CN111145289A - Extraction method and device of pelvis three-dimensional data - Google Patents
Extraction method and device of pelvis three-dimensional data Download PDFInfo
- Publication number
- CN111145289A CN111145289A CN201911404207.4A CN201911404207A CN111145289A CN 111145289 A CN111145289 A CN 111145289A CN 201911404207 A CN201911404207 A CN 201911404207A CN 111145289 A CN111145289 A CN 111145289A
- Authority
- CN
- China
- Prior art keywords
- data
- dimensional
- dimensional data
- transformation
- pelvis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
- G06T11/008—Specific post-processing after tomographic reconstruction, e.g. voxelisation, metal artifact correction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/64—Three-dimensional objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30008—Bone
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/467—Encoded features or binary features, e.g. local binary patterns [LBP]
Abstract
The invention discloses a method and a device for extracting pelvis three-dimensional data. Wherein, the method comprises the following steps: acquiring tomography medical image CT data of pelvis, wherein the CT data is data to be subjected to three-dimensional data extraction; processing the CT data to obtain a three-dimensional data extraction model; inputting the CT data into a three-dimensional data extraction model; acquiring an output result of the three-dimensional data extraction model; and converting the output result into three-dimensional data corresponding to the CT data. The invention solves the technical problem that the accuracy of extracted pelvis three-dimensional data is lower by removing the metal artifact in the CT data through manual operation when the pelvis three-dimensional data is extracted in the related technology.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a method and a device for extracting pelvis three-dimensional data.
Background
As patients who undergo total hip arthroplasty continue to grow in age and exercise capacity, the number of hip revision surgeries continues to increase. The failure of the primary total hip replacement requires revision of the hip for many reasons, including infection, wearing of the polyethylene lining of the prosthesis, and habitual dislocation of the prosthesis. The dissolution of bone around the prosthesis caused by the abrasion debris of the prosthesis and the aseptic loosening of the prosthesis are always the main reasons for the hip revision surgery. In the face of revision of hip joint after primary hip joint replacement failure, the strategy and the operation technology of the acetabulum side defect repair and reconstruction operation are one of the challenges for orthopedists.
The extraction of the pelvis three-dimensional data aims to provide more specific anatomical information for an operating doctor, preliminarily judges how much bone mass possibly remains in an operating patient and provides powerful support for diagnosis, treatment, operation and the like of the doctor.
However, the medical image visualization technology has made great progress in the three-dimensional reconstruction of human bones, and doctors can perform three-dimensional reconstruction on CT tomography medical image data through three-dimensional image generation and editing processing software to obtain a three-dimensional model. The conventional processing software segments bones based on image threshold values, metal artifacts exist in CT data of a hip revision surgery patient, the metal artifacts cannot be selectively removed through the threshold values, the metal artifacts can only be manually edited layer by a user, and the accuracy and the extraction efficiency of extracting the pelvis three-dimensional data are influenced by manual operation.
In view of the above problems, no effective solution has been proposed.
Disclosure of Invention
The embodiment of the invention provides a method and a device for extracting pelvis three-dimensional data, which at least solve the technical problem that when the pelvis three-dimensional data is extracted in the related art, metal artifacts in CT data are removed by manual operation, so that the accuracy of the extracted pelvis three-dimensional data is lower.
According to an aspect of an embodiment of the present invention, there is provided a method for extracting pelvis three-dimensional data, including: acquiring tomography medical image CT data of a pelvis, wherein the CT data is data to be subjected to three-dimensional data extraction; processing the CT data to obtain a three-dimensional data extraction model; inputting the CT data into the three-dimensional data extraction model; acquiring an output result of the three-dimensional data extraction model; and converting the output result into three-dimensional data corresponding to the CT data.
Optionally, processing the CT data comprises: obtaining two-dimensional medical image data based on the CT data; and analyzing the CT data to obtain the metal artifact degree of the CT data and identifying the metal artifact degree.
Optionally, processing the CT data comprises: performing a predetermined transformation on the CT data to obtain the transformed CT data, wherein the predetermined transformation includes at least one of the following modes: rotation transformation, translation transformation and noise addition transformation; extracting Local Binary Pattern (LBP) characteristics I of the CT data before transformation and the CT data after transformation; and performing dimension reduction processing on the LBP feature I to obtain the LBP feature I after dimension reduction.
