self-adaptive filtering data.
After the normalized mean value filtering is performed on the target vascular tissue, the target vascular tissue data needs to be screened more precisely, and a blood vessel range needing to be considered in preoperative planning and a blood vessel range not needing to be considered are selected, so that the interference blood vessel is abandoned in final modeling, and an operator can observe the three-dimensional model more clearly. Thus, the filtered average value can be used for filtering the data valueMake a judgment when->(N is the filter radius, and is usually 3), that is, the vessel is too thin, and is far from the center of the target vessel, no observation is needed, so the +.>Corresponding vessel voxel data; when->When the radius of the part of the vessel is larger, the vessel cannot collapse even though the vessel is subjected to smoothing operation; while whenWhen the vessel is in the range to be observed, but after the subsequent smoothing operation, the vessel may collapse or disappear, and the vessel to be thickened is the vessel, so after judging the vessel meeting the thickening condition, the mean value filtering data value ∈meeting the condition is obtained>Corresponding voxel->I.e. finding the range of positions of the vessel to be thickened, for voxel +.>And carrying out self-adaptive filtering on the vessel voxel initial data VO in the range, and thickening the vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data.
Wherein, in some preferred embodiments, voxels are paired with a filter radius NThe self-adaptive filtering is carried out on the initial data VO of the vascular voxels in the range, and the step of obtaining self-adaptive filtering data of the vascular voxels comprises the following steps:
filtering data values according to the mean value satisfying the conditionCalculating voxel->Filtering probability of the vessel voxel initial data VO in the range;
vessel voxel adaptive filter data is calculated based on the filter probabilities.
Wherein the voxels areBlood vessels, voxels, indicating need for enhancement>Range represents voxel->At voxel +.>When the initial data VO of the vascular voxels is filtered within the range of (1) by the filter radius N, in order to ensure the filter effect, fine blood vessels in the target vascular are better enhanced, and the mean value filter data value of the vascular which can be enhanced according to the need is +.>And calculating the filtering probability of the vessel voxel initial data VO, and then calculating to obtain the reinforced vessel voxel self-adaptive filtering data according to the filtering probability.
Wherein, in some preferred embodiments, voxels are computedThe specific formula of the filtering probability of the vessel voxel initial data VO in the range is as follows: />Wherein->Filtering data values to meet a condition,/>Is the filtering probability.
Wherein, in some preferred embodiments, the specific calculation formula for calculating the vessel voxel adaptive filtering data according to the filtering probability is as follows:,/>whereinFor voxel->Filtering result of the in-range vessel voxel initial data VO, a +.>Representing an arithmetic function rounding the filtering result,/->The data is adaptively filtered for vessel voxels. And rounding the filtering result to obtain the vascular voxel self-adaptive filtering data.
In step A3, voxel iso-surface extraction, that is, three-dimensional modeling is performed on the vessel voxel adaptive filtering data obtained in step A2 to obtain a three-dimensional model of the target vessel, and the model surface at this time is rough, so in step A4, a taubin network smoothing method is adopted to perform smoothing operation on the three-dimensional model established in step A3 to obtain vessel smoothing image information, so that an operator can accurately plan before operation through the vessel smoothing image information.
From the above, the present application provides a smoothing method of a vessel model, by obtaining vessel voxel initial data VO, where the vessel voxel initial data VO is voxel data of a target vessel; and then, in order to avoid the end collapse of the initial vascular voxel data VO after network smoothing, the initial vascular voxel data VO can be subjected to self-adaptive filtering operation to thicken the end of a blood vessel and the radius of a tiny blood vessel so as to obtain self-adaptive filtering data of the filtered vascular voxel, then, a voxel isosurface is extracted according to the self-adaptive filtering data of the vascular voxel, a model of a target vascular is displayed on a display, at the moment, the surface of the model is rough, and the model can be subjected to smoothing treatment by adopting taubin network smoothing, so that vascular smooth image information can be obtained. Therefore, the application adopts a mode of combining self-adaptive filtering and network smoothing to reconstruct the vascular tissue in three dimensions, and has the beneficial effects of ensuring the smoothing effect of the model and avoiding the collapse of the tail end of the model.
Referring to fig. 2, the present application provides a smoothing device for a vascular model, the device comprising:
the acquisition module 201: the method comprises the steps of acquiring vessel voxel initial data VO;
the filtering module 202: the method comprises the steps of performing self-adaptive filtering on vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data;
extraction module 203: the voxel isosurface is extracted according to the vessel voxel self-adaptive filtering data;
smoothing module 204: and the method is used for carrying out taubin network smoothing on the voxel isosurface to obtain vessel smoothing image information.
