【摘要】： 目的 近幾年興起的針對膝關(guān)節(jié)損傷的數(shù)字化虛擬研究很多，人體膝關(guān)節(jié)的三維重建是這些研究的基礎(chǔ)。但由于膝關(guān)節(jié)結(jié)構(gòu)的復(fù)雜性以及影像技術(shù)的限制，這些三維重建往往只重建了膝關(guān)節(jié)的骨結(jié)構(gòu)。本研究將結(jié)合膝關(guān)節(jié)的CT、MRI圖像，利用三維重建技術(shù)及圖像配準技術(shù)，構(gòu)建膝關(guān)節(jié)的骨、半月板、前／后交叉韌帶、關(guān)節(jié)軟骨，為開展膝關(guān)節(jié)的數(shù)字醫(yī)學(xué)研究進行有關(guān)模型構(gòu)建及應(yīng)用等方面的探索。 臨床膝關(guān)節(jié)手術(shù)中要求盡可能保護局部血液循環(huán)，以避免局部骨壞死的出現(xiàn)。由于膝關(guān)節(jié)局部血供復(fù)雜，傳統(tǒng)影像學(xué)檢查無法觀察血管，血管造影可以較好顯示局部血管走行，但其質(zhì)量受造影劑流動速度控制、圖像分割處理技術(shù)等影響。為克服以上不足，本研究將進行動脈造影灌注后行CT掃描，重建膝關(guān)節(jié)骨與血管，旨在準確、完整地顯示膝關(guān)節(jié)動脈的走行與分布，為膝關(guān)節(jié)外科及解剖的教學(xué)、科研提供一種清楚、準確、有價值的顯像手段。 人體解剖學(xué)是一門形態(tài)科學(xué)，目前基于二維圖像的教學(xué)方法教學(xué)難度很大。隨著計算機圖形圖像技術(shù)的不斷完善以及虛擬人研究的不斷進展，已出現(xiàn)多種圖像的三維重建方法，但目前大多數(shù)三維圖像的顯示都需要專業(yè)軟件，價格昂貴且存在知識產(chǎn)權(quán)問題，無法用于三維解剖圖譜的普及與推廣應(yīng)用。本研究將利用先進的VRML語言，編寫相關(guān)的程序，建立基于web的膝關(guān)節(jié)三維瀏覽網(wǎng)頁，為完成三維解剖圖譜奠定基礎(chǔ)。 盡管膝關(guān)節(jié)的穩(wěn)定性測試目前已應(yīng)用于膝關(guān)節(jié)運動功能的評價，但由于膝關(guān)節(jié)的離體標本研究無法模擬膝關(guān)節(jié)的真實運動，而在體的運動測試又無法獲得骨結(jié)構(gòu)的運動信息，因而不能得到準確的膝關(guān)節(jié)穩(wěn)定性數(shù)據(jù)，也就無法對膝關(guān)節(jié)微損傷的早期診斷及防治措施進行深入的研究。本研究將利用圖像三維重建、2D／3D圖像配準技術(shù)以及圖像處理技術(shù)探討建立膝關(guān)節(jié)在體穩(wěn)定性評價系統(tǒng)的可行性。 逆向工程是是基于一個已獲得的實物模型來構(gòu)造出CAD模型，并通過調(diào)整相關(guān)參數(shù)來達到對實物模型的逼近、修改和完善，進而將這些CAD模型用于產(chǎn)品的分析和制造�；诶萌S激光掃描儀開展的逆向工程測量精度高的優(yōu)點，本研究將探討在離體狀態(tài)下利用逆向工程技術(shù)開展膝關(guān)節(jié)運動還原的方法，并對利用激光三維掃描儀的逆向工程運動還原方法的測試精度進行標定。 應(yīng)用計算機三維重建技術(shù)、2D／3D圖像配準技術(shù)及圖像處理技術(shù)構(gòu)建了膝關(guān)節(jié)在體穩(wěn)定性測試系統(tǒng)。本研究將以逆向工程運動還原方法為對照，研究膝關(guān)節(jié)在體穩(wěn)定性測試系統(tǒng)的精度。 材料與方法 人體成年新鮮膝關(guān)節(jié)標本1例，進行CT掃描共387層，層厚為0.299mm，隨后進行MRI掃描，共64層，層厚為1.497mm，利用三維重建軟件Mimics及逆向工程軟件Geomagic對圖像進行三維重建及圖像配準，構(gòu)建膝關(guān)節(jié)骨、軟骨、韌帶及半月板等結(jié)構(gòu)。 新鮮成人完整下肢標本1例，用填充劑配成合適濃度，對標本的動脈進行灌注。隨后進行CT掃描，層厚0.499mm，取膝關(guān)節(jié)部分共671層用于本研究。采用Mimic進行膝關(guān)節(jié)骨及血管等結(jié)構(gòu)的三維重建，分別采用Mimics和3ds max進行動脈的透明化顯示及效果比較。 