Optionally, the extracting local binary pattern LBP features of the CT data before transformation and the CT data after transformation, which includes: setting a first pixel size of the CT data and a second pixel size of a sub-block of the CT data, wherein the sub-block is a sub-block with a preset size obtained by dividing the CT data; dividing the CT data into two pairs according to the pixel size of the sub-blocks of the CT data to obtain a preset number of sub-blocks; determining a histogram of each subblock in the preset number of subblocks, and performing normalization processing on the histogram of each subblock to obtain a statistical histogram of each subblock; and connecting the statistical histograms of each sub-block to obtain a first LBP characteristic of the CT data.
Optionally, performing dimension reduction processing on the LBP feature one to obtain the LBP feature one after dimension reduction includes: and performing dimensionality reduction on the LBP feature I by utilizing a Principal Component Analysis (PCA) technology to obtain the LBP feature I after dimensionality reduction.
Optionally, processing the CT data comprises: extracting depth features II of the CT data before transformation and the CT data after transformation by using a convolutional neural network; and fusing the LBP feature I subjected to dimensionality reduction and the depth feature II to obtain a fused feature III.
Optionally, the processing the CT data to obtain a three-dimensional data extraction model includes: inputting the three fusion characteristics into a full connection layer with preset dimensionality, and classifying by using a preset classifier to obtain the three-dimensional data extraction model.
According to another aspect of the embodiments of the present invention, there is also provided an apparatus for extracting three-dimensional data of a pelvis, including: the device comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring tomography medical image CT data of pelvis, and the CT data is data to be subjected to three-dimensional data extraction; the processing unit is used for processing the CT data to obtain a three-dimensional data extraction model; an input unit for inputting the CT data to the three-dimensional data extraction model; a second obtaining unit configured to obtain an output result of the three-dimensional data extraction model; and the conversion unit is used for converting the output result into three-dimensional data corresponding to the CT data.
Optionally, the processing unit comprises: an acquisition module for obtaining two-dimensional medical image data based on the CT data; and the analysis module is used for analyzing the CT data to obtain the metal artifact degree of the CT data and identifying the metal artifact degree.
Optionally, the processing unit comprises: a transformation module, configured to perform a predetermined transformation on the CT data to obtain the transformed CT data, where the predetermined transformation includes at least one of the following ways: rotation transformation, translation transformation and noise addition transformation; the first extraction module is used for extracting Local Binary Pattern (LBP) characteristics I of the CT data before transformation and the CT data after transformation; and the dimension reduction module is used for carrying out dimension reduction processing on the LBP feature I to obtain the LBP feature I after dimension reduction.
Optionally, the first extraction module includes: the setting submodule is used for setting a first pixel size of the CT data and a second pixel size of a sub-block of the CT data, wherein the sub-block is a sub-block with a preset size obtained by dividing the CT data; the dividing submodule is used for dividing the CT data into two pairs according to the pixel size of the sub-blocks of the CT data to obtain a preset number of sub-blocks; the determining submodule is used for determining the histogram of each subblock in the preset number of subblocks and normalizing the histogram of each subblock to obtain a statistical histogram of each subblock; and the connecting sub-module is used for connecting the statistical histograms of all the sub-blocks to obtain a first LBP characteristic of the CT data.
Optionally, the dimension reduction module includes: and the dimension reduction submodule is used for carrying out dimension reduction processing on the first LBP characteristic by utilizing a Principal Component Analysis (PCA) technology to obtain the first LBP characteristic after the dimension reduction processing.
Optionally, the processing unit comprises: the second extraction module is used for extracting depth features II of the CT data before transformation and the CT data after transformation by using a convolutional neural network; and the fusion module is used for fusing the LBP feature I subjected to dimension reduction with the depth feature II to obtain a fusion feature III.
Optionally, the processing unit comprises: and the input module is used for inputting the three fusion characteristics into a full connection layer with preset dimensionality and classifying by using a preset classifier to obtain the three-dimensional data extraction model.
According to another aspect of the embodiments of the present invention, there is also provided a storage medium including a stored program, wherein the program executes the method for extracting three-dimensional pelvic data according to any one of the above.
According to another aspect of the embodiment of the present invention, there is further provided a processor, configured to execute a program, where the program executes the method for extracting three-dimensional pelvic data according to any one of the above.
