Research Of Knee-joint-related Digital Anatomy And Establishment Of Stability Test System Of Knee Joint In Vivo

ObjectivesIn recent years there were so many digitized virtual researches of knee jointinjury. 3D reconstruction was the foundation in these researches. However, becauseof the complexity of the knee joint structure, as well as the limitation of radiologicaltechnology, these 3D reconstructions only reconstructed the bone structure of kneejoint as usual. This research would combine the images such as CT, MRI with 3Dreconstruction and registration technology. At last the structure knee joint such asbone, meniscus, ACL/PCL and articular cartilage were reconstructed to accumulatethe experience for development of digital medical research of knee joint.Protection of the blood circulation was requested as far as possible in clinicalknee surgery to avoid the appearance of regional osteonecrosis. Because of thecomplexity of blood supplies of knee joint, the traditional radiological inspectioncould not observe vessels. Angiography was the good demonstration of distribution ofvessels. However, it was influenced by the flow rate control of contrast medium andimage segmentation technology. In order to overcome above deficiency, this researchwould perform the CT scanning of knee which the artery was perfused by contrastmedium. The bone and artery of knee joint were reconstructed separately todemonstrate the direction and distribution of artery of knee joint accurately andthoroughly.The anatomy is a morphology science. It was quite difficult teaching anatomybased on the two-dimensional picture. With the continual flawless, as well as the continual advancement of digital human studies, several kinds of 3D reconstructionmethods had appeared. However, the display of 3D models all needed the specializedsoftware. Limited the expensive price and also the intellectual property rightsquestion, it was unable to use in 3D anatomical atlas for the popularized application.The research would use the advanced VRML language to compile the relatedprocedure, and finally establish 3D anatomical atlas of knee joint based on the web. Itwould lay the foundation for 3D anatomical atlas research.Although the stability test of knee joint had widely applied in the evaluation thefunction of knee joint. However, the existing exsomatize research of knee joint wasunable to simulate the proper motion of knee joint, and the existing movement test invivo could not obtain movement information of bone structure. Therefore it could notobtain the accurate stability data of knee joint. And also it could not carry out theearly diagnosis of micro damage and thorough research about prevention and curemethods of knee joint. The research would combine 3D reconstruction technology,2D/3D registration technology and image processing technology to investigate thefeasibility of establishment of stability test system of knee joint.The reverse-engineering was the progress that firstly the CAD model wasobtained based on the entity model, secondly the model was rectified by adjustingrelatively parameter, and finally the model was used for the analysis and manufactureof production. Based on the high accuracy of reverse-engineering measurement whichperformed by 3D laser scanner, the research would discuss the method of movementsimulation of knee joint under the exsomatize condition using the reverse-engineeringtechnology. And then the accuracy of this method was standardized.The stability test system of knee joint in vivo was established using 3Dreconstruction technology, 2D/3D registration technology and the image processingtechnology. But because the registration progress was performed by manualregistration, the research would study the precision stability test system of knee jointin vivo which,compared with movement recovery method using reverse-engineeringtechnology.Methods387 slices of CT data with 0.299 mm interval were harvested by CT scanningusing 1 adult knee joint specimen. And then 64 slices of MRI data with 1.497mm interval were harvested by MRI scanning using the same knee joint specimen. Thestructures such as bone, atricular cartilage, meniscus and ACL/PCL of knee joint werereconstructed using 3D reconstruction software Mimics and reverse-engineeringsoftware Geomagic 6.The artery of 1 whole lower limb was perfused by the proper density bulkingagent. And then the CT scan was performed. 671 slices with 0.499 mm interval wereselected for the study. The bone and artery structures of knee were 3D reconstructedseparately using Mimics. The method and its effect of transparence demonstrationwere carried out using Mimics and 3ds max methods.BS Contact VRML 6.1 and Vrmlpad were selected as VRML demonstrationplug-in unit and procedure edition software. The 3D reconstruction knee joint weretransferred to "wrl" document. And the VRML programming and the homepagemanufacture were carried on.CT data of knee from a healthy adult specimen was harvested. And the twoorthogonal images of the knee were captured using a G-arm fluoroscope at a certainflexion angles. The 3D computer model of knee was reconstructed. A virtualfluoroscope was then created in solid modeling software and was used to reproducethe relative positions of the orthogonal images. Two virtual cameras in the softwarewere used to represent the X-ray sources. The 3D computer model of the femur andtibia was then introduced into the virtual fluoroscope respectively and was projectedonto the orthogonal images by the two virtual cameras. By matching the projectionsof the knee model to the orthogonal images of the knee, the relatively displacementand angle of knee were determined.A cadaver specimen of knee was embedded by Polymethylmethacrylate(PMMA) at the superior and inferior ends. The superior and inferior ends positionswere got by a 3D laser scan with the type of point clouds. With the software themovement parameters of knee joint were calculated. The scales were fixed on theKOHZU precise measurement set whose precision could get the degree of 0.01mmand 0.01°. The rotation positions were gathered by the 3D laser scan, and the angleswere calculated by the software. Compared with the angle measured with KOHZUmeasurement set, the precision of the technology was got.3 human knee joint specimens were used in the study. And the two orthogonal images of the knee were captured using a G-arm fluoroscope at a certain flexionangles. At the same time, the superior and inferior ends positions were got by a 3Dlaser scan with the type of point clouds. With the software the movement parametersof knee joint were calculated. The above performance