采用BS Contact VRML 6.1和Vrmlpad作為VRML顯示插件及程序編輯軟件，將已三維重建膝關(guān)節(jié)各部分結(jié)構(gòu)轉(zhuǎn)為wrl文件，并進行VRML編程及網(wǎng)頁制作。 對健康成年志愿者膝關(guān)節(jié)進行CT掃描，采集志愿者保持某一姿勢時互成直角的正側(cè)位X線平片，進行膝關(guān)節(jié)CT圖像的三維重建，并在軟件中建立虛擬X線放射系統(tǒng)再現(xiàn)2張互成直角的X線平片，采用2D／3D圖像配準還原攝平片時膝關(guān)節(jié)的位置，計算兩種位置之間的相對位移及角度變化。 人體膝關(guān)節(jié)標本上、下端包埋后由激光三維掃描儀采集包埋塊的位置信息，通過Geomagic軟件進行膝關(guān)節(jié)的位置還原，并將標志物固定于精度可達0.01mm及0.01°的KOHZU精密測試平臺上，由激光三維掃描儀采集標志物的運動信息，通過軟件計算其運動參數(shù)，檢測逆向工程運動還原方法的測試精度。 人體膝關(guān)節(jié)標本3例，利用G型臂X線機獲得正側(cè)位X線片，同時利用激光三維掃描儀掃描包埋塊的三維點云信息，每個標本采集2次任意角度的正側(cè)位X線片和包埋塊的三維點云信息。采用虛擬X線放射系統(tǒng)和逆向工程運動還原方法分別進行位置還原，將還原出來的位置分別計算與CT圖像構(gòu)建的膝關(guān)節(jié)模型的相對運動，并進行配對樣本T檢驗。對由虛擬X線放射系統(tǒng)和逆向工程技術(shù)系統(tǒng)導(dǎo)出膝關(guān)節(jié)的相對運動數(shù)據(jù)，采用描述性統(tǒng)計觀察兩者在平移和旋轉(zhuǎn)各個軸上的差異。 結(jié)果 利用CT對骨結(jié)構(gòu)顯示較好的特點，將CT圖像導(dǎo)入圖像三維重建軟件Mimics10.01中進行重建，然后將骨的三維模型分別以stl文件格式保存于電腦中。將MRI與CT分別構(gòu)建的骨模型導(dǎo)入Geomagic軟件中進行位置配準，對CT的三維模型進行坐標變換使其適合MRI圖像的坐標系，使CT構(gòu)建的骨模型可以準確地導(dǎo)入MRI圖像中。利用MRI對軟骨、韌帶等組織顯示較好的特點，用Mimics10.01軟件構(gòu)建出半月板、前／后交叉韌帶和關(guān)節(jié)軟骨的三維模型，導(dǎo)入Geomagic軟件中進行三維修飾，使其結(jié)構(gòu)稍平滑。關(guān)節(jié)軟骨需要再次導(dǎo)入CT影像中使其邊緣平滑。最終建立具有骨、半月板、前／后交叉韌帶和關(guān)節(jié)軟骨的三維膝關(guān)節(jié)模型。 利用血管造影灌注后CT掃描數(shù)據(jù)，用Mimics軟件成功構(gòu)建膝關(guān)節(jié)骨-血管三維立體結(jié)構(gòu)模型，能夠清晰地顯示膝關(guān)節(jié)骨結(jié)構(gòu)及血管之間的相互位置關(guān)系及立體形態(tài)。對于關(guān)節(jié)的局部血供也可以清楚展現(xiàn)，如髕周的血管網(wǎng)。利用Mimics和3ds max都可實現(xiàn)膝關(guān)節(jié)動脈三維模型的透明化，兩者顯示效果相近。相比較而言，利用Mimics軟件操作比較簡單，但軟件設(shè)定的透明度不能任意調(diào)節(jié)。而3ds max軟件操作相對復(fù)雜一些，特別是對于容量較大的文件，導(dǎo)入3ds max的過程較慢，，但它可以任意設(shè)定透明度，具有較強的可視化設(shè)計能力。 通過Vrmlpad編輯軟件，將色彩、三維文字標注及各結(jié)構(gòu)組裝等功能通過各個程序?qū)崿F(xiàn)，并用Frontpage軟件建立了膝關(guān)節(jié)的三維解剖圖譜，此三維解剖圖譜包括了文字內(nèi)容、二維及三維圖像，便于學(xué)生理論與標本相結(jié)合、二維與三維相對照地學(xué)習(xí)解剖學(xué)知識，提高學(xué)習(xí)效率。 通過圖像重建、2D／3D圖像配準及相應(yīng)的圖像處理后，計算出股骨在G型臂X線機攝片位與CT掃描位之間的相對位移及角度。與CT掃描時的位置相比，攝X線平片時股骨下段前屈5.72°，內(nèi)翻1.02°，左旋13.22°。 通過點云重建、位置還原等步驟，可計算出激光三維掃描時與CT掃描時的膝關(guān)節(jié)位置的相對運動。