In the embodiment of the invention, CT data of tomography medical images of pelvis are acquired, wherein the CT data are data to be subjected to three-dimensional data extraction; processing the CT data to obtain a three-dimensional data extraction model; inputting the CT data into a three-dimensional data extraction model; acquiring an output result of the three-dimensional data extraction model; the output result is converted into three-dimensional data corresponding to the CT data, the three-dimensional data corresponding to the CT data is obtained by establishing a three-dimensional data extraction model through the pelvis three-dimensional data extraction method provided by the embodiment of the invention, the technical effect of improving the accuracy in extracting the pelvis three-dimensional data is achieved, and the technical problem that in the related technology, when the pelvis three-dimensional data is extracted, metal artifacts in the CT data are removed through manual operation, so that the accuracy of the extracted pelvis three-dimensional data is low is solved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention. In the drawings:
fig. 1 is a flowchart of an extraction method of pelvic three-dimensional data according to an embodiment of the present invention;
FIG. 2 is a flowchart of an alternative method for extracting three-dimensional pelvic data according to an embodiment of the invention;
fig. 3 is a schematic diagram of an apparatus for extracting three-dimensional pelvic data according to a second embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
For convenience of description, the following provides a detailed description of the terms or phrases in the embodiments of the present invention.
Local Binary Patterns (LBP for short): that is, the linear reflection projection method is a simple imaging algorithm, which accumulates all the projection rays passing through a certain point and then estimates the density value of the point in reverse.
Digital Imaging And Communication In Medicine (DICOM): is an international standard for medical images and related information defining a medical image format that can be used for data exchange with a quality that meets clinical needs.
Principal component Analysis (Principal Components Analysis, PCA for short): the idea of reducing dimensions is utilized to convert multiple indexes into a few comprehensive indexes.
Example one
In accordance with an embodiment of the present invention, there is provided a method embodiment of a method for extracting three-dimensional data of a pelvis, it is noted that the steps illustrated in the flowchart of the drawings may be performed in a computer system such as a set of computer executable instructions, and that while a logical order is illustrated in the flowchart, in some cases the steps illustrated or described may be performed in an order different than here.
Fig. 1 is a flowchart of an extraction method of pelvic three-dimensional data according to an embodiment of the present invention, as shown in fig. 1, the extraction method of pelvic three-dimensional data includes the following steps:
step S102, CT data of tomography medical images of the pelvis are obtained, wherein the CT data are data to be subjected to three-dimensional data extraction.
And step S104, processing the CT data to obtain a three-dimensional data extraction model.
Optionally, a CT tomography medical image of the pelvis to be extracted is obtained, two-dimensional medical image DICOM data is obtained, and the degree of metal artifacts of the data is determined and marked.
In an alternative embodiment, processing the CT data may include: obtaining two-dimensional medical image data based on the CT data; and analyzing the CT data to obtain the metal artifact degree of the CT data and identifying the metal artifact degree.
Step S106, inputting the CT data into the three-dimensional data extraction model.
And step S108, acquiring an output result of the three-dimensional data extraction model.
Step S110, converting the output result into three-dimensional data corresponding to the CT data.
From the above, CT data of a tomography medical image of the pelvis can be obtained, wherein the CT data is data to be subjected to three-dimensional data extraction; processing the CT data to obtain a three-dimensional data extraction model; inputting the CT data into a three-dimensional data extraction model; acquiring an output result of the three-dimensional data extraction model; and converting the output result into three-dimensional data corresponding to the CT data, thereby achieving the purpose of obtaining the three-dimensional data corresponding to the CT data by establishing a three-dimensional data extraction model.
It is easy to note that since the output of the three-dimensional data extraction model can be obtained by inputting the acquired CT data into the three-dimensional data extraction model, the three-dimensional data corresponding to the CT data is obtained by establishing the three-dimensional data extraction model, and the technical effect of improving the accuracy of extracting the pelvis three-dimensional data is achieved.
Therefore, the method for extracting the pelvis three-dimensional data provided by the embodiment of the invention solves the technical problem that the accuracy of the extracted pelvis three-dimensional data is lower by removing the metal artifact in the CT data through manual operation when the pelvis three-dimensional data is extracted in the related technology.
According to the above embodiment of the present invention, in step S104, the processing the CT data may include: performing a predetermined transformation on the CT data to obtain transformed CT data, wherein the predetermined transformation comprises at least one of the following modes: rotation transformation, translation transformation and noise addition transformation; extracting local binary pattern LBP characteristics I of the CT data before transformation and the CT data after transformation; and performing dimension reduction processing on the LBP feature I to obtain the LBP feature I after dimension reduction.
