CN101366639A - Three-dimensional modelling approach for children's funnel chest - Google Patents
Three-dimensional modelling approach for children's funnel chest Download PDFInfo
- Publication number
- CN101366639A CN101366639A CNA2008101508824A CN200810150882A CN101366639A CN 101366639 A CN101366639 A CN 101366639A CN A2008101508824 A CNA2008101508824 A CN A2008101508824A CN 200810150882 A CN200810150882 A CN 200810150882A CN 101366639 A CN101366639 A CN 101366639A
- Authority
- CN
- China
- Prior art keywords
- costicartilage
- cartilage
- funnel chest
- model
- osseum
- 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.)
- Granted
Links
Images
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses a three-dimensional modeling method for the funnel chest of a child. Firstly, aiming at an original spiral CT picture of the funnel chest of the child, a region search method is adopted to model a thoracic hard bone part; a gray value of a cartilage region of a thoracic three-dimensional model is measured, and masking separation costal cartilage image data are created after the modification; a region growing method is adopted to apply region growing separation costal cartilage image data to a combining end of a hard bone and a cartilage of each rib; the costal cartilage data which is subjected to secondary data separation is three-dimensionally drafted to generate a three-dimensional funnel costal cartilage model; the three-dimensional funnel costal cartilage model is subjected to Gaussian smooth processing, and the model of the thoracic hard bone part of the funnel chest of the child and the three-dimensional costal cartilage model are overlapped to obtain an integral three-dimensional model of the funnel chest of the child. The region search method is adopted to process cartilages at the positions where with hard bones are combined, and the region growing method is adopted to solve the problem that a CT image of the cartilages is invisible, so the created three-dimensional model ensures that costal cartilages are not adhered to each other and each costal cartilage is round and smooth, and has high simulation degree.
Description
Technical field
The present invention relates to a kind of modeling method of funnel chest, particularly three-dimensional modelling approach for children's funnel chest.
Background technology
The sickness rate of funnel chest and pigeon chest is about 1～4 ‰, and wherein the funnel chest infant accounts for 90%.It is that sternal body caves inward that funnel chest mainly shows, and is funnel-form at sternal body and xiphoid-process junction.The lopsided thorax of funnel chest not only causes breastbone compressing heart, cardiac output minimizing, total lung capacity, TLC and the vital capacity of depression to descend, and easily causes infant psychology to feel oneself inferior, so that has a strong impact on the psychology and the physiological health of infant.The most scholar of medical circle thinks by due to costicartilage and the unusual excessive increase of breastbone.
Document " the Minimally Invasive Surgery corrective therapy present situation of funnel chest. high inferior; Li Gongcai. " Chinese pediatric surgery magazine "; 2005; (08): 431-432 " and a kind of method of utilizing Minimally Invasive Surgery to carry out the funnel chest corrective therapy is disclosed, the advantage of funnel chest Wicresoft corrective procedure is that otch and wound are little, hemorrhage and complication is few; Operating time is short; Effectively improved cardio-pulmonary function; Can obtain satisfied thorax orthopedic effect.But in fact funnel chest Wicresoft orthomorphia also exists many problems, and is not visible as the shaped position of costicartilage in the orthomorphia, only relies on CT picture and doctor's feel experience to judge, thereby can't make accurate judgement to pathological changes cartilage shaped position; The orthopedic frame position is difficult to the science location, causes orthopedic frame to be shifted easily, makes once more operability up to 30%.Therefore,, then can before art, carry out emulation experiment, not only solve the visualization problem of operation, can also solve the fixation problem of orthopedic frame, thereby improve success rate of operation if can before art, obtain the thorax threedimensional model of infant.
Human body cartilage density is very near soft tissue, and the CT image is difficult to distinguish cartilage and soft tissue, so the cartilage modeling is relatively difficult.
Document " foundation and the Stress Analysis and Research of human body thorax three-dimensional finite element model. Hu Huiying; Zhong Shizhen; open U.S. superfine. " Chinese emergency medicine ", 2007, (12): 1098-1100 " a kind of method of utilizing modeling of thorax three-dimensional finite element and stress analysis disclosed.This method adopts spiral CT that one adult male fresh cadaver specimen is carried out chest scan, sweep spacing 1.0mm, altogether 440 tension fault pictures, the medical image three-dimensional reconstruction software Mimics10.1 that uses Belgian Materialise company to produce, the tomography picture is set up illiteracy cover intercepting cartilage and os osseum view data respectively, then these numbers are called three-dimensional volume render operation generation model, set up the three-dimensional finite element model that comprises clavicle, rib, the isostructural human body thorax of spinal column.Spinal column and rib os osseum department pattern that the document is set up are more complete, costicartilage part the 4th rib and the 5th rib are sticked together at the xiphoid-process position, and 4th, costicartilage models such as 5,6,7 are too flat, the part is too thick, obviously depart from the actual physiology shape of human body, model is more coarse aspect costicartilage.