與CT掃描時的位置相比，激光三維掃描時的股骨后伸6°，內(nèi)翻1.2°，右旋6.16°。激光三維掃描儀在各角度時的測試結(jié)果與KOHZU精密測試臺的測試結(jié)果相比，各角度時測試精度均可控制在0.1°以內(nèi)。利用SPSS10.0統(tǒng)計軟件進行單因素方差分析發(fā)現(xiàn)各角度組之間的誤差率無顯著性差異(P=0.206)。 對膝關(guān)節(jié)在體穩(wěn)定性測試系統(tǒng)與逆向工程運動還原系統(tǒng)導(dǎo)出的圖像分別與CT掃描時的位置計算其相對運動數(shù)據(jù)，經(jīng)配對樣本T檢驗后發(fā)現(xiàn)，膝關(guān)節(jié)在體穩(wěn)定性測試系統(tǒng)計算的相對運動與逆向工程技術(shù)計算的相對運動之間不存在顯著性差異(t=0.132，P=0.895)。對平移及旋轉(zhuǎn)各個軸單獨進行配對樣本T檢驗，發(fā)現(xiàn)除Z軸平移數(shù)據(jù)兩者存在顯著性差異(t=3.214，P=0.024)外，其余各個軸的平移及旋轉(zhuǎn)數(shù)據(jù)均不存在顯著性差異(P＞0.05))。將兩個配準系統(tǒng)導(dǎo)出的股骨、脛骨三維位置直接進行相對運動計算，可發(fā)現(xiàn)各個軸的配準誤差。其中在X軸的平均平移誤差最大，達到6.98mm；Z軸的平均旋轉(zhuǎn)誤差最大，達到了6.92°。 結(jié)論 本研究通過采集同一標本的CT和MRI數(shù)據(jù)，利用CT顯示精度高、對骨的顯示好的特點以及MRI對軟組織顯示好的特點，取長補短，對CT或MRI數(shù)據(jù)在個人計算機上進行圖像重建，并且結(jié)合先進的逆向工程技術(shù)，構(gòu)建了具有骨、關(guān)節(jié)軟骨、半月板及前／后交叉韌帶的膝關(guān)節(jié)。 本研究模型清楚顯示膝關(guān)節(jié)周圍血管的形態(tài)，三維直觀再現(xiàn)了Scapinelli描述的髕周動脈環(huán)結(jié)構(gòu)。重建的膝部骨—動脈三維模型，直觀再現(xiàn)膝部動脈走行、分布特點與常用手術(shù)入路之間的關(guān)系，并能測量其空間距離，有利于幫助判斷手術(shù)操作對血運的影響程度�？捎糜诟倪M解剖教學(xué)手段，同時利于直觀、形象地與臨床應(yīng)用要點相結(jié)合地開展教學(xué)，利于醫(yī)學(xué)生迅速、準確掌握局部解剖特點。 三維解剖圖譜在解剖學(xué)教學(xué)中可以方便教學(xué)，提高教學(xué)效率。重建的三維模型及二維圖像可以在普通的個人計算機上使用，幫助學(xué)生們直觀地理解并記憶各解剖結(jié)構(gòu)，提高學(xué)習(xí)效率。同時網(wǎng)頁版的三維解剖圖譜可以建立虛擬解剖學(xué)實驗室，實現(xiàn)解剖學(xué)的遠程教學(xué)。 本研究采用計算機圖像重建技術(shù)對患者膝關(guān)節(jié)進行三維重建，采用G型臂X線機采集患者膝關(guān)節(jié)骨結(jié)構(gòu)的運動信息，通過二維／三維(2D／3D)圖像配準技術(shù)，將二維動態(tài)的X線影像轉(zhuǎn)化為三維模型的仿真運動，建立膝關(guān)節(jié)在體穩(wěn)定性測試系統(tǒng)。 利用逆向工程技術(shù)可以實現(xiàn)膝關(guān)節(jié)的位置還原，具有以下優(yōu)點：1．實驗精度高，精度可達到0.1°；2．測量為非接觸式，對實驗影響較小。3．對于運動范圍較大的測試，可采用點云拼接的方法，在實驗精度不受影響的情況下實現(xiàn)穩(wěn)定性評價。 本研究建立的膝關(guān)節(jié)在體穩(wěn)定性測試精度與逆向工程運動運動還原系統(tǒng)的測試精度無顯著性差異，說明本系統(tǒng)可以用于膝關(guān)節(jié)的在體穩(wěn)定性研究。但精度不高，今后要注意實驗細節(jié)，改進實驗方法，進一步提高測試精度以更好地服務(wù)臨床。
[Abstract]:objective