Optionally, the acquired CT data is changed, and the CT data of the pelvis before and after the change are both used as training data, so that the training data volume is increased, overfitting is reduced, and the generalization capability of the model is improved, specifically: and respectively carrying out rotation, translation, noise adding operation and the like on the acquired CT data.
In an alternative embodiment, extracting the local binary pattern LBP features of the pre-transformed CT data and the transformed CT data includes: setting a first pixel size of CT data and a second pixel size of a sub-block of the CT data, wherein the sub-block is a sub-block with a preset size obtained by dividing the CT data; dividing two pairs of CT data according to the pixel size of the sub-blocks of the CT data to obtain a preset number of sub-blocks; determining a histogram of each subblock in a preset number of subblocks, and normalizing the histogram of each subblock to obtain a statistical histogram of each subblock; and connecting the statistical histograms of each sub-block to obtain a first LBP characteristic of the CT data.
For example, extracting CT data of the pelvis before and after the change to obtain an LBP feature a (i.e., LBP feature one), and performing dimension reduction processing on the LBP feature a, specifically: setting the pixel size of CT data as m multiplied by m and the size of the subblock as n multiplied by n, wherein m and n are positive integers and m can be divided by n; dividing each pelvis CT data into m according to the set sub-block size2/n2Sub-blocks of the same size, for one pixel in each sub-block, using
The operator compares the gray values of the pixels of the adjacent p sub-blocks with it,
wherein R is the radius, P is the number of sampling points, and the cycle isIf the gray value of the pixel of the surrounding sub-block is larger than that of the pixel of the sub-block, the position of the pixel point of the sub-block is marked as 1, otherwise, the position is 0; thus, in the circular neighborhood taking R as the radius, after the gray values of the pixels of the central subblock and the surrounding p subblocks are compared, a p-bit binary number is generated, and the LBP value of the window central pixel point of the central subblock is obtained. As shown in equation (1):
wherein (x)c，yc) Is the window center pixel point, icIs the gray value of the pixel of the central sub-block, ipIs the gray value of the pixels of the neighboring sub-blocks, s is a sign function, as shown in (2):
then, calculating a histogram of each sub-block, then carrying out normalization processing on the histogram, and finally connecting the obtained statistical histograms of each sub-block into an LBP feature a of each pelvic CT data to finally obtain the LBP feature a with dimensions of 2 Xn 2 Xm/n.
In the embodiment of the present invention, performing dimension reduction on the LBP feature one to obtain the dimension-reduced LBP feature one includes: and performing dimensionality reduction on the LBP feature I by utilizing a Principal Component Analysis (PCA) technology to obtain the LBP feature I after dimensionality reduction.
For example, dimension reduction processing can be performed by PCA to obtain LBP feature a of h dimension. The PCA method is to form new variables by linear projection of the original variables, and the principal components of the features are generally calculated by formula (3): y is equal to UT(xi-x) (3), wherein y represents principal component features, x represents a feature mean of the training samples, xiFor features requiring dimension reduction, UTThe formula is calculated for the covariance matrix as shown in formula (4):
according to the above embodiment of the present invention, processing the CT data may include: extracting depth features II of the CT data before transformation and the CT data after transformation by using a convolutional neural network; and fusing the LBP feature I subjected to dimensionality reduction and the depth feature II to obtain a fused feature III.
For example, a convolutional neural network is used to extract a depth feature b (i.e., a depth feature two) in the CT data of the pelvis before and after the change, specifically: adopting an vgg-16 model, carrying out a series of convolution pooling operations, and finally obtaining a depth feature b through a full-connection layer; for convolutional layers, the output characteristics of each convolutional layer are obtained by convolving a set of M1 × M2 filters with the output characteristics of the previous convolutional layer, and the output formula of the convolution operation is shown in formula (5):
wherein, YjOutput characteristics, X, obtained for the jth convolutional layeriRepresenting input features of convolutional layers, Wj，iIs a weight matrix of M1 XM 2 filters, bjFor the bias of the j-th layer,
representing convolution operation, and N is the number of all or part of the features of the convolution layer in the previous layer. For the pooling layer, obtaining corresponding output characteristics by performing maximum sampling on each characteristic map of the convolution layer of the previous layer; for a fully-connected layer, each neuron in the fully-connected layer connects all neurons in the previous pooling layer, resulting in a depth feature b. Fusing the LBP feature a and the depth feature b after dimensionality reduction to obtain a fused feature c (namely, a fused feature III), wherein the formula is as follows: c ═ a1，b1；a2，b2；a3，b3；…ap，bp](6) Where p is the total number of CT data for the trained pelvis.