Summary of the invention
In order to overcome prior art owing to the transparency, density and soft tissues of thorax's difference of costicartilage are little, and cause spiral CT poor image quality, the sightless deficiency of cartilage, the invention provides a kind of three-dimensional modelling approach for children's funnel chest, adopt the range searching method to handle and os osseum binding site cartilage, analyze the cartilage characteristic parameter, adopt the region growing method to solve the invisible problem of cartilage CT image, can realize the modeling of human body cartilage.
The technical solution adopted for the present invention to solve the technical problems: a kind of three-dimensional modelling approach for children's funnel chest is characterized in comprising the steps:
(a) at the original spiral CT picture of children's funnel chest, adopt the range searching method, the thorax os osseum is partly carried out modeling;
(b) be 65～80 with Medical Image Processing software Mimics measurement thorax terminal cartilage of threedimensional model rib and os osseum binding site cartilage area gray value, revise back establishment illiteracy cover and separate the costicartilage view data;
(c) adopt the region growing method, the judgement formula | F (G
k)-F (s) | whether＜T sets up, and selects the terminal isolated cartilage view data of every root os osseum, threshold value T=5～8, characteristic area function F (G
k)=68～78, characteristic vector F (s)=51～70 uses region growing to separate the costicartilage view data as the growth point to every root bone os osseum and cartilage binding end;
In the formula, T is a threshold value, F (G
k) be the characteristic area function, F (s) is a characteristic vector;
(d) adopt the ray cast method to carrying out volume drawing, adopt the ray cast method to carry out volume drawing after the costicartilage data integrity separates and generate funnel chest costicartilage threedimensional model through the costicartilage data after step (b), (c) twice data separating;
(e) the costicartilage threedimensional model is carried out Gauss's smoothing processing, make it mellow and full smooth;
(f) the children's funnel chest thorax os osseum department pattern that step (a) is set up superposes with the children's funnel chest costicartilage threedimensional model of step (b)～(e) set up, and obtains complete children's funnel chest threedimensional model.
The invention has the beneficial effects as follows: owing to adopt the range searching method to handle and os osseum binding site cartilage, analyze the cartilage characteristic parameter, adopt the region growing method to solve the invisible problem of cartilage CT image, realized the modeling of human body cartilage, the threedimensional model costicartilage of building does not have adhesion, each costicartilage is mellow and full smooth, emulation degree height.
Below in conjunction with drawings and Examples the present invention is elaborated.
Description of drawings
Fig. 1 is the used funnel chest spiral CT picture of three-dimensional modelling approach for children's funnel chest of the present invention.
Fig. 2 is a tomography picture after Fig. 1 cartilage data separating.
Fig. 3 is that Fig. 1 cartilage data area increases back tomography picture.
Fig. 4 is the three-dimensional initial model of costicartilage that the inventive method is set up.
Fig. 5 is a costicartilage threedimensional model after the smoothing processing set up of the inventive method.
Fig. 6 is the funnel chest thorax threedimensional model that the inventive method is set up.
The specific embodiment
The method for building up of funnel chest thorax threedimensional model of the present invention, specifically carry out according to the following steps:
With the original spiral CT picture input of certain infant funnel chest shown in Figure 1 computer.The CT picture is a sectioning image, and grayscale mode shows that the branch master looks, overlooks and three directions are looked on a left side.Photo resolution is 512 * 512 pixels, and slice numbers is 227, and slice distance is 0.4mm.Only can tell the difference of rib os osseum and soft tissue from the original CT gradation of image, can't tell cartilage and soft tissue.
At first the original spiral CT picture of certain infant funnel chest is carried out Flame Image Process.For the situation that the original spiral CT picture of certain infant funnel chest is a gray level image, earlier it is carried out the costicartilage gradation data and extract.Tissue CT gradation of image value scope is-1024～1250, and the different tissues material has different gray values, to the isolating costicartilage of desire, except the cartilaginous tissue remainder all can be considered noise.