In recent years, there are a lot of digital virtual research on knee joint injuries. Three-dimensional reconstruction of human knee joint is the basis of these studies. But because of the complexity of knee joint structure and the limitation of imaging technology, these three-dimensional reconstruction usually only reconstruct the bone structure of knee joint. The reconstruction of bone, meniscus, anterior/posterior cruciate ligament and articular cartilage of the knee joint were constructed by using 3D reconstruction technique and image registration technique.
In clinical knee surgery, local blood circulation should be protected as much as possible to avoid the occurrence of local osteonecrosis. Because of the complex local blood supply of knee joint, traditional imaging can not observe the blood vessels. Angiography can better display the local blood vessels, but its quality is affected by the flow rate of contrast media, image segmentation and processing technology. In order to overcome the above shortcomings, CT scans were performed after perfusion of arteriography to reconstruct the bone and blood vessels of the knee joint. The purpose of this study is to accurately and completely display the course and distribution of the arteries of the knee joint, and to provide a clear, accurate and valuable imaging method for the teaching and scientific research of the surgery and anatomy of the knee joint.
Human anatomy is a morphological science. At present, teaching methods based on two-dimensional images is very difficult. With the continuous improvement of computer graphics and image technology and the continuous progress of virtual human research, there have been a variety of three-dimensional image reconstruction methods, but most of the three-dimensional image display needs professional software, which is expensive and expensive. In this study, we will use advanced VRML language to compile related programs and establish web-based three-dimensional browsing web pages of knee joints to lay the foundation for the completion of three-dimensional anatomical maps.