In an alternative embodiment, the processing the CT data to obtain the three-dimensional data extraction model includes: inputting the three fusion characteristics into a full connection layer with preset dimensionality, and classifying by using a preset classifier to obtain a three-dimensional data extraction model.
In an alternative embodiment, in step S110, the output result is converted into three-dimensional data corresponding to the CT data in the following manner: inputting the fusion characteristics c into a classifier for classification to obtain a trained model and obtain a final three-dimensional data result, wherein the method specifically comprises the following steps: and (3) connecting the fusion characteristics c with a K-dimensional full-connected layer, classifying by using a softmax classifier to obtain a trained model and obtain a final three-dimensional data result, wherein the essence of the softmax function is to compress an arbitrary K-dimensional real number vector into another K-dimensional real number vector, the value of each element in the real number vector is between (0 and 1), and the calculation formula is shown as a formula (7):
it should be noted that, in the embodiment of the present invention, the input of the three-dimensional data extraction model is a pixel point of a picture (a matrix is read in a computer), and the output is a three-dimensional data result.
Fig. 2 is a flowchart of an alternative method for extracting three-dimensional pelvic data according to an embodiment of the present invention, as shown in fig. 2, first, CT tomography medical image data of a pelvis is obtained; then, changing the acquired CT data; extracting LBP characteristic a of CT data of the pelvis before and after change, and performing dimension reduction processing on the LBP characteristic a; extracting depth features b in the CT data of the pelvis before and after the change by using a convolutional neural network; fusing the LBP characteristic a and the depth characteristic b after dimensionality reduction to obtain a fusion characteristic c; inputting the fusion characteristics c into a classifier for classification to obtain a trained model and obtain final three-dimensional data; and importing the CT tomography medical image data of the pelvis to be extracted into the trained model to obtain a three-dimensional data result, and displaying the three-dimensional data on a preset display module.
By the extraction method of the pelvis three-dimensional data provided by the embodiment of the invention, the pelvis three-dimensional data is extracted based on LBP characteristics and deep learning, so that an operator can quickly and accurately obtain the pelvis three-dimensional data before hip revision. Specifically, CT tomography medical image data of pelvis are obtained and changed; obtaining fusion characteristics based on LBP characteristics and deep learning; and inputting the fusion characteristics into the three-dimensional data extraction model so as to obtain the three-dimensional data of the CT data and displaying the obtained three-dimensional data.
Meanwhile, the pelvis three-dimensional data is extracted based on LBP characteristics and deep learning, so that an operator can quickly and accurately obtain the pelvis three-dimensional data before hip revision surgery, and powerful support is provided for preoperative diagnosis, treatment, surgery and the like.
Example two
According to another aspect of the embodiments of the present invention, there is further provided an embodiment of an apparatus for performing the method for extracting three-dimensional pelvic data according to the first embodiment, and fig. 3 is a schematic diagram of an apparatus for extracting three-dimensional pelvic data according to the second embodiment of the present invention, as shown in fig. 3, the apparatus for extracting three-dimensional pelvic data includes: a first acquisition unit 31, a processing unit 33, an input unit 35, a second acquisition unit 37 and a conversion unit 39. The device for extracting three-dimensional pelvic data will be described in detail below.
The first acquisition unit 31 is configured to acquire tomographic medical image CT data of a pelvis, where the CT data is data to be subjected to three-dimensional data extraction.
And the processing unit 33 is configured to process the CT data to obtain a three-dimensional data extraction model.
An input unit 35 for inputting the CT data to the three-dimensional data extraction model.
And a second obtaining unit 37 for obtaining an output result of the three-dimensional data extraction model.
And a conversion unit 39 for converting the output result into three-dimensional data corresponding to the CT data.
