CN104794749A - Method for establishing three-dimensional electromagnetic breast simulation model on basis of clinical MRI (magnetic resonance imaging) images - Google Patents
Method for establishing three-dimensional electromagnetic breast simulation model on basis of clinical MRI (magnetic resonance imaging) images Download PDFInfo
- Publication number
- CN104794749A CN104794749A CN201510163898.9A CN201510163898A CN104794749A CN 104794749 A CN104794749 A CN 104794749A CN 201510163898 A CN201510163898 A CN 201510163898A CN 104794749 A CN104794749 A CN 104794749A
- Authority
- CN
- China
- Prior art keywords
- breast
- model
- thoracic cavity
- tissue
- simulation model
- 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
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention relates to a method for establishing a three-dimensional electromagnetic breast simulation model on the basis of clinical MRI (magnetic resonance imaging) images. The method includes the step of (1) performing stretching processing on each breast section image by a bicubic interpolation method; (2) stacking breast sections subjected to the stretching processing sequentially to obtain a three-dimensional pixel-point breast array; (3) obtaining a peripheral outline of a breast model; (4) subjecting breast tissue to discrete division; (5) obtaining a peripheral outline of a thoracic cavity, and determining a part, positioned within the peripheral outline of the thoracic cavity, of the breast model to be the thoracic cavity; (6) assigning different electrical property values for different breast tissue obtained in the step (4), and establishing the electromagnetic breast simulation model. The method for establishing the three-dimensional electromagnetic breast simulation model on the basis of the clinical MRI images has the advantages that the established breast model is capable of morphologically reflecting breasts in aspects of shape and internal tissue, such as fat, glands and the thoracic cavity.
Description
Art
The invention belongs to field of biological medicine, relate to a kind of breast Electromagnetic Simulation model
Background technology
Tumor of breast is the malignancy disease that women's incidence of disease is the highest, and mortality ratio occupies first of women's mortality of malignant tumors.The diagnosis of early stage tumor of breast all has decisive significance for the survival rate at a specified future date improving mastotic treatment rate and patient undoubtedly.The detection method of early-stage breast cancer conventional at present comprises mammography, ultrasonograph technology, computed tomography, nmr imaging technique, thermal imaging detection etc., but all multi-methods all exist certain shortcoming, as low, costly etc. in produced radiation injury, image contrast to human body.The principle of ultra broadband Electromagnetic Wave Detection breast cancer is that different biological tissue is different to electromagnetic absorption, reflection and transmissison characteristic, and the electromagnetic field produced when the pulse signal of antenna transmission is propagated in breast tissue can reflect the abundant information of malignant tissue.Simultaneously ultra-wideband microwave signal has that radiation power is low, the target information amount of carrying large, provides grade to locate and the advantage such as testing cost is lower, can as the conventional means of early stage breast cancer.At present, this technology is still in the development phase, and research mainly based on numerical simulation, but needs time from really practical.In tumor of breast detect delay in early days, model normally comparatively simple model, such as the square model of employing, half spherical model etc.This class model has symmetry, and does not consider the gland tissue impact of internal mammary, and the signal obtained during detection has a lot of similarity, in later stage signal transacting, can carry out noise elimination relatively easily.But real breast environment is complicated more than this kind of naive model, wherein has non-homogeneous skin, gland tissue, thoracic cavity etc.These tissues can cause larger scattering to signal, and body of gland can cause signal to produce larger decay simultaneously, and this is all undoubtedly larger challenge concerning signal transacting and imaging.Along with the propelling of problem, be necessary to study, to verify the applicability of all kinds of algorithm under complex environment in more real breast environment.
Summary of the invention
In order to simulate the electromagnetic transmission situation at internal mammary better, the invention provides a kind of method setting up three-dimensional breast electromagnetic wave model based on clinical nuclear magnetic resonance (MRI) image.Adopt the breast model of method establishment provided by the invention on morphology, shape and the interior tissue of breast can be reflected truly, as fat, body of gland, thoracic cavity etc.Technical scheme of the present invention is as follows:
A kind of method setting up three-dimensional breast Electromagnetic Simulation model based on clinical MRI image:
(1) bicubic interpolation method is adopted to carry out stretch processing to each breast slice map.
(2) the breast section after stretched process be stacked up in order, obtain the breast array of a three-dimensional image vegetarian refreshments, third dimension data correspond to breast and to cut into slices perpendicular direction;
(3) to each xsect of breast array, elder generation for boundary value, carries out image two divisional processing with fatty tissue of breast's gray-scale value; Recycling canny Boundary Recognition algorithm processes, and obtains the circumference of breast model;
(4) discrete division is carried out to breast tissue, method is as follows: inner in breast array, set different threshold ranges, the pixel of gray-scale value in same threshold range is all classified as the breast tissue of a certain medium, after discrete processes, obtains breast model; Add one deck skin at the boundary of breast model, Adding Way is: the circumference of the breast model identified in (3) step, centered by each profile frontier point, the pixel region near it is set as skin;
(5) for thoracic tissues sets a boundary value, adopt the method for step (3), obtain the circumference in thoracic cavity, and the part be positioned at by breast model within the circumference of thoracic cavity is all set as thoracic cavity;
(6) the different breast tissues obtained for step (4) give different electrical characteristics values, set up breast Electromagnetic Simulation model.
Wherein, in step (4), altogether can set four kinds of normal breast tissue classifications, be respectively: fat 1, fat 2, body of gland 1 and body of gland 2, the fat that corresponding specific inductive capacity is lower and higher respectively and body of gland, the gray-scale value corresponding to organizing four kinds is set as: 255-120,119-90,89-60 and 59-30.
The present invention is based on clinical medicine MRI figure and establish the breast model being applicable to FDTD Electromagnetic Simulation, this model simulates the complex distributions situation of breast contours and internal mammary tissue well.Utilize the breast model of method establishment of the present invention, simulation study can be made to carry out in more complicated breast environment, to verify the applicability of all kinds of imaging algorithm under complicated breast environment, and signal processing method, imaging algorithm are optimized, strengthen the robustness of algorithm.
Accompanying drawing explanation
The breast source figure of Fig. 1 medical MRI cuts into slices schematic diagram.
Stacking and the sagittal of Fig. 2 breast image intercepts schematic diagram.
The sagittal figure of Fig. 3 image stack rear udder attachment.
Sagittal figure after Fig. 4 breast vertical direction interpolation.
Some cross sectional representation of Fig. 5 interpolation rear udder attachment.
The breast xsect of Fig. 6 after binary conversion treatment.
Breast border after the process of Fig. 7 canny operator identification.
The border schematic diagram of Fig. 8 internal mammary and air.
Fig. 9 organizes discrete division rear udder attachment illustraton of model.
Figure 10 adds the breast model figure after skin.
Figure 11 breast model numbering schematic diagram.
Figure 12 adds the electrology characteristic schematic diagram of tumour rear udder attachment model.
Embodiment
First the present invention carries out stretch processing to full group medical image MRI source figure, matches to make each pixel and FDTD grid.Then, all MRI source figure carried out stacking and carries out interpolation processing in the direction perpendicular to figure, making each pixel in vertical direction and FDTD Grid Align.Secondly, Boundary Recognition is carried out to the xsect after each interpolation, finds out the border between internal mammary and air.Again, discrete division is organized to internal mammary, according to gray-scale value, breast is divided into different tissue class, and adds skin layer along breast border.Finally different tissues is numbered and electrology characteristic assignment, model can be applied in FDTD Electromagnetic Simulation model.Concrete steps are as follows:
1. source figure stretch processing
The T1 source figure that hospital provides is resolution 512x512 substantially, i.e. total 512x512 pixel.Here each pixel, we are thought of as a grid.The actual size of every pictures representative is by the actual size represented by the single pixel of decision.Be illustrated in figure 1 certain MRI image of patient A, engineer's scale in figure pixel shared by above-below direction is 256 points, these 256 pixels represent the length of real 15cm, so can know that whole Figure 51 2x512 pixel just represents the length of real 30x30cm.Therefore if do not carry out any adjustment to picture, so each pixel of original graph represents the sizing grid of 0.586x0.586mm (300mm/512).But, under normal circumstances, the grid of 0.5x0.5mm is used in Fdtd Method (FDTD) algorithm, therefore, when carrying out model derivation, first stretch processing is carried out to source figure, the grid of the corresponding 0.5x0.5mm of each pixel after making it stretch, for patient as shown in Figure 1, need picture to be stretched as 600x600 pixel, the size of corresponding 30x30cm.This step adopts bicubic (bicubic) picture interpolation method to realize, and the method is consuming time the longest, but the picture degree of distortion after stretching is smaller.
2. the stacking interpolation of source figure
The MRI striograph of each patient is formed by one group of section of multiple diverse locations, to reflect the inner case of breast diverse location.In order to set up breast three-dimensional model, need to process to reduce to these sections the information of other positions of breast here.First cut into slices the stretch processing carried out in first step to adapt to FDTD grid to all MRI, such as patient A has 24 MRI source slice maps, is stretched as 600X600.Next be stacked up in order by the picture after stretching, as shown in Figure 2, will obtain the voxel point array of a 600X600X24, third dimension data are corresponding with breast cuts into slices perpendicular direction.According to the slice location information in the figure of MRI source, known spacing of often opening between section is approximately 5mm, each pixel therefore in the third dimension is corresponding 5mm.But three-dimensional FDTD grid is chosen for 0.5mmx0.5mmx0.5mm usually, therefore need to carry out interpolation processing to meet grid requirement to section stacking direction.
In this step, first as shown in Figure 2 sagittal plane intercepting is carried out to the voxel of heap poststack, more successively each sagittal figure is carried out interpolation stretch processing to mate FDTD grid.Fig. 3 is the image after the intercepting of an example, and this figure is the sagittal profile diagram of breast, and the pixel of this figure is 600x24, and each pixel of vertical direction represents 5mm.In order to mate conventional FDTD grid, being carried out bicubic interpolation process, being stretched as 600x240 pixel here, each like this pixel represents the grid of 0.5mmx0.5mm.Fig. 4 is the sagittal figure after interpolation.Sagittal plane sectional drawing after interpolation is combined, just obtains a complete corresponding FDTD grid array of pixels, the grid of the corresponding 0.5mmx0.5mmx0.5mm of each pixel.The array obtaining a 600x600x240 in this instance.
3. breast Boundary Recognition
For completing steps 1, each xsect of the breast array after 2 interpolation, all carries out the operation of a Boundary Recognition, to tell inside and the air of breast.If exemplified by Fig. 5, this figure is some xsects.Before carrying out Boundary Recognition, first image two divisional processing is carried out to this figure.For the pixel of all 600x600, with fatty tissue of breast's gray-scale value for boundary value, part assignment gray-scale value being greater than this value is 255, is white; Part assignment gray-scale value being less than this value is 0, is black.Effect after two divisional processing as shown in Figure 6, is desired to make money or profit to process with canny Boundary Recognition algorithm for this, can obtain a boundary graph, as shown in Figure 7.The frontier point that canny algorithm now identifies has a lot, but only needs to obtain outermost internal mammary and air interface, and therefore next step will outermost Boundary Extraction out, as shown in Figure 8.Finally, a Boundary Recognition is all carried out to all xsects, so just obtains the circumference of breast model.
4. the discrete division of breast tissue
This step is the committed step that model is derived.Namely the pixel of gray-scale value in certain threshold value, at internal mammary, is all classified as a certain medium by principle here, and is unified assignment for certain particular value is to carry out mark display.In the present invention, altogether set four kinds of normal tissue class, be respectively: fat 1, fat 2, body of gland 1 and body of gland 2, the fat that corresponding specific inductive capacity is lower and higher and body of gland.Gray-scale value corresponding to organizing four kinds is set as: 255-120,119-90,89-60 and 59-30.After discrete processes, breast model as shown in Figure 9.Next one deck skin will be added at breast boundary, Adding Way is the breast contours (for Fig. 8) identified in the 3rd step, centered by each profile frontier point, the pixel region of the 3x3 around it is set as skin, as shown in Figure 10.Due to the gray-scale value of thoracic cavity and the gray-scale value of breast tissue similar, some thoracic cavity part has been classified as breast tissue, therefore needs to identify border, thoracic cavity.Repeat the process being similar to breast Boundary Extraction, identify border, thoracic cavity, and the part within border is all set as thoracic cavity.
5. breast model is set up and is imported
Different tissues same numbering is marked, and gives corresponding electrical characteristics values to when importing FDTD realistic model often kind of tissue.
Table 1 is the organization number of an example and organizes electrology characteristic correspondence table.
The electrical characteristics values of different tissues in table 1 breast model
Carry out assignment according to upper table to breast model, Figure 11 is the part that breast model intercepts, the organization number that wherein different colours is corresponding different.Figure 12 is after manually adding a tumour, the electrology characteristic figure of this breast model.Through above 5 steps, just establish a complete breast Electromagnetic Simulation model.
Claims (2)
1. set up the method for three-dimensional breast Electromagnetic Simulation model based on clinical MRI image for one kind:
(1) bicubic interpolation method is adopted to carry out stretch processing to each breast slice map.
(2) the breast section after stretched process be stacked up in order, obtain the breast array of a three-dimensional image vegetarian refreshments, third dimension data correspond to breast and to cut into slices perpendicular direction;
(3) to each xsect of breast array, elder generation for boundary value, carries out image two divisional processing with fatty tissue of breast's gray-scale value; Recycling canny Boundary Recognition algorithm processes, and obtains the circumference of breast model;
(4) discrete division is carried out to breast tissue, method is as follows: inner in breast array, set different threshold ranges, the pixel of gray-scale value in same threshold range is all classified as the breast tissue of a certain medium, after discrete processes, obtains breast model; Add one deck skin at the boundary of breast model, Adding Way is: the circumference of the breast model identified in (3) step, centered by each profile frontier point, the pixel region near it is set as skin;
(5) for thoracic tissues sets a boundary value, adopt the method for step (3), obtain the circumference in thoracic cavity, and the part be positioned at by breast model within the circumference of thoracic cavity is all set as thoracic cavity;
(6) the different breast tissues obtained for step (4) give different electrical characteristics values, set up breast Electromagnetic Simulation model.
2. the method setting up three-dimensional breast Electromagnetic Simulation model according to claim 1, it is characterized in that, in step (4), altogether set four kinds of normal breast tissue classifications, be respectively: fat 1, fat 2, body of gland 1 and body of gland 2, the fat that corresponding specific inductive capacity is lower and higher respectively and body of gland, the gray-scale value corresponding to organizing four kinds is set as: 255-120,119-90,89-60 and 59-30.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510163898.9A CN104794749A (en) | 2015-04-08 | 2015-04-08 | Method for establishing three-dimensional electromagnetic breast simulation model on basis of clinical MRI (magnetic resonance imaging) images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510163898.9A CN104794749A (en) | 2015-04-08 | 2015-04-08 | Method for establishing three-dimensional electromagnetic breast simulation model on basis of clinical MRI (magnetic resonance imaging) images |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104794749A true CN104794749A (en) | 2015-07-22 |
Family
ID=53559528
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510163898.9A Pending CN104794749A (en) | 2015-04-08 | 2015-04-08 | Method for establishing three-dimensional electromagnetic breast simulation model on basis of clinical MRI (magnetic resonance imaging) images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104794749A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106023277A (en) * | 2016-05-18 | 2016-10-12 | 华北电力大学（保定） | Magnetic induction magnetoacoustic endoscopic image modeling and simulation method |
| CN107103605A (en) * | 2016-02-22 | 2017-08-29 | 上海联影医疗科技有限公司 | A kind of dividing method of breast tissue |
| CN107212884A (en) * | 2017-05-11 | 2017-09-29 | 天津大学 | A kind of supine body position oppresses breast imaging method |
| CN109363675A (en) * | 2018-10-09 | 2019-02-22 | 中国人民解放军第四军医大学 | The suppressing method of edge effect in a kind of detection of Breast electrical impedance scanning imagery |
| CN112507647A (en) * | 2020-12-15 | 2021-03-16 | 重庆邮电大学 | Electromagnetic coupling time domain modeling analysis method for space electromagnetic field action bifurcation line |
| CN115115770A (en) * | 2021-03-22 | 2022-09-27 | 天津大学 | Three-dimensional electromagnetic simulation breast model construction method based on MRI (magnetic resonance imaging) image |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102792189A (en) * | 2009-12-02 | 2012-11-21 | 上海联影医疗科技有限公司 | A method of utilization of high dielectric constant (HDC) materials for reducing SAR and enhancing SNR in MRI |
| CN103549953A (en) * | 2013-10-25 | 2014-02-05 | 天津大学 | Method for extracting microwave detection breast model based on medical magnetic resonance imaging |
-
2015
- 2015-04-08 CN CN201510163898.9A patent/CN104794749A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102792189A (en) * | 2009-12-02 | 2012-11-21 | 上海联影医疗科技有限公司 | A method of utilization of high dielectric constant (HDC) materials for reducing SAR and enhancing SNR in MRI |
| CN103549953A (en) * | 2013-10-25 | 2014-02-05 | 天津大学 | Method for extracting microwave detection breast model based on medical magnetic resonance imaging |
Non-Patent Citations (1)
| Title |
|---|
| ZONGJIE WANG ET AL: "Development of Anatomically Realistic Numerical Breast Phantoms Based on T1- and T2-Weighted MRIs for Microwave Breast Cancer Detection", 《IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107103605A (en) * | 2016-02-22 | 2017-08-29 | 上海联影医疗科技有限公司 | A kind of dividing method of breast tissue |
| CN106023277A (en) * | 2016-05-18 | 2016-10-12 | 华北电力大学（保定） | Magnetic induction magnetoacoustic endoscopic image modeling and simulation method |
| CN106023277B (en) * | 2016-05-18 | 2018-07-27 | 华北电力大学（保定） | A kind of modeling and simulation method of magnetic induction magnetosonic endoscopic picture |
| CN107212884A (en) * | 2017-05-11 | 2017-09-29 | 天津大学 | A kind of supine body position oppresses breast imaging method |
| CN107212884B (en) * | 2017-05-11 | 2020-07-24 | 天津大学 | Supine position compression breast imaging method |
| CN109363675A (en) * | 2018-10-09 | 2019-02-22 | 中国人民解放军第四军医大学 | The suppressing method of edge effect in a kind of detection of Breast electrical impedance scanning imagery |
| CN109363675B (en) * | 2018-10-09 | 2022-03-29 | 中国人民解放军第四军医大学 | Method for inhibiting edge effect in mammary gland electrical impedance scanning imaging detection |
| CN112507647A (en) * | 2020-12-15 | 2021-03-16 | 重庆邮电大学 | Electromagnetic coupling time domain modeling analysis method for space electromagnetic field action bifurcation line |
| CN112507647B (en) * | 2020-12-15 | 2023-07-21 | 重庆邮电大学 | Electromagnetic coupling time domain modeling analysis method for space electromagnetic field acting bifurcation line |
| CN115115770A (en) * | 2021-03-22 | 2022-09-27 | 天津大学 | Three-dimensional electromagnetic simulation breast model construction method based on MRI (magnetic resonance imaging) image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108460809B (en) | Deep convolutional encoder-decoder for prostate cancer detection and classification | |
| CN104794749A (en) | Method for establishing three-dimensional electromagnetic breast simulation model on basis of clinical MRI (magnetic resonance imaging) images | |
| CN106600609B (en) | Spine segmentation method and system in medical image | |
| US7646902B2 (en) | Computerized detection of breast cancer on digital tomosynthesis mammograms | |
| Viji et al. | Automatic detection of brain tumor based on magnetic resonance image using CAD System with watershed segmentation | |
| Sinha et al. | Medical image processing | |
| CN109009110A (en) | Axillary lymphatic metastasis forecasting system based on MRI image | |
| CN105488781B (en) | A kind of dividing method based on CT images liver neoplasm lesion | |
| CN109512464A (en) | A kind of disorder in screening and diagnostic system | |
| CN115496771A (en) | Brain tumor segmentation method based on brain three-dimensional MRI image design | |
| CN111583246A (en) | Method for classifying liver tumors by utilizing CT (computed tomography) slice images | |
| Sparks et al. | Fully automated prostate magnetic resonance imaging and transrectal ultrasound fusion via a probabilistic registration metric | |
| CN111784701A (en) | Ultrasonic image segmentation method and system combining boundary feature enhancement and multi-scale information | |
| Gao et al. | Detection and recognition of ultrasound breast nodules based on semi-supervised deep learning: a powerful alternative strategy | |
| Hasan | A hybrid approach of using particle swarm optimization and volumetric active contour without edge for segmenting brain tumors in MRI scan | |
| Ganvir et al. | Filtering method for pre-processing mammogram images for breast cancer detection | |
| Narayanan et al. | Adaptation of a 3D prostate cancer atlas for transrectal ultrasound guided target-specific biopsy | |
| Liu et al. | 3-D prostate MR and TRUS images detection and segmentation for puncture biopsy | |
| CN109741439A (en) | A kind of three-dimensional rebuilding method of two dimension MRI fetus image | |
| Sparks et al. | Multiattribute probabilistic prostate elastic registration (MAPPER): application to fusion of ultrasound and magnetic resonance imaging | |
| Byra et al. | Impact of ultrasound image reconstruction method on breast lesion classification with neural transfer learning | |
| Alvarez-Jimenez et al. | Differentiating cancerous and non-cancerous prostate tissue using multi-scale texture analysis on MRI | |
| Chen et al. | Research related to the diagnosis of prostate cancer based on machine learning medical images: A review | |
| KR101227272B1 (en) | Image registration method of ultrasound imaging and magnetic resonance imaging | |
| Lee et al. | Breast ultrasound automated ROI segmentation with region growing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150722 |