was performed twice to everyknee joint specimen in the random angle. The positions were recovered by the virtualradiology system and reverse-engineering movement recovery system separately. Thedata of relative motion was calculated between these two positions with the CTposition. Paired-sample T test were carried on for the statistics. And descriptivestatistics was carried on the data which extracted from these two positions directly.ResultsWith the merit of CT to demonstrate the bone structure well, segmentation and3D reconstruction of bone were performed using Mimics 10.01 in CT data. And thenstl format document of bone was saved in the disk of computer. The registration wasperformed after the 3d models which reconstructed by MRI and CT separately wereimported in the software Geomagic to transform the coordinate of CT data whichenabled the CT model be fit for the MRI image. With the merit of MRI todemonstrate the soft tissue structure well, segmentation and 3D reconstruction ofmeniscus, articular cartilage and ACL/PCL were performed using Mimics 10.01 inCT data. The 3D models of these structures were imported to Geomagic software to3D modification to let the structure smooth. The articular cartilage models wereimported into CT image again to testify the region of the segmentation. Finally, the3D knee joint model with bone, meniscus, ACL/PCL and articular cartilage wasreconstructed.3D models of bone-blood vessel structures were constructed using theangiography-irrigated CT scanning data. It could demonstrate the clearly positionrelations and 3D morphology. And it could also clearly demonstrate the regionalblood supply, such as peri-patella artery link. The transparence demonstrations werecarried out using both Mimics and 3ds max methods. Both demonstration effecst isclose. Compares says, operation of using the Mimics software was quite simple, butthe software could not adjust willfully. And 3ds the max software operation isrelatively complex some. Furthermore, if the capacity big document, the process of3ds max would be very slow. But it could establish the transparency willfully, and had the strong visualization designed capacity.Through the Vrmlpad edition software, functions such as color, 3D wordlabeling and various structures assembly were realized through each procedure. Andthe homepage of 3D anatomical atlas of knee joint was established with theFrontPage software. The 3D anatomical atlas included the writing content, 2D and 3Dimages which were easy for students to study in theory and the specimen combination,2D and 3D images comparison to enhance the study efficiency.After 2D/3D registration, the relatively displacement and angle were calculated.Compared with the CT scan position, the angle of fluoroscope position of lower partof left femur was 5.72°flexion, 1.02°inversion and 13.22°left rotation.The point clouds were imported into the software. With the register andtransform function of the software, the 3D position of scales were got. Comparedwith the CT scan position, the angle of position of femur which scanned by 3D laserscanner was 6°extension, 1.2°inversion and 6.16°right rotation. And the precision ofthe test technology could get the degree of 0.1°. There was no difference from theerrors among the different angle groups.The relative movement data were calculated by the positions derived from thestability test system and the reverse-engineering movement recovery systemseparately vs the position of CT data. After paired-sample T test, there was nodifference between the relative motion of virtual X radiology system vs CT data andthe relative motion of reverse-engineering movement recovery system separately vsthe position of CT data (t=-0.132, P=0.895). And then the paired-sample T testswere performed in every freedom degree, it was found there were no difference in thefreedom degrees (P＞0.05) except for the translation in Z axis (t=3.214, P=0.024).The relative motion was calculated between the positions derived from virtual Xradiology system and CT scan data directly. It was found that the error of every axis.In these errors, the most translation error was the X axis error which achieves6.98mm, and the most rotation error was the Z axis error which achieves 6.92°.ConclusionsCombined the merits of CT demonstrating bone structures and MRIdemonstrating soft tissue, the 3D models are reconstructed on the personal computerusing the 3D reconstruction technology, as well as the advanced reverse-engineering technology. Finally several structures such as bone, articular cartilage, meniscus andACL/PCL are reconstructed.The research demonstrates clearly the morphology of blood vessel around theknee joint. The reconstruction models of bone-artery of knee joint reappears thecourse and distribution of artery of knee joint and the relationship with commonlyused operative approach. And it is a great help for judgment surgery to the bloodsupply with the space length measurement. It can also improve the anatomicalteaching method. It is good for medical students to understand the anatomicalcharacter rapidly with the combination of the 3D reconstruction models and clinicalpractice point.The 3D anatomical atlas may facilitate the teaching in anatomy teaching andmay enhance the teaching efficiency. The 3D reconstructed models and 2D imagescan be used on the ordinary personal computer. It can help students to understand andto memory the anatomical knowledge intuitively. Simultaneously, the homepageversion of the 3D anatomical atlas may build the virtual anatomy laboratory. It canrealize the long-distance teaching of anatomy education.The research uses the image reconstruction technology and 2D/3D registrationtechnology and image processing technology to come true the movement recovery ofknee joint. Thus, the stability test system of knee joint in Vivo is established.The position recovery is come true using the reverse-engineering technology. Ithas the following merits: 1. Experimental precision is high. The precision mayachieve 0.1°. 2. The measurement method is non-contact type. It affects less to theexperiment. 3. Regarding the big scope movement test, it can conjunct the 3D cloudpoints to maintain the high precision.There is no difference of measurement precision from the stability test of kneein vivo and the movement recovery system using reverse-engineering technology. Itindicates that the established system in this study can use for the research of stabilitytest of knee joint in vivo. However, there is still a higher error. The further researchwill pay more attention to improve the experimental technique to increase the testprecision.