It should be noted here that the first acquiring unit 31, the processing unit 33, the input unit 35, the second acquiring unit 37 and the converting unit 39 correspond to steps S102 to S110 in the first embodiment, and the above units are the same as the examples and application scenarios realized by the corresponding steps, but are not limited to the disclosure in the above embodiment 1. It should be noted that the above-described elements as part of an apparatus may be implemented in a computer system, such as a set of computer-executable instructions.
As can be seen from the above, in the above embodiments of the present application, the first obtaining unit may be used to obtain CT data of a tomography medical image of the pelvis, where the CT data is data to be subjected to three-dimensional data extraction; then processing the CT data by using a processing unit to obtain a three-dimensional data extraction model; inputting the CT data into a three-dimensional data extraction model by using an input unit; acquiring an output result of the three-dimensional data extraction model by using a second acquisition unit; and converting the output result into three-dimensional data corresponding to the CT data by using a conversion unit. The pelvis three-dimensional data extraction device provided by the embodiment of the invention achieves the purpose of obtaining the three-dimensional data corresponding to the CT data by establishing the three-dimensional data extraction model, achieves the technical effect of improving the accuracy of extracting the pelvis three-dimensional data, and solves the technical problem that the accuracy of the extracted pelvis three-dimensional data is lower because the metal artifact in the CT data is removed by manual operation when the pelvis three-dimensional data is extracted in the related technology.
In an alternative embodiment, the processing unit comprises: an acquisition module for obtaining two-dimensional medical image data based on the CT data; and the analysis module is used for analyzing the CT data to obtain the metal artifact degree of the CT data and identifying the metal artifact degree.
In an alternative embodiment, the processing unit comprises: the transformation module is used for carrying out predetermined transformation on the CT data to obtain transformed CT data, wherein the predetermined transformation comprises at least one of the following modes: rotation transformation, translation transformation and noise addition transformation; the first extraction module is used for extracting local binary pattern LBP (first binary pattern) characteristics I of the CT data before transformation and the CT data after transformation; and the dimension reduction module is used for carrying out dimension reduction processing on the LBP feature I to obtain the LBP feature I after dimension reduction.
In an alternative embodiment, the first extraction module comprises: the setting submodule is used for setting a first pixel size of the CT data and a second pixel size of a subblock of the CT data, wherein the subblock is a subblock with a preset size obtained by dividing the CT data; the dividing submodule is used for dividing two pairs of CT data according to the pixel size of the sub-blocks of the CT data to obtain a preset number of sub-blocks; the determining submodule is used for determining the histogram of each subblock in a preset number of subblocks and normalizing the histogram of each subblock to obtain a statistical histogram of each subblock; and the connecting sub-module is used for connecting the statistical histograms of the sub-blocks to obtain a first LBP characteristic of the CT data.
In an alternative embodiment, the dimension reduction module comprises: and the dimensionality reduction submodule is used for carrying out dimensionality reduction on the LBP characteristic I by utilizing a Principal Component Analysis (PCA) technology to obtain the LBP characteristic I after dimensionality reduction.
In an alternative embodiment, the processing unit comprises: the second extraction module is used for extracting depth features II of the CT data before transformation and the CT data after transformation by using a convolutional neural network; and the fusion module is used for carrying out fusion processing on the LBP characteristic I and the depth characteristic II after dimension reduction to obtain a fusion characteristic III.
In an alternative embodiment, the processing unit comprises: and the input module is used for inputting the three fusion characteristics into a full connection layer with preset dimensionality and classifying by using a preset classifier to obtain a three-dimensional data extraction model.
EXAMPLE III
According to another aspect of the embodiments of the present invention, there is also provided a storage medium including a stored program, wherein the program performs the method of extracting three-dimensional pelvic data according to any one of the above.
Example four
According to another aspect of the embodiments of the present invention, there is also provided a processor, configured to execute a program, where the program executes the method for extracting three-dimensional pelvic data according to any one of the above methods.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
In the above embodiments of the present invention, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
In the embodiments provided in the present application, it should be understood that the disclosed technology can be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units may be a logical division, and in actual implementation, there may be another division, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, units or modules, and may be in an electrical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (10)
1. A method for extracting pelvis three-dimensional data is characterized by comprising the following steps:
acquiring tomography medical image CT data of a pelvis, wherein the CT data is data to be subjected to three-dimensional data extraction;
processing the CT data to obtain a three-dimensional data extraction model;
inputting the CT data into the three-dimensional data extraction model;
acquiring an output result of the three-dimensional data extraction model;
and converting the output result into three-dimensional data corresponding to the CT data.