In practical applications, the acquisition module 201 may be a CT scanning instrument or an MRI scanning instrument, and after acquiring CT data or MRI data scanned by the foregoing instrument, the data is put into a model trained in advance, so as to obtain the vessel voxel initial data VO. Wherein, before acquiring the vessel voxel initial data VO, the automatic segmentation of the vessel system needs to be completed, and the automatic segmentation step may include:
acquiring a preset number of liver venous period data;
labeling the data of a target tissue in the liver venous period data, and preparing a training set;
training the training set by adopting nnunet to obtain a training model;
inputting the target vessel to be modeled into the training model to complete the segmentation of the target vessel.
The preset number is set manually, and can be 100 sets, the target tissue in the data is marked to form a training set, the training set is input into nnunet to train the training set, wherein nnunet is a self-adaptive framework based on 2D and 3D original U-Nets, the framework can automatically adjust all super parameters according to the attribute of the given data set, and the whole process does not need manual intervention, in other words, the training set of the target tissue is input into a training model formed by training in nnunet, and specific data of the target tissue in the current vascular data input into the training model can be automatically deduced, so that the segmentation of the target tissue is completed. In practical application, the acquisition module 201 can acquire the initial data VO of the vascular voxels faster by adopting the method, so that the efficiency of three-dimensional modeling is improved.
In practical application, the filtering module 202 filters the vessel voxel initial data VO acquired by the acquiring module 201 to obtain vessel voxel self-adaptive filtering data, wherein the specific filtering mode is as follows: firstly, carrying out normalized mean value filtering on vessel voxel initial data VO, carrying out initial smoothing on a vessel system to blur out tiny blood vessels far away from a target center, avoiding interference of the tiny blood vessels far away from the target center on three-dimensional modeling to obtain vessel voxel mean value filtering data VM in an observation range, and then traversing voxels i in the vessel voxel initial data VO to obtain an initial data value of the voxels iWherein, the initial data VO of the vascular voxel can be understood as a data set, so that the voxel i is a pointer of the data set, a specific data value corresponding to the voxel i can be obtained through the pointer, and all data values of the VO are combined, thus obtaining the initial data value of the initial data VO of the vascular voxel>The method comprises the steps of carrying out a first treatment on the surface of the The vessel voxel mean value filtering data VM is a data set obtained by filtering the vessel voxel initial data VO, so VM and VO have the same fingerIn other words, the voxel i may indicate two sets of data, i.e. the pre-filtered vessel voxel initial data VO and the filtered vessel voxel mean value filtered data VM, respectively, so that after traversing the VO, the mean value filtered data value +_ at the position corresponding to the same position in the VM dataset can be obtained from the position of the voxel i in the VO>VO is the data set before voxel i filtering; VM is voxel i filtered dataset, < +.>Data value for each specific voxel in VM, < >>For the data value of each specific voxel in VO, -, is given>And->And a pair of corresponding. Wherein the initial data value before filtering +.>Generally 1, after normalized mean filtering, mean filtered data values +.>Typically between 0 and 1.
In some preferred embodiments, the filtering module 202 needs to determine a specific value of the filter radius N when performing normalized mean filtering on the vessel voxel initial data VO, and because the mean filtering obscures data of a fine portion of the information to be processed, if the filter radius N of the mean filtering is larger, the blurring range will also increase accordingly, and in the present application, that is, as the image of the target vasculature is larger, the value of the filter radius N should also increase, where the value of the filter radius N is set in advance by a technician, but in actual operation, in order to avoid the mean filtering from excessively smoothing the vessel voxel initial data VO, the filter radius N is set to 3, so as to meet the filtering requirement of the present application.
In some preferred embodiments, after the filtering module 202 performs the normalized mean filtering on the target vascular tissue, it is further required to screen the target vascular tissue data more precisely, and select the range of the blood vessels that need to be considered in the preoperative planning, and the range of the blood vessels that need not to be considered, so as to discard the interfering blood vessels in the final modeling, so that the operator can observe the three-dimensional model more clearly. Thus, the filtered average value can be used for filtering the data valueMake a judgment when->(N is the filter radius, and is usually 3), that is, the vessel is too thin, and is far from the center of the target vessel, no observation is needed, so the +.>Corresponding vessel voxel data; when->When the radius of the part of the vessel is larger, the vessel cannot collapse even though the vessel is subjected to smoothing operation; while->When the vessel is in the range to be observed, but after the subsequent smoothing operation, the vessel may collapse or disappear, and the application needs to thicken, namely the vessel is the part, so after judging the vessel meeting the thickening condition, the mean value filtering data value meeting the condition is obtainedCorresponding voxel->I.e. finding the range of positions of the vessel to be thickened, for voxel +.>And carrying out self-adaptive filtering on the vessel voxel initial data VO in the range, and thickening the vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data.