The range searching method.Children's's costicartilage and soft-tissue density are approximate, and the transparency is high, can't distinguish cartilage and soft tissue on the CT image, and cartilage CT gradation of image value scope non-avaible can reference, therefore can't directly separate the costicartilage view data; And the rib os osseum is because tissue density and surrounding soft tissue's difference are bigger, and the CT clear picture is visible and edge contour is clear; From physiological structure, costicartilage links to each other with the rib os osseum, and the terminal adjacent tissue of rib os osseum promptly is the costicartilage tissue.Therefore, as long as set up rib os osseum model, accurately the costicartilage zone can be determined in positioning rib os osseum edge of model zone.Os osseum gradation of image value scope is generally 226～1250, create and cover cover Bone Mask, tonal range 226～1250, divide image data can comprise hard bone view data such as thoracic vertebra, rib, breastbone, adopt volume render operation to set up rib os osseum position threedimensional model to these hard bone view data, position at threedimensional model rib os osseum tip a mouse click, then cursor of mouse appears at opposite position automatically in the three-view diagram, and this position promptly is the terminal costicartilage section start of rib os osseum.
Cartilage zone and parameter are determined.Because costicartilage density and soft tissues of thorax's difference are little, CT poor image quality, vision determine that the costicartilage zone has difficulties.But from physiological structure, costicartilage links to each other with the rib os osseum, and os osseum CT picture quality height, modeling is simple, sets up chest rib os osseum model, and its rib stub area must be a costicartilage.Create and cover cover, set up rib os osseum threedimensional model.Covering cover promptly is the prospect background data extractor, and being provided with and covering the cover parameter is minimal gray 226, and maximum gray scale 1250 is separated rib os osseum data and drawing three-dimensional model.Because children's's costicartilage transparency is high, density is with soft tissues of thorax's difference is little, spiral CT picture quality is relatively poor, and cartilage spiral CT tonal range rawness value can reference, so the actual spiral CT picture costicartilage of palpus measurement district real data could be determined.The present invention is according to the CT picture of concrete funnel chest, recording the cartilage area gray value with the area grayscale survey tool Profile Line of Medical Image Processing software Mimics is 65～80, revise the back and create the illiteracy cover, parameter is made as minimum gradation value 60, maximum gradation value is 85, average gray value is 68.88, standard deviation is 6.1751, use the illiteracy cover and can leach all most of costicartilage gray-scale pixels and part soft tissue and spleen gray-scale pixels, obtain the isolated costicartilage view data of representing with different colours.After Fig. 2 cartilage data separating the tomography picture as can be seen, use cover cover isolation of cartilage data after because separating out of view data shows as the different gray scale of different tissues, obviously be different from the Lycoperdon polymorphum Vitt of Fig. 1 basic stitch image and surrounding soft tissue.But tomography picture middle part is a spleen tissue after the cartilage data separating, and because of its CT gradation of image value is identical with cartilaginous tissue, also separated coming out needs further to handle.
The present invention adopts the further data separating method based on the region growing method.Promptly in foreground area to be split, select the starting point of a seed points that links to each other with the rib os osseum as growth; The similar features degree of search and seed points in the seed points neighborhood can guarantee to satisfy the pixel of specifying the growth criterion, and pixel and seed points region are merged.The pixel that will newly merge adopts cyclic search and merging method as new seed points again, until obtaining satisfied pixel.Make F represent to weigh the characteristic function of pixel similarity degree, G
kRepresent the zone of k growth during the step, then G
kThe provincial characteristics function be F (G
k).If current pixel is that (i, j), its characteristic vector is F (s) to s, judges whether following inequality is set up
|F(G
k)-F(s)|<T (1)
In the formula, T is certain threshold value.
If (1) formula is set up, then pixel s is incorporated into G
k+ 1=G
k+ s searches for next neighbor otherwise skip pixel s.The core of algorithm is exactly to compare F (G
k) whether satisfy certain threshold value T with the characteristic distance of F (s).
For the costicartilage view data of having separated, the present invention adopts the further mask data of region growing method.Because infant soft tissue and the very approximate cartilage CT of spleen tissue CT gray scale gray scale also can be separated as the costicartilage data.With the region growing instrument Region Growing of Medical Image Processing software Mimics, setting regions increases tool parameters: threshold value T=5～8, characteristic area function F (G
k)=68～78, characteristic vector F (s)=51～70 clicks the zone that rib os osseum end has been confirmed as costicartilage, carries out the secondary cartilage data separating of region growing.After secondary data separation and Extraction costicartilage data, can obtain costicartilage gradation of image data with the third color mark, than the data separating first time, the costicartilage data were more complete after secondary data was separated.As can be seen from Figure 3, after covering cover and adding the region growing method and handle because the region growing source is cartilage but not spleen, spleen tissue as the noise signal conductively-closed get up.
Object plotting method of the present invention adopts the ray cast method.Through the costicartilage data after twice data separating is the two-dimensional geometry data, need carry out could generating 3-D geometric model after the volume drawing, and the present invention adopts the ray cast method to realize volume drawing, reproduces the costicartilage 3-D geometric model.Planar each pixel of this method analog image is all launched a ray along direction of visual lines, and this ray passes volumetric data set.Select k equidistant sampled point along this ray, and make the cubic curve interpolation by the colour and the opacity value of nearest 8 data points of a certain sampled point of distance, obtain the opacity and the colour of this sampled point, at last with the colour of each sampled point on this ray and opacity value by forward direction after or by after synthesized forward, can obtain the colour of this pixel, and then combination reflects the contour structure of object.
Volume drawing generates threedimensional model.The 3D view is to generate by the regional triangulation calculation of separating, there is the costicartilage pixel that leaches through preceding twice data separating in every CT tomography picture, any pixel along arbitrary tomography picture sends a sight line to next tomography picture pixel, these 2 have just constituted a leg-of-mutton limit, if this edge extends to more and still has the same pixel point on lower floor's tomography picture, then this limit has just connected three tension fault pictures simultaneously; Can form leg-of-mutton second, third limit as a same reason.Number of triangles has determined the quality of 3D model, and triangle is many more, and quality is good more, and the internal memory that needs is many more.Call volume drawing instrument Calculate 3D from mask, setup parameter is selected the family for use for the 3D quality, and interpolating method selects the profile interpolation, matrix compression XY resolution is 2, Z resolution is 1, and all the other parameters are given tacit consent to, and calculates triangle and draw the body profile to obtain the three-dimensional initial model of Fig. 4 funnel chest costicartilage.From Fig. 4 threedimensional model as can be seen, set up through initial threedimensional model after the aforementioned operation, just model is more coarse, further optimizes below.
It is the further optimization that 3-D geometric model is handled that surface smoothing is handled.The surface smoothing main purpose is to reduce picture noise, and smoothing method of the present invention adopts Gauss's smoothing method.Adopt Gauss's smoothing method, it is 0.4 that smoothing factor is set, and level and smooth number of times selects 3, through getting the final geometric model of funnel chest costicartilage 3D after three smoothing processing.As can be seen from Figure 5, complete through funnel chest costicartilage model after the smoothing processing, natural shape, the clear smooth adhesion that do not have of each costicartilage.
With aforesaid children's funnel chest thorax os osseum department pattern, superpose with aforesaid children's funnel chest costicartilage threedimensional model, obtain complete children's funnel chest threedimensional model.As can be seen from Figure 6, the children's funnel chest threedimensional model image of being set up is clear, and costicartilage does not have adhesion, and each costicartilage is mellow and full smooth, emulation degree height.
Claims (1)
1. a three-dimensional modelling approach for children's funnel chest is characterized in that comprising the steps:
(a) at the original spiral CT picture of children's funnel chest, adopt the range searching method, the thorax os osseum is partly carried out modeling;
(b) be 65～80 with Medical Image Processing software Mimics measurement thorax terminal cartilage of threedimensional model rib and os osseum binding site cartilage area gray value, revise back establishment illiteracy cover and separate the costicartilage view data;
(c) adopt the region growing method, the judgement formula | F (G
k)-F (s) | whether＜T sets up, and selects the terminal isolated cartilage view data of every root os osseum, threshold value T=5～8, characteristic area function F (G
k)=68～78, characteristic vector F (s)=51～70 uses region growing to separate the costicartilage view data as the growth point to every root bone os osseum and cartilage binding end;
In the formula, T is a threshold value, F (G
k) be the characteristic area function, F (s) is a characteristic vector;
(d) adopt the ray cast method to carrying out volume drawing, adopt the ray cast method to carry out volume drawing after the costicartilage data integrity separates and generate funnel chest costicartilage threedimensional model through the costicartilage data after step (b), (c) twice data separating;
(e) the costicartilage threedimensional model is carried out Gauss's smoothing processing, make it mellow and full smooth;