2. The method of claim 1, wherein processing the CT data comprises:
obtaining two-dimensional medical image data based on the CT data;
and analyzing the CT data to obtain the metal artifact degree of the CT data and identifying the metal artifact degree.
3. The method of claim 1, wherein processing the CT data comprises:
performing a predetermined transformation on the CT data to obtain the transformed CT data, wherein the predetermined transformation includes at least one of the following modes: rotation transformation, translation transformation and noise addition transformation;
extracting Local Binary Pattern (LBP) characteristics I of the CT data before transformation and the CT data after transformation;
and performing dimension reduction processing on the LBP feature I to obtain the LBP feature I after dimension reduction.
4. The method of claim 3, wherein extracting Local Binary Pattern (LBP) features of the CT data before transformation and the CT data after transformation comprises:
setting a first pixel size of the CT data and a second pixel size of a sub-block of the CT data, wherein the sub-block is a sub-block with a preset size obtained by dividing the CT data;
dividing the CT data into two pairs according to the pixel size of the sub-blocks of the CT data to obtain a preset number of sub-blocks;
determining a histogram of each subblock in the preset number of subblocks, and performing normalization processing on the histogram of each subblock to obtain a statistical histogram of each subblock;
and connecting the statistical histograms of each sub-block to obtain a first LBP characteristic of the CT data.
5. The method of claim 3, wherein performing dimension reduction on the first LBP feature to obtain the dimension-reduced first LBP feature comprises:
and performing dimensionality reduction on the LBP feature I by utilizing a Principal Component Analysis (PCA) technology to obtain the LBP feature I after dimensionality reduction.
6. The method of claim 3, wherein processing the CT data comprises:
extracting depth features II of the CT data before transformation and the CT data after transformation by using a convolutional neural network;
and fusing the LBP feature I subjected to dimensionality reduction and the depth feature II to obtain a fused feature III.
7. The method of claim 6, wherein processing the CT data to obtain a three-dimensional data extraction model comprises:
inputting the three fusion characteristics into a full connection layer with preset dimensionality, and classifying by using a preset classifier to obtain the three-dimensional data extraction model.
8. An extraction device of pelvis three-dimensional data, comprising:
the device comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring tomography medical image CT data of pelvis, and the CT data is data to be subjected to three-dimensional data extraction;
the processing unit is used for processing the CT data to obtain a three-dimensional data extraction model;
an input unit for inputting the CT data to the three-dimensional data extraction model;
a second obtaining unit configured to obtain an output result of the three-dimensional data extraction model;
and the conversion unit is used for converting the output result into three-dimensional data corresponding to the CT data.
9. A storage medium characterized by comprising a stored program, wherein the program executes the extraction method of pelvic three-dimensional data according to any one of claims 1 to 7.
10. A processor, characterized in that the processor is configured to execute a program, wherein the program executes the method for extracting three-dimensional pelvic data according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911404207.4A CN111145289A (en) | 2019-12-30 | 2019-12-30 | Extraction method and device of pelvis three-dimensional data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911404207.4A CN111145289A (en) | 2019-12-30 | 2019-12-30 | Extraction method and device of pelvis three-dimensional data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111145289A true CN111145289A (en) | 2020-05-12 |
Family
ID=70522307
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911404207.4A Pending CN111145289A (en) | 2019-12-30 | 2019-12-30 | Extraction method and device of pelvis three-dimensional data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111145289A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112562033A (en) * | 2020-12-24 | 2021-03-26 | 司法鉴定科学研究院 | Conversion method for obtaining data for quickly identifying injured individual based on pelvis CT (computed tomography) image |
| WO2022217660A1 (en) * | 2021-04-15 | 2022-10-20 | 卡本(深圳)医疗器械有限公司 | Posterior superior iliac spine automatic detection and location method based on ct images |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006065955A2 (en) * | 2004-12-14 | 2006-06-22 | Orthoclear Holdings, Inc. | Image based orthodontic treatment methods |