Although the stability test of knee joint has been used to evaluate the motion function of knee joint, it is impossible to simulate the real motion of knee joint in vitro and to obtain the motion information of bone structure in body motion test, so it is impossible to obtain the accurate stability data of knee joint. In this study, three-dimensional image reconstruction, two-dimensional/three-dimensional image registration and image processing technology will be used to explore the feasibility of establishing a knee joint stability evaluation system in vivo.
Reverse engineering is to construct a CAD model based on an acquired physical model, and to approximate, modify and perfect the physical model by adjusting the relevant parameters, and then apply these CAD models to product analysis and manufacturing. This paper discusses the method of knee joint motion reduction in vitro using reverse engineering technology, and calibrates the measuring accuracy of the method of knee joint motion reduction using laser three-dimensional scanner.
The in vivo stability test system of knee joint was constructed by computer three-dimensional reconstruction technology, two-dimensional/three-dimensional image registration technology and image processing technology.
Materials and methods
One fresh adult knee joint specimen was scanned by CT with a total thickness of 387 layers (0.299 mm). Then MRI was performed with a total thickness of 64 layers (1.497 mm). The three-dimensional reconstruction software Mimics and reverse engineering software Geomagic were used to reconstruct and register the images, and the bone, cartilage, ligament and meniscus of the knee joint were constructed.
A fresh adult complete lower limb specimen was perfused with a suitable concentration of filler. Then CT scan was performed with a thickness of 0.499 mm. A total of 671 layers of the knee joint were taken for the study. Three-dimensional reconstruction of the bone and blood vessels of the knee joint was performed with Mimic, and the arteries were transparently displayed with Mimic and 3ds max, respectively. Comparison of effects.
BS Contact VRML 6.1 and Vrmlpad are used as VRML display plug-in and program editing software. The structure of each part of the knee joint which has been reconstructed in three dimensions is transformed into WRL file, and VRML programming and web page making are carried out.
The knee joints of healthy adult volunteers were scanned by CT, and the right and lateral X-ray plain films were collected when the volunteers maintained a certain posture. The three-dimensional reconstruction of the knee joints was carried out. The virtual X-ray system was established in the software to reproduce two X-ray plain films with right angles. The relative displacement and angle change between the two positions are calculated.
The position information of the embedding block was collected by laser three-dimensional scanner after embedding on the lower part of the human knee joint specimen. The position of the embedding block was restored by Geomagic software. The markers were fixed on the precision testing platform of KOHZU with the accuracy of 0.01 mm and 0.01 degrees. Its motion parameters are calculated to detect the accuracy of the kinematic reduction method in reverse engineering.
Three human knee joint specimens were examined with a G-arm X-ray machine and three-dimensional point cloud information of the embedding block was scanned with a laser three-dimensional scanner. The relative motion of the knee joint model constructed from CT images was calculated and the paired sample T-test was performed. The relative motion data of the knee joint were derived from the virtual X-ray radiography system and reverse engineering technology system, and the differences between the two axes of translation and rotation were observed by descriptive statistics.
Result
CT images were imported into the 3D reconstruction software Mimics 10.01 to reconstruct the bone structure, and then the three-dimensional models of the bone were saved in STL file format in the computer. The three-dimensional models of meniscus, anterior/posterior cruciate ligament and articular cartilage were constructed by Mimics 10.01 software, and were imported into Geomagic software to modify the structure slightly. Articular cartilage needs to be re-introduced into CT images to smooth its edges. Finally, a three-dimensional knee joint model with bone, meniscus, anterior/posterior cruciate ligament and articular cartilage is established.
Using CT scan data after perfusion of angiography, a three-dimensional model of bone-vessel structure of the knee joint was successfully constructed with Mimics software. The relationship between bone structure and blood vessels of the knee joint and their three-dimensional shape were clearly displayed. The local blood supply of the joint, such as the vascular network around the patella, could also be clearly displayed. Comparatively speaking, the operation of Mimics software is relatively simple, but the transparency of software settings can not be adjusted arbitrarily. And the operation of 3DS MAX software is relatively complex, especially for large files, the process of importing 3ds Max is relatively slow, but it can be tolerated. Transparency and strong visual design ability.
Through Vrmlpad editing software, the functions of color, three-dimensional character labeling and structure assembling are realized by each program, and the three-dimensional anatomical map of knee joint is established by Frontpage software. The three-dimensional anatomical map includes text content, two-dimensional and three-dimensional images, which is convenient for students to combine theory with specimen, and two-dimensional and three-dimensional comparative geology. Learn anatomy knowledge and improve learning efficiency.
After image reconstruction, 2D/3D image registration and corresponding image processing, the relative displacement and angle of femur between the X-ray position of G-arm X-ray machine and CT scanning position were calculated.
The relative motion of the knee joint position during laser three-dimensional scanning and CT scanning can be calculated by means of point cloud reconstruction and position restoration. Compared with the position of CT scanning, the femur of laser three-dimensional scanning is extended 6 degrees, inverted 1.2 degrees and dextral 6.16 degrees. Compared with the test results, the test accuracy can be controlled within 0.1 degree at all angles. The single factor analysis of variance using SPSS10.0 statistical software showed that there was no significant difference in the error rate between the different angle groups (P = 0.206).
The relative motion data of the knee joint in vivo stability testing system and reverse engineering motion restoring system were calculated by the position of the knee joint in CT scanning, respectively. After the paired sample T test, it was found that there was no significant difference between the relative motion calculated by the in vivo stability testing system and the relative motion calculated by reverse engineering technology. The results of T-test showed that there was no significant difference (P > 0.05) in the translation and rotation data of the other axes except the Z-axis translation data (t = 3.214, P = 0.024). The three-dimensional position of femur and tibia derived from the two registration systems was directly entered. The registration error of each axis can be found by calculating the relative motion. The average translation error of X axis is the largest, reaching 6.98 mm, and the average rotation error of Z axis is the largest, reaching 6.92 degrees.
conclusion
In this study, CT and MRI data of the same specimen were collected, and the characteristics of CT display precision, bone display and MRI display of soft tissue were used to reconstruct the image of CT or MRI data on personal computer. Combined with advanced reverse engineering technology, bone, articular cartilage, meniscus and anterior were constructed. The knee joint of posterior cruciate ligament.
The model clearly shows the shape of the blood vessels around the knee joint and visually reproduces the structure of the peripatellar artery ring described by Scapinelli in three dimensions. It can be used to improve the teaching methods of anatomy, and it is also helpful to carry out teaching visually and vividly in combination with the key points of clinical application.
The reconstructed three-dimensional model and two-dimensional image can be used in ordinary personal computer to help students understand and memorize anatomical structures intuitively and improve learning efficiency. Room for remote teaching of anatomy.
In this study, computer image reconstruction technology was used to reconstruct the knee joints of patients in three-dimensional, G-arm X-ray machine was used to collect the motion information of the knee joint bone structure, and through two-dimensional / three-dimensional (2D / 3D) image registration technology, the two-dimensional dynamic X-ray images were transformed into three-dimensional model of the simulation movement, and the in vivo stability testing system of the knee joints was established. Unification.
Reverse engineering technology can be used to restore the position of the knee joint, with the following advantages: 1. high experimental accuracy, accuracy can reach 0.1 degrees; 2. measurement is non-contact, less impact on the experiment.
【學(xué)位授予單位】：第一軍醫(yī)大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2007
【分類號】：R322

【引證文獻】

相關(guān)碩士學(xué)位論文 前1條

1 王蕊;基于RE/RP技術(shù)的標準膝關(guān)節(jié)假體的研究與開發(fā)[D];天津理工大學(xué);2011年

本文編號：2234523