(f) the children's funnel chest thorax os osseum department pattern that step (a) is set up superposes with the children's funnel chest costicartilage threedimensional model of step (b)～(e) set up, and obtains complete children's funnel chest threedimensional model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200810150882A CN100588370C (en) | 2008-09-10 | 2008-09-10 | Three-dimensional modeling method for funnel chest of children |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200810150882A CN100588370C (en) | 2008-09-10 | 2008-09-10 | Three-dimensional modeling method for funnel chest of children |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101366639A true CN101366639A (en) | 2009-02-18 |
| CN100588370C CN100588370C (en) | 2010-02-10 |
Family
ID=40410747
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200810150882A Expired - Fee Related CN100588370C (en) | 2008-09-10 | 2008-09-10 | Three-dimensional modeling method for funnel chest of children |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100588370C (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI396111B (en) * | 2009-12-30 | 2013-05-11 | Pou Yuen Technology Co Ltd | Single bone region separation method |
| CN105590340A (en) * | 2015-12-17 | 2016-05-18 | 上海市第六人民医院 | Human neck three-dimensional reconstruction method |
| CN105654546A (en) * | 2014-12-02 | 2016-06-08 | 上海金仕达卫宁软件股份有限公司 | Modeling method and device |
| CN108135552A (en) * | 2015-09-29 | 2018-06-08 | 皇家飞利浦有限公司 | The improved visualization of projected X-rays image |
| WO2019164214A1 (en) * | 2018-02-23 | 2019-08-29 | 삼성전자주식회사 | Magnetic resonance imaging device and control method therefor |
-
2008
- 2008-09-10 CN CN200810150882A patent/CN100588370C/en not_active Expired - Fee Related
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI396111B (en) * | 2009-12-30 | 2013-05-11 | Pou Yuen Technology Co Ltd | Single bone region separation method |
| CN105654546A (en) * | 2014-12-02 | 2016-06-08 | 上海金仕达卫宁软件股份有限公司 | Modeling method and device |
| CN105654546B (en) * | 2014-12-02 | 2019-05-07 | 卫宁健康科技集团股份有限公司 | Modeling method and model building device |
| CN108135552A (en) * | 2015-09-29 | 2018-06-08 | 皇家飞利浦有限公司 | The improved visualization of projected X-rays image |
| CN108135552B (en) * | 2015-09-29 | 2021-10-29 | 皇家飞利浦有限公司 | Improved visualization of projected X-ray images |
| CN105590340A (en) * | 2015-12-17 | 2016-05-18 | 上海市第六人民医院 | Human neck three-dimensional reconstruction method |
| WO2019164214A1 (en) * | 2018-02-23 | 2019-08-29 | 삼성전자주식회사 | Magnetic resonance imaging device and control method therefor |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100588370C (en) | 2010-02-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI535423B (en) | Method for simulating thoracic 4dct | |
| CN100588370C (en) | Three-dimensional modeling method for funnel chest of children | |
| CN107945169B (en) | Coronary artery image analysis method | |
| CN107993293B (en) | Three-dimensional interactive anatomy teaching method containing Chinese population difference | |
| CN109157284A (en) | A kind of brain tumor medical image three-dimensional reconstruction shows exchange method and system | |
| Van de Velde et al. | An anatomically validated brachial plexus contouring method for intensity modulated radiation therapy planning | |
| Zhang et al. | Three-dimensional reconstruction of medical images based on 3D slicer | |
| Macrini et al. | Digital cranial endocast of Pucadelphys andinus, a Paleocene metatherian | |
| KR20140015964A (en) | Method and apparatus for creating model of patient specified target organ based on blood vessel structure | |
| JP5322662B2 (en) | Image processing device | |
| Dorfling et al. | Facial approximations: Characteristics of the eye in a South African sample | |
| Kamiya et al. | Automated segmentation of recuts abdominis muscle using shape model in X-ray CT images | |
| Dahdouh et al. | A comprehensive tool for image-based generation of fetus and pregnant women mesh models for numerical dosimetry studies | |
| WO2002003304A2 (en) | Predicting changes in characteristics of an object | |
| Zhu et al. | 3D automatic MRI level set segmentation of inner ear based on statistical shape models prior | |
| US9552663B2 (en) | Method and system for volume rendering of medical images | |
| Vasilyev et al. | 3D reconstruction of the ancient Egyptian mummy skeleton from the Pushkin State Museum of Fine Arts (I, 1 1240) | |
| WO2022031564A1 (en) | System and method for augmented reality spine surgery | |
| US10692208B2 (en) | Chest wall estimation from optical scans | |
| Tan et al. | Multi-needle particle implantation computer assisted surgery based on virtual reality | |
| Zhou et al. | Enhancement of lesions in CT images of lung cancer based on MIMICS software | |
| EP2725546B1 (en) | Display processing method and apparatus | |
| WO2022170264A1 (en) | System and method for processing black bone mri data | |
| US20210241534A1 (en) | System and method for augmenting and synchronizing a virtual model with a physical model | |
| Nakao et al. | Volumetric surgical planning system for fibular transfer in mandibular reconstruction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100210 Termination date: 20120910 |