In some preferred embodiments, in order to ensure the filtering effect, the fine blood vessels in the target vessel are better enhanced, and the average value of the vessels which can be enhanced according to needs is used for filtering the data valueAnd calculating the filtering probability of the vessel voxel initial data VO, and then calculating to obtain the reinforced vessel voxel self-adaptive filtering data according to the filtering probability. The specific calculation formula is as follows: />Wherein->Mean value filtered data value for satisfying the condition +.>,/>Is the filtering probability;
,/>wherein->For voxel->Filtering result of the in-range vessel voxel initial data VO, a +.>Representing an arithmetic function that rounds the filtering result,/>the data is adaptively filtered for vessel voxels.
In some specific embodiments, the extracting module 203 extracts the voxel iso-surface from the voxel adaptive filtering data of the vessel obtained in the filtering module 202, that is, three-dimensional modeling is performed to obtain a three-dimensional model of the target vessel, and the model surface at this time is rough, so that the smoothing module 204 performs smoothing operation on the three-dimensional model established in the step A3 by adopting a taubin network smoothing method to obtain vessel smoothing image information, so that an operator can accurately plan before operation through the vessel smoothing image information.
As can be seen from the above, the present application provides a smoothing device for a vessel model, by obtaining vessel voxel initial data VO, where the vessel voxel initial data VO is voxel data of a target vessel; and then, in order to avoid the end collapse of the initial vascular voxel data VO after network smoothing, the initial vascular voxel data VO can be subjected to self-adaptive filtering operation to thicken the end of a blood vessel and the radius of a tiny blood vessel so as to obtain self-adaptive filtering data of the filtered vascular voxel, then, a voxel isosurface is extracted according to the self-adaptive filtering data of the vascular voxel, a model of a target vascular is displayed on a display, at the moment, the surface of the model is rough, and the model can be subjected to smoothing treatment by adopting taubin network smoothing, so that vascular smooth image information can be obtained. Therefore, the application adopts a mode of combining self-adaptive filtering and network smoothing to reconstruct the vascular tissue in three dimensions, and has the beneficial effects of ensuring the smoothing effect of the model and avoiding the collapse of the tail end of the model.
Referring to fig. 4, a left image is a three-dimensional model reconstructed by the method of the present application and a right image is a three-dimensional model reconstructed by the method of the present application, wherein the vessels reconstructed by the left and right images are the same vessel, and the positions indicated by arrows in the images are the vessel ends of the three-dimensional model after smoothing, so that the vessel ends of the smoothing method of the present application can be clearly distinguished from the figures, which is more clearly visible than the vessel ends of the smoothing method of the present application, thereby facilitating the operator to plan the operation more precisely.
Referring to fig. 3, fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application, and the present application provides an electronic device 3, including: processor 301 and memory 302, the processor 301 and memory 302 being interconnected and in communication with each other by a communication bus 303 and/or other form of connection mechanism (not shown), the memory 302 storing computer readable instructions executable by the processor 301, which when executed by an electronic device, the processor 301 executes the computer readable instructions to perform the methods in any of the alternative implementations of the above embodiments to perform the functions of: acquiring vessel voxel initial data VO; performing self-adaptive filtering on the vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data; extracting a voxel isosurface according to the vessel voxel self-adaptive filtering data; and carrying out taubin network smoothing processing on the voxel isosurface to obtain vascular smooth image information.
An embodiment of the present application provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs a method in any of the alternative implementations of the above embodiments to implement the following functions: acquiring vessel voxel initial data VO; performing self-adaptive filtering on the vessel voxel initial data VO to obtain vessel voxel self-adaptive filtering data; extracting a voxel isosurface according to the vessel voxel self-adaptive filtering data; and carrying out taubin network smoothing processing on the voxel isosurface to obtain vascular smooth image information.
The computer readable storage medium may be implemented by any type or combination of volatile or non-volatile Memory devices, such as static random access Memory (Static Random Access Memory, SRAM for short), electrically erasable Programmable Read-Only Memory (Electrically Erasable Programmable Read-Only Memory, EEPROM for short), erasable Programmable Read-Only Memory (Erasable Programmable Read Only Memory, EPROM for short), programmable Read-Only Memory (ROM for short), magnetic Memory, flash Memory, magnetic disk, or optical disk.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be other manners of division in actual implementation, and for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.
Further, the units described as separate units may or may not be physically separate, and units displayed as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
Furthermore, functional modules in various embodiments of the present application may be integrated together to form a single portion, or each module may exist alone, or two or more modules may be integrated to form a single portion.
In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and variations will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.