| CN109727236A (en) * | 2018-12-27 | 2019-05-07 | 北京爱康宜诚医疗器材有限公司 | The appraisal procedure and device of acetabular bone defect, storage medium and processor |
| CN110288170A (en) * | 2019-07-29 | 2019-09-27 | 重庆电子工程职业学院 | Prediction model construction method, traffic flow forecasting method, device and electronic equipment |
| CN110473193A (en) * | 2019-08-12 | 2019-11-19 | 北京爱康宜诚医疗器材有限公司 | Detection method, detection device, storage medium and the processor of acetabular bone defect |
-
2019
- 2019-12-30 CN CN201911404207.4A patent/CN111145289A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006065955A2 (en) * | 2004-12-14 | 2006-06-22 | Orthoclear Holdings, Inc. | Image based orthodontic treatment methods |
| CN109727236A (en) * | 2018-12-27 | 2019-05-07 | 北京爱康宜诚医疗器材有限公司 | The appraisal procedure and device of acetabular bone defect, storage medium and processor |
| CN110288170A (en) * | 2019-07-29 | 2019-09-27 | 重庆电子工程职业学院 | Prediction model construction method, traffic flow forecasting method, device and electronic equipment |
| CN110473193A (en) * | 2019-08-12 | 2019-11-19 | 北京爱康宜诚医疗器材有限公司 | Detection method, detection device, storage medium and the processor of acetabular bone defect |
Non-Patent Citations (1)
| Title |
|---|
| 张意等: "基于一致性扩散的非迭代快速CT金属伪影去除算法" * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112562033A (en) * | 2020-12-24 | 2021-03-26 | 司法鉴定科学研究院 | Conversion method for obtaining data for quickly identifying injured individual based on pelvis CT (computed tomography) image |
| CN112562033B (en) * | 2020-12-24 | 2024-01-19 | 司法鉴定科学研究院 | Transformation method for obtaining data for rapidly identifying injured individuals based on pelvis CT pictures |
| WO2022217660A1 (en) * | 2021-04-15 | 2022-10-20 | 卡本(深圳)医疗器械有限公司 | Posterior superior iliac spine automatic detection and location method based on ct images |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mahfouz et al. | Patella sex determination by 3D statistical shape models and nonlinear classifiers | |
| Lim et al. | Introducing Willmore flow into level set segmentation of spinal vertebrae | |
| Li et al. | AutoImplant 2020-first MICCAI challenge on automatic cranial implant design | |
| US11869670B2 (en) | Systems and methods for automated segmentation of patient specific anatomies for pathology specific measurements | |
| WO2024001140A1 (en) | Vertebral body sub-region segmentation method and apparatus, and storage medium | |
| CN112037200A (en) | Method for automatically identifying anatomical features and reconstructing model in medical image | |
| CN106846346A (en) | Sequence C T image pelvis profile rapid extracting methods based on key frame marker | |
| Alawneh et al. | Computer-aided diagnosis of lumbar disc herniation | |
| CN111145289A (en) | Extraction method and device of pelvis three-dimensional data | |
| CN106780491A (en) | The initial profile generation method used in GVF methods segmentation CT pelvis images | |
| Tang et al. | Automatic lumbar spinal CT image segmentation with a dual densely connected U-Net | |
| CN116258726A (en) | Temporal-mandibular joint MRI image important structure segmentation method based on deep learning | |
| CN113838048B (en) | Center positioning and ligament length calculating method for pre-operation dead center of cruciate ligament | |
| Zou et al. | Semi-automatic segmentation of femur based on harmonic barrier | |
| Wang et al. | Automatic segmentation of lumbar spine MRI images based on improved attention U-net | |
| Mahdi et al. | A u-net based system for cranial implant design with pre-processing and learned implant filtering | |
| Castro-Gutierrez et al. | Detection of acetabulum fractures using X-Ray imaging and processing methods focused on noisy images | |
| Yousefi et al. | Lumbar spine vertebral compression fracture case diagnosis using machine learning methods on CT images | |
| CN106780492A (en) | A kind of extraction method of key frame of CT pelvises image | |
| Raja'S et al. | Desiccation diagnosis in lumbar discs from clinical MRI with a probabilistic model | |
| Kurochka et al. | An algorithm of segmentation of a human spine X-ray image with the help of Mask R-CNN neural network for the purpose of vertebrae localization | |
| Shinde et al. | Multidomain feature level fusion for classification of lumbar intervertebral disc using spine MR images | |
| Hou et al. | An MRI image automatic diagnosis model for lumbar disc herniation using semi-supervised learning | |
| Hu et al. | Multirigid registration of MR and CT images of the cervical spine | |
| Mir | Segmentation of the herniated intervertebral discs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |