基于CTA影像大肠三维数字解剖模型构建及可视化的基础和临床初步应用研究
|
摘要
背景
现代科学技术的发展越来越体现出多门学科的交叉和渗透。集医学,生物力学,机械学,材料学,计算机图形学,数学分析,自动控制等多学科为一体的新型交叉研究领域-医学虚拟现实技术(Medical Virtual Reality Technology,MVRT)目前正在飞速发展,促使现代的医疗诊断和治疗的手段与观念正在发生着很大的变化。计算机辅助外科(Computer Aided Surgery,CAS)是一种基于计算机对大量数据信息的高速处理及控制能力,通过虚拟手术环境为外科医生从技术上提供支援,是使手术更安全、更准确的一门新技术。近年来,随着CT、MRI等图像诊断仪的发展,使计算机虚拟现实技术在医学中的应用得到了飞速的发展。计算机利用这些图像信息进行三维图像重建,为外科医生进行手术模拟、手术导航(navigator)、手术定位、制订手术方案提供了客观、准确、直观、科学的手段。自1994年Vining等首次报道螺旋CT仿真内窥镜以来,在国内外已进入临床试用阶段。但在临床实际工作中,其价值如何,尚需评估。
结、直肠癌是胃肠道常见的恶性肿瘤,近年来发病率明显上升,发病率仅次于胃癌和食管癌,全球结直肠癌每年新发病例数达94万,每年近50万人死于结直肠癌。死亡数居癌症死因第三位。手术治疗一直是直肠癌的主要治疗手段,然而迄今为止没有建立起令人满意的精细三维解剖结构模型供临床参考。随着计算机技术向影像学领域渗透和应用,CT、MRI等断层影像技术实现了数字化图像处理,医学图像信息的采集、储存、传输、处理、显示等发生了前所未有的变革。由此形成全新医学图像处理和诊断模式,诞生了仿真影像学。仿真影像学以仿真内镜(VE)为核心,包括各种数字化图像三维重建技术和后处理技术。利用CT、MRI等二维断层图像为数据源,应用计算机软件重建直观地显示人体器官,尤其是管腔器官内表面解剖及病变三维仿真图像,能够准确地直观反映人体三维结构,进行疾病诊断、立体定位、模拟手术等。大肠及周围结构是普通外科重要的解剖结构,基于CT、MRI工作站对断层影像处理主要用于疾病诊断,目的比较单一,对CT、MRI等断层影像有效信息挖掘不够深入。结直肠癌是胃肠道常见的恶性肿瘤,常规钡剂灌肠和纤维结肠镜是主要的检查手段。三维成像及CT仿真内镜(CT virtual endoscopy,CTVE)技术对结肠病变的研究逐日增多特别是64层螺旋CT以其高速度、更加薄层、更大范围的容积扫描,为这一技术的发展起到了很大的促进作用。
为此,我们基于不同类型CTA断层数据集建立了(1)直肠及周围结构解剖结构的三维可视化模型;(2)直肠肿瘤及周围解剖结构的三维可视化模型;(3)在个人电脑重建大肠体绘制和面绘制模型,大肠和周围解剖结构空间毗邻关系模型及大肠虚拟内镜,为深入挖掘断层影像有效信息做了有益尝试;(4)在个人电脑对10例经术后病理证实的结肠癌和结肠息肉患者进行了大肠虚拟内镜检查及大肠和周围解剖结构模型进行重建,并和CT三维成像与CT仿真结肠镜进行形态学观察,探讨其临床应用价值。
研究目的和方法
目的
1、寻求基于CTA断层图像重建直肠及周围结构数字模型及三维可视化的方法,简便、高效地建立直肠及周围结构三维模型。
2、研究CTA断层图像重建直肠肿瘤数字模型及三维可视化方法建立的模型为解剖教学、生理模拟、手术培训提供医疗教学平台。
3、寻求基于64层螺旋CT断层图像在个人电脑重建大肠三维数字模型及虚拟内镜检查方法。
4、探讨基于64层螺旋CT断层图像在个人电脑重建大肠三维数字模型及虚拟内镜检查方法,根据患者的数字化仿真模型,可制订个体化的诊断、术前规划和手术模拟,提高手术的安全性,降低手术并发症。
方法
1.研究对象:
第一部分基于CTA断层图像直肠及周围结构数字模型的重建及三维可视化研究研究对象为健康男性志愿者一名,27岁;
第二部分直肠肿瘤数字解剖模型构建及三维可视化研究对象为直肠肿瘤病人,男性,68岁;
第三部分基于64层螺旋CT断层图像的影像后处理:大肠三维重建及虚拟内镜研究对象为健康志愿者一名,男性,31岁;
第四部分基于64层CTA断层图像虚拟技术和CT仿真结肠镜临床应用研究研究对象:搜集2007年7～9月结肠癌和结肠息肉患者10例。其中结肠癌8例,男5例,女3例,年龄37～65岁,平均50.4岁;结肠息肉2例,男1例,女1例,年龄40～57岁。
2.检查前肠道准备:
检查前3日进流质饮食,检查前晚服蓖麻油30ml,当日禁食,检查前2小时作清洁灌肠,直到排出液澄清为止,术前15分钟肌注阿托品0.5mg。
3.CT扫描前空气灌肠使肠管充气膨胀:
使用1S-818E电脑遥控空气灌肠机灌肠。肛门及肛管涂抹液体石蜡,经肛门插入Foley氏18F气囊管(代肛管),插入深度为10 cm,注人20-30ml气体将气囊充起,防止气囊管从肛门脱出。X线透视观察下开始灌肠。以稳定低压4-5kPa持续向直肠内注入空气使肠管充气膨胀,固定维持压力在4kPa。
4.数据采集设备
64排螺旋CT-PHILIPS Brilliance64(荷兰)。高压注射器采用MEDRAD双筒高压注射器(美国),图像后处理工作站为PHILIPS Brilliance64层螺旋CT自带的Mxview工作站。
5.CT扫描
空气灌肠成功后行CTA动脉造影,亚毫米状态下高分辨力容积扫描,常规平扫时患者取仰卧位,头足方向,扫描范围从胸12至股骨中上部分,扫描条件120KV、300mAs;采用0.625×64排探测器组合,以层厚0.5mm、间隔0.5mm,螺距(Pitch)0.809mm,球管旋转一周时间0.5s,选择软组织窗,准直器宽41.40mm×0.625 mm,开始常规平扫。动态CT增强扫描:平扫完成后行动脉期及静脉期连续跟踪扫描,经肘静脉注射对比剂(应用套管针),使用双筒CT高压注射器,注射速率5ml/s,所用对比剂为高浓度非离子型碘必乐370(370mgI/ml)或优维显370(370mgI/ml),剂量1.5ml/kg体重,对比剂注射完毕后以生理盐水50ml冲管,扫描技术条件同平扫。扫描完成后,应用增强原始数据进行1mm的薄层重建,并将图像数据传至Mxview工作站。
6.CT薄层扫描数据的采集
在Mxview诊断工作站上,利用光盘刻录全部的数据,其中包括平扫期、动脉期、静脉期数据。格式为DICOM(Digital Imaging and Communications inMedicine)3.0。
7.建立直肠及周围结构数字模型
Mimics软件分别读入Dicom格式的动脉期及静脉期CT断层图像,软件自动设定CT原始扫描参数,层间距均为0.5mm,动脉期螺距0.809mm,静脉期螺距0.824mm,以.MCS文件格式保存。在动脉期数据集分别重建骨骼、肠道、动脉、皮肤、膀胱及前列腺三维模型,根据不同研究的需要还分别建立动脉系统及肿瘤供血动脉、直肠肿瘤等。在静脉期数据集重建输尿管三维模型,以STL格式导出在动脉期数据集行三维模型立体配对。
8.肠管面绘制和外显镜
Mimics软件读入Dicom格式CT断层图像,软件自动设定CT原始扫描参数,每层原始图像灰度值以内插值处理,达到亚体素的精度。调整窗位窗宽,对比度达到最佳视度。空气灌肠使肠管充气膨胀,肠道面绘制模型按空气阈值建模,界定阈值在-1024至-1024亨氏单位,并膨胀1—2个像素达到选择像素充满全部肠腔,区域增长工具分割出直肠及乙状结肠,升结肠,横结肠及部分小肠,去除冗余数据,经3D计算建立肠道的三维数字模型。面绘制模型以三角面片显示。在动脉期数据集重建骨骼系统,动脉系统及皮肤,前列腺三维模型。在110层及82层二维图像分别手动分割出膀胱及前列腺,3D计算建立膀胱及前列腺三维数字模型。在静脉期数据集根据残余造影剂浓度重建输尿管三维模型,阈值在220-1641亨氏单位,经区域增长3D计算建立输尿管三维模型,以STL格式导出在动脉期数据集根据输尿管和双肾及膀胱解剖结构关系行三维模型立体配对。
9.虚拟内镜
对基于CT断层图像建立的肠管面绘制模型,用虚拟内窥镜技术可对肠腔解剖和病变进行观察,观看三角面片内部信息能够获得有效肠道内解剖图像和诊断信息。可在二维平面图像进行大体观察来指导内镜重点观察目标,给合多平面裁剪功能可以综合分析。在肠腔内“漫游”可观察全结肠。对观察到的病变对发现病变后,变换视角、视距,甚至反向观察。返回到二维平面图像测量其最大径作为直径。
10.CT三维成像和仿真内镜检查
使用Philips公司Brilliance 64 CT用5 mm层厚完成全腹部容积扫描,在后处理工作站(Extended Brilliance workspace)重建层厚0.5mm,间隔0.5 mm,使用容积再现(volume rendering,VR)和内镜软件,通过合理调整阈值获取三维图像及内镜图像。对表面遮盖显示(shaded surface display,SSD)和透明显示(RaySum)图像进行大体观察来指导内镜的重点观察目标,结合轴面图像及多平面重建(MPR)综合分析。
结果
1.基于CTA断层图像直肠及周围结构数字模型的重建及三维可视化研究成功建立了直肠及周围结构三维数字模型,对腰椎、骨盆及股骨等骨骼结构,腹主动脉及髂内动脉及髂外动脉及分支、皮肤、直肠及乙状结肠、升结肠、横结肠及部分小肠、膀胱、前列腺、输尿管分别建立数字模型。
建立的数字模型几何外型逼真,具有较佳的视觉效果,可以多彩色、透明或任意组合显示,整体清晰、实体感强。通过不同平面的切面可以观察各组件的内部结构关系,能够任意角度的旋转及缩放观察,动态显示可制作成AVI格式电影存储、输出,画面清晰流畅。
2.直肠肿瘤数字解剖模型构建及三维可视化
建立直肠及直肠肿瘤结构三维数字模型,真实再现了肿瘤在直肠内的确切位置,对腰椎,骨盆及股骨等骨骼结构,腹主动脉,髂内动脉,髂外动脉及分支,皮肤,直肠,肛管,直肠肿瘤及乙状结肠,升结肠,横结肠及部分小肠,膀胱,前列腺,输尿管分别建立数字模型。肿瘤与周围其他组织的位置关系,浸润情况一目了然。
3.基于64层螺旋CT断层图像的影像后处理:大肠三维重建及虚拟内镜
基于64层螺旋CT断层图像用虚拟内窥镜技术对肠腔的解剖和病变进行了观察。能成功内镜样展现胃肠道的任何部位与转角处,并能从狭窄、梗阻处两端观察官腔的解剖和病变,可靠地显示了病变部位和大小。
4.基于64层CTA断层图像虚拟技术和CT仿真结肠镜临床应用研究
成功建立了结肠面绘制三维数字模型,效果良好。用虚拟内镜对大肠解剖和病理结构进行了观察,其肿瘤形态与术中所见几乎完全一致。
结论
1.直肠及周围结构的三维重建方法应用可视化技术利用人类的视觉特性,运用计算机图像学技术,将二维断层图像序列进行处理,重构出具有立体效果的空间三维形态结构图像。基于空气灌肠造影及CTA血管造影技术,可更精确还原肠管及动脉,输尿管在活体内的原有真实结构形态,可简便、高效地建立直肠及周围结构三维模型。其创新性及意义有(1)、制定手术方案:能够利用图像数据,帮助医生合理、定量地制定手术方案,对于选择最佳手术路径、减小手术损伤、减少对临近组织损害、提高肿瘤定位精度、使手术安全性增加,而手术的风险性降低、并发症减少,对执行复杂外科手术和提高手术成功率等具有十分重要的意义。(2)、增进医患交流,减少医疗纠纷:医生可以用重建的三维模型对病人及家属讲解病情及手术情况,使患者及家属对病情及风险有直观的了解。(3)、降低手术费用:由于医生有较好的手术预案,能够缩短病人的恢复周期、降低病人和医院的开支。
2.基于CT断层图像的虚拟内镜技术源于螺旋CT连续扫描获得的容积数据,和传统的内窥镜技术相比,虚拟内窥镜和CT工作站仿真内窥镜有着共同独特的优点:(1)一种理想非侵入性的检查方法,病人无不适感,检查过程中不会产生穿孔、出血或感染等副作用;(2)常规的纤维镜存在视野的局限和无法评价腔外解剖与病变等缺点。虚拟内窥镜能够充分显示结肠的解剖形态以及病变部位,并能从狭窄、梗阻处两端观察肠腔的解剖和病变。除了解腔内情况外,结合三维图像还可了解肠壁以及腔外的情况,更有利于肿瘤的定性以及分期诊断,为临床制定手术方案提供依据;(3)可以针对不同要求,制定不同的漫游计划,任意地重复检查过程;(4)降低了医疗检查复杂性、危险性和医疗成本;(5)对绝大多数结肠肿瘤性病变可做出定性诊断。同时能完整地保存原始数据,具有很好的可重复性,更加灵活,不受入路限制,可任意到达所需观察的解剖部位。反复多次观察有利于小病灶以及多发性病灶的检查,可避免因人为因素导致的漏诊。
3.基于64层CTA断层图像虚拟技术比CT仿真结肠镜技术优势:能更精确还原骨骼,肠管及动脉,输尿管在活体内的原有真实结构形态,简便、高效地建立大肠及周围结构三维模型。以三维可视化的方式从不同角度进行展示,结合传统教材的图像与实体解剖,能更好地理解真实的三维人体结构,是对传统教材中的完善和补充。三维重建图像利于整体直观地显示病变,帮助明确诊断并指导手术。在可视化技术的基础上可以进一步实现计算机模拟及手术规划。肠管柔软易变形,离体后不易定形,其在人体的真实外形不易理解,本研究以三维可视化的形式,通过任意角度的旋转,全方位显示大肠及其周围结构,便于对周围结构的观察理解,对理解掌握直肠周围结构的解剖关系有极大帮助。因此,与结肠三维成像和CT仿真内镜相比,基于64层CTA断层图像虚拟技术能达到与CT仿真内镜结合三维成像同样的敏感性和特异性,加上各种组织三维重建技术可以提供比仿真内镜更丰富的信息,其应用前景十分广阔。
建立的模型为解剖教学、生理模拟、手术培训提供医疗教学平台。临床医师根据患者的数字化仿真模型,可制订个体化的诊断、术前规划和手术模拟,提高手术的安全性,降低手术并发症,推动普通外科的发展。
Background
The development of modern science and technology more and more reflects interdisciplinary intersection and penetration. The new interdisciplinary studies field-Medical Virtual Reality Technology (Medical Virtual Reality Technology, MVRT) which integrates medicine, biomechanics, mechanics, hylology, computer graphics, mathematical analysis, automatic control and the like is developing rapidly at present, bringing about rapid change on modern medical diagnosis and therapeutic methods and concepts. Computer Aided Surgery (Computer Aided Surgery, CAS) is a new technology which makes the surgery safer and more accurate through providing technical support for surgeons by virtual surgical environment. CAS is based on the high-speed processing and controlling capacity on a vast amount of data information of the computer. In recent years, with the development of image diagnostic apparatus, such as CT, MRI and the like, the application of computer virtual reality technology in medicine has developed rapidly. The computer provides objective, accurate, intuitive and scientific methods for surgery simulation, surgery navigator, surgery orientation and surgery planning through using the image information to reconstruct the three-dimensional images. Since Vining and so on have firstly reported the helical CT virtual endoscope in 1994, the helical CT virtual endoscope has entered in the stage of clinical trial. However, its value in clinical practical work needs to be estimated.
Colorectal cancer is the common malignant tumour in gastric bowel path. In recent years, the disease incidence has increased obviously. The disease incidence is only second to gastric cancer and esophageal cancer. The number of new cases of colorectal cancer every year around the world reaches 940,000. Nearly 500, 000 people die of colorectal cancer every year. Mortality is the third of cancer-related death. Surgical treatment is always the main treatment of rectum cancer. So far, the satisfactory fine three-dimensional anatomical structural model for clinical reference has not been established. The computer technology has been applied in imageology field, so slice image technology, such as CT, MRI and the like has realized digital image processing. Collection, storage, transmission, processing, display and so on of the medical image information have changed greatly. Thus, a new medical image processing and diagnostic mode is formed and the virtual imageology is born. The virtual imageology consists of various digital image three-dimensional reconstruction techniques and after-treatment techniques with virtual endoscopy (VE) as the core. The virtual imageology uses the two-dimensional slice images, such as CT, MRI and the like as the data source and utilizes the computer software reconstruction to directly display human organs, especially anatomical images of inner surfaces of lumen organs and three-dimensional virtual images of pathological changes. The virtual imageology can accurately and directly reflect the three-dimensional structures of human body and carry out disease diagnosis, stereoscopic localization, surgical simulation and so on. Large intestine and the surrounding structures are important anatomical structures of the general surgery. To slice image processing, CT and MRI workspace is mainly used for disease diagnosis, the purpose is single. Effective information mining of the slice image, such as CT, MRI and the like is not thorough. Colorectal cancer is the common malignant tumour in gastric bowel path, the regular barium enema and fibercoloscope are the main detection methods. Increasing researches on colonic pathological changes by three-dimensional imaging technology and CT virtual endoscopy (CT virtual endoscopy, CTVE) technology, especially 64-slice helical CT and the volume scanning with high-speed, thinner slice, wider range have largely promoted the development of this technology.
Thus, based on CTA slice data sets of different types, we (1) have established the three-dimensional visualization models of the anatomical structures of rectum and the surrounding structures; (2) have established the three-dimensional visualization models of rectal tumor and the surrounding anatomical structures; (3) have reconstructed the volume rendering model and the surface rendering model of large intestine, the spatial adjacent relation model of large intestine and the surrounding anatomical structures and large intestine virtual endoscope in personal computers and have tried to mine the effective information of the slice image and (4) have carried out the large intestine virtual endoscopy and reconstructed the structural models of large intestine and the surrounding structures on 10 cases of colon cancer patients and intestinal polyp patients in personal computers, viewed morphology through CT three-dimensional imaging and CT virtual colonoscope and discussed the clinical application value.
Study objectives and methods Purpose
1. Seek for the method for reconstructing the digital models of rectum and the surrounding structures and for three-dimensional visualizing based on CTA slice images.
2. Research the method for reconstructing the digital model of rectum cancer tumour and for three-dimensional visualizing based on CTA slice images.
3. Seek for the method for reconstructing the three-dimensional digital models and for virtual endoscope detecting in personal computers based on 64-slice helical CT slice images.
4. Discuss the method for reconstructing the three-dimensional digital models and for virtual endoscope detecting in personal computers based on 64-slice helical CT slice images.
Method
1. Study object:
The first part The study object of reconstruction of the digital models of rectumand the surrounding structures and three-dimensional visualization based on CTAslice images is a healthy male volunteer who is 27-year-old.
The second part The study object of the construction of the digital model ofrectum cancer tumour and three-dimensional visualization is a male rectum tumourpatient who is 68-year-old.
The third part Image after-treatment based on 64-slice helical CT slice images: thestudy object of the three-dimensional reconstruction of large intestine and virtualendoscopy is a healthy male volunteer who is 31-year-old.
The fourth part The study object of the virtual technology based on 64-slice helicalCT slice images and the CT virtual colonoscope clinical application research: collect10 cases of colon cancer patients and intestinal polyp patients during July toSeptember in 2007. The colon cancer patients are 8 cases, wherein, 5 cases are maleand 3 cases are female, who are 36 to 65 years old. The intestinal polyp patients are 2cases, wherein, 1 case is male and 1 case is female, who are 40 to 57 years old.
2. Bowel preparation before testing
Eat liquid food 3 days before testing. Eat 30ml caster oil the night before testing. Fast on the day of testing. Carry out cleansing enema 2 hours before testing till the discharge liquor is clear. Intramuscularly inject 0.5mg atropine 15 minutes before operation.
3. Carrying out air enema to inflate intestinal canal before CT scanning
Use 1S-818E computer to remote-control an air enemator to carry out enema. Smear liquid paraffin on anus and anal canal. Insert Foley 18F sac duct (anal tube) per anus. The depth of insertion is 10cm. Inject 20-30ml gas to inflate the sac in order to prevent the sac duct from falling out. Begin to carry out enema under X-ray perspective observation. Continuously inject air into rectum with the stable 4-5kPa low pressure to inflate intestinal canal. Keep the pressure at 4kPa.
4. Data acquisition equipment:
64-slice helical CT-PHILIPS Brilliance64 (Holland) is used. The high-pressure injector uses MEDRAD dual-tube high-pressure injector (America). The image after-treatment processing workspace is attached Mxview workspace of PHILIPS Brilliance64-slice helical CT.
5. CT scanning:
Carry out CTA arteriography and high-resolution volume scanning under submillimeter state after the air enema is successful. When carry out routine scanning, the patient is in supine position. The scanning area is from twelfth thoracic vertebra to the middle upper part of thighbone along head-foot direction. The scanning condition is 120KV, 300mAs. 0.625×64-slice detector combination is used. Slice thickness is 0.5mm, interval is 0.5mm and pitch is 0.809mm. Time of a pipet rotating one circle is 0.5s. Soft tissue window is selected. Width of the collimator is 41.40mm×0.625 mm. Begin to routine scan. Dynamic CT enhanced scan: carry out arterial phase and venous phase continuous tracking scanning after the routine scanning is finished. Inject a contrast agent (using trocar) per cubital vein. Use a dual-tube CT high-pressure injector. The injection rate is 5ml/s. The used contrast agent is high-concentration non-ionic Iopamiro 370 (370mgI/ml) or Ultravist 370 (370mgI/ml). The dosage is 1.5 ml/kg weight. Use 50ml normal saline solution to rinse the tube after injecting the contrast agent. The scanning condition is the same with the routine scanning. Use the enhanced original data to reconstruct 1mm lamel after finishing scanning and transmit the image data to Mxview workspace.
6. Collection of CT lamel scanning data: at Mxview diagnosis workspace, use a compact disk to burn all data, comprising routine scanning phase, arterial phase and venous phase data. The format is DICOM (Digital Imaging and Communications in Medicine)3.0.
7. Construct digital models of rectum and the surrounding structures
Mimics software respectively reads in arterial phase and venous phase CT slice images of Dicom format. The software automatically sets CT original scanning parameters. The interlamellar spacing is 0.5 mm, the arterial phase pitch is 0.809mm and the venous phase is 0.824mm. The parameters are stored by .MCS file format. Respectively reconstruct the three-dimensional models of bone, intestinal canal, artery, skin, bladder and prostate at the arterial phase data set. Respectively reconstruct arterial system, tumour artery, rectum tumour and so on according to different research needs. Reconstruct the three-dimensional model of ureter at the venous phase data set. Lead out the three-dimensional models in STL format and carry out the three-dimensional model stereoscopic pairing at the arterial phase data set.
8. Intestinal canal surface rendering and external microscope Mimics software reads in CT slice images of Dicom format. The software automatically sets CT original scanning parameters. Gray value of each original image is processed by interpolated value in order to reach the precision of subvoxel. Adjust the window position and the window width to make the contrast reach the optimum visibility. Carry out air enema to inflate the intestinal canal. Construct the intestinal canal surface rendering model according to the air threshold value. The threshold value is defined between -1024 and -1024 Hounsfield unit. Inflate 1-2 pixels to make intestinal cavity be full with the selected pixels. Regional growth tools divide rectum and sigmoid colon, ascending colon, transverse colon and part of small intestine. Remove the redundant data. Construct the three-dimensional digital model of rectum through 3D computation. Surface rendering model is displayed by tri-patch. Reconstruct the three-dimensional models of bone system, arterial system, skin and prostate at the arterial phase data set. Manually divide bladder and prostate at 110-layer and 82-layer two-dimensional images. Construct the three-dimensional digital models of bladder and prostate through 3D computation. Reconstruct the three-dimensional model of ureter at the venous phase data set according to the concentration of the residual contrast agent. The threshold value is between 220 and 1641 Hounsfield unit. Reconstruct the three-dimensional model of ureter through regional growth 3D computation. Lead out the three-dimensional models in STL format and carry out the three-dimensional model stereoscopic pairing according to the anatomical structural relation of ureter, double kidneys and bladder at the arterial phase data set.
9. Virtual endoscope
To the intestinal canal surface rendering model constructed based on CT slice image, virtual endoscopy can be used to anatomize intestinal cavity and observe pathological changes. Effective intestinal canal anatomical image and diagnosis information can be obtained through observing the internal information of the tri-patch. Generally observe the two-dimensional images to guide the important observing target of the endoscope. Comprehensively analyze the target through multi-plane cutting function. Roam in intestinal cavity to observe the whole colon. Change the visual angle and the viewing distance, even reversely observe when the observed pathological changes have occurred. Return to the two-dimensional plane image to measure the largest diameter as the diameter.
10. CT three-dimensional imaging and virtual endoscopy checking
Use Philips Brilliance 64 CT and 5mm thickness to finish the whole abdomen volume scanning. Reconstruct 0.5mm thickness and 0.5mm interval at the after-treatment workspace (Extended Brilliance workspace) to make volume rendering (volume rendering, VR) and endoscope software. Obtain the three-dimensional image and the endoscope image through rationally adjusting the threshold value. Generally observe shaded surface display (shaded surface display, SSD) and Raysum (RaySum) images to guide the important observing target of the endoscope. Comprehensively analyze the target through combining the axial plane image and multi-plane reconstruction (MRP).
Result
1. Reconstruction of the digital models of rectum and the surrounding structures and three-dimensional visualization research based on CTA slice images
Successfully construct the digital models of rectum and the surrounding structures. Respectively construct the digital models of bone structures, such as lumber, pelvis, thighbone and so on, abdominal aorta, internal iliac artery, external iliac artery, branch, skin, rectum, sigmoid colon, ascending colon, transverse colon, part of small intestine, bladder, prostate and ureter.
The constructed models have vivid geometrical appearance, better visualization effect, clear view and strong reality. The models can be displayed through several colours, tranparancy or arbitrary combinations. The internal structural relation between each component can be observed through the sections of different planes. The models can be rotated or observed at any angles. Dynamic display can be made into AVI format film to store and output. The pictures are clear and smooth.
2. Construction of the digital anatomical model of rectum tumour and three-dimensional visualization
Construct the three-dimensional digital models of rectum and rectum tumour, truly rendering the accurate position of the tumour in rectum. Respectively construct the digital models of bone structures, such as lumber, pelvis, thighbone and so on, abdominal aorta, internal iliac artery, external iliac artery, branch, skin, rectum, anal canal, rectum tumour, sigmoid colon, ascending colon, transverse colon, part of small intestine, bladder, prostate and ureter. The position relation between tumour and the surrounding tissues and the infiltration situation are clear.
3. Image after-treatment based on 64-slice helical CT slice image: large intestine three-dimensional reconstruction and virtual endoscopy
Observe anatomization and pathological changes of intestinal cavity through using the virtual endoscopy based on 64-slice helical CT slice image. Any part and corner in gastrointestinal tract can be successfully rendered. Anatomization and pathological changes of intestinal cavity can be observed from narrow and block part. Position and size of pathological change can be reliably displayed.
4. Virtual technology and CT virtual colonoscope clinical application research based on 64-slice helical CT slice image.
Successfully construct the colon surface rendering three-dimensional digital model.
Observe anatomization and pathological changes of large intestine through using thevirtual endoscopy. The tumour form is exactly the same with the form seen in theoperation.
Conclusion
1. The three-dimensional reconstruction method of rectum and the surrounding structures reconstructs the spatial three-dimensional images with stereoscopic effect through using visualization technology, human visual characteristic and computer graphics to process the two-dimensional slice image series. The original and actual structures of intestinal canal, artery and ureter in vivo can be more accurately reconstructed and the three-dimensional models of rectum and the surrounding structures can be simply and effectively constructed based on air enema and CTA angiography.
2. The virtual endoscopy based on CT slice image derives from the volume data obtained through helical CT continuous scanning. Compared with the traditional endoscopy, the virtual endoscope and CT workspace virtual endoscope have the same special advantages. (1) It is a perfect non-invasively check method. Patients are free of discomfort. There are no side effects of perforating, bleeding or infecting during the checking process. (2) The regular fiber scope has limited view and can not evaluate outer cavity anatomization and pathological changes. The virtual endoscope can fully display anatomization and pathological change position of colon and can observe anatomization and pathological changes of intestinal cavity from narrow and block part. Besides conditions in the cavity, the conditions of intestinal wall and out of the cavity can be known through combining with the three-dimensional images. It is better for determining tumour and stage diagnosis. The virtual endoscope provides evidence for clinic operation plan. (3) Different roaming plans can be planned out according to different needs and checking process can be arbitrarily repeated. (4) Medical check difficulty, danger and medical cost are reduced. (5) Most colon tumour pathological changes can be determined. At the same time, original data can be completely stored. The virtual endoscopy has better repeatability and flexibility and it is not limited by approaches and can reach the part which is needed to be observed arbitrarily. Repeated observation is good for checking small lesions and multiple lesions, avoiding missed diagnosis caused by human factors.
3. Compared with CT virtual colonoscope technology, the advantages of the virtual technology based on 64-slice CTA slice image are: it can more accurately reconstruct bone, intestinal canal, artery and ureter in vivo and simply and effectively construct the three-dimensional models of large intestine and the surrounding structures. The virtual technology based on 64-slice CTA slice image can display from different angles through the three-dimensional visualization method. It is good for learning the actual three-dimensional human body through combining with the images in the traditional teaching material and substantial anatomization. The virtual technology based on 64-slice CTA slice image is supplement to the traditional teaching material. The three-dimensional reconstructed images are good for directly displaying the pathological changes, determining diagnosis and guiding operation. Computer simulation and operation planning can be further realized on the basis of visualization technology. The intestinal canal is soft and easy to deform, and is hard to shape in vitro. The actual shape of the intestinal canal in the human body is hard to understand. The study displays large intestine and the surrounding structure in an all-round manner through the three-dimensional visualization and rotation at any angle, which is good for observing and understanding the surrounding structures and helpful for understanding the anatomic relation of the surrounding structures of rectum. Compared with colon three-dimensional imaging and CT virtual endoscopy, the virtual technology based on 64-slice CTA slice image can realize the same sensitivity and speciality of the CT virtual endoscopy and three-dimensional imaging. Additionally, various tissue three-dimensional reconstruction technologies can provide richer information than the virtual endoscopy, its application future is wide.
The constructed models provide a medical teaching platform for anatomic teaching, physical simulation and operation training. Clinic doctors can work out individual diagnosis, plan before operation and carry out operation simulation according to the digital simulated models of patients. Thus, operation safety is enhanced and complications of surgery are reduced. Development of general surgery is promoted.
现代科学技术的发展越来越体现出多门学科的交叉和渗透。集医学,生物力学,机械学,材料学,计算机图形学,数学分析,自动控制等多学科为一体的新型交叉研究领域-医学虚拟现实技术(Medical Virtual Reality Technology,MVRT)目前正在飞速发展,促使现代的医疗诊断和治疗的手段与观念正在发生着很大的变化。计算机辅助外科(Computer Aided Surgery,CAS)是一种基于计算机对大量数据信息的高速处理及控制能力,通过虚拟手术环境为外科医生从技术上提供支援,是使手术更安全、更准确的一门新技术。近年来,随着CT、MRI等图像诊断仪的发展,使计算机虚拟现实技术在医学中的应用得到了飞速的发展。计算机利用这些图像信息进行三维图像重建,为外科医生进行手术模拟、手术导航(navigator)、手术定位、制订手术方案提供了客观、准确、直观、科学的手段。自1994年Vining等首次报道螺旋CT仿真内窥镜以来,在国内外已进入临床试用阶段。但在临床实际工作中,其价值如何,尚需评估。
结、直肠癌是胃肠道常见的恶性肿瘤,近年来发病率明显上升,发病率仅次于胃癌和食管癌,全球结直肠癌每年新发病例数达94万,每年近50万人死于结直肠癌。死亡数居癌症死因第三位。手术治疗一直是直肠癌的主要治疗手段,然而迄今为止没有建立起令人满意的精细三维解剖结构模型供临床参考。随着计算机技术向影像学领域渗透和应用,CT、MRI等断层影像技术实现了数字化图像处理,医学图像信息的采集、储存、传输、处理、显示等发生了前所未有的变革。由此形成全新医学图像处理和诊断模式,诞生了仿真影像学。仿真影像学以仿真内镜(VE)为核心,包括各种数字化图像三维重建技术和后处理技术。利用CT、MRI等二维断层图像为数据源,应用计算机软件重建直观地显示人体器官,尤其是管腔器官内表面解剖及病变三维仿真图像,能够准确地直观反映人体三维结构,进行疾病诊断、立体定位、模拟手术等。大肠及周围结构是普通外科重要的解剖结构,基于CT、MRI工作站对断层影像处理主要用于疾病诊断,目的比较单一,对CT、MRI等断层影像有效信息挖掘不够深入。结直肠癌是胃肠道常见的恶性肿瘤,常规钡剂灌肠和纤维结肠镜是主要的检查手段。三维成像及CT仿真内镜(CT virtual endoscopy,CTVE)技术对结肠病变的研究逐日增多特别是64层螺旋CT以其高速度、更加薄层、更大范围的容积扫描,为这一技术的发展起到了很大的促进作用。
为此,我们基于不同类型CTA断层数据集建立了(1)直肠及周围结构解剖结构的三维可视化模型;(2)直肠肿瘤及周围解剖结构的三维可视化模型;(3)在个人电脑重建大肠体绘制和面绘制模型,大肠和周围解剖结构空间毗邻关系模型及大肠虚拟内镜,为深入挖掘断层影像有效信息做了有益尝试;(4)在个人电脑对10例经术后病理证实的结肠癌和结肠息肉患者进行了大肠虚拟内镜检查及大肠和周围解剖结构模型进行重建,并和CT三维成像与CT仿真结肠镜进行形态学观察,探讨其临床应用价值。
研究目的和方法
目的
1、寻求基于CTA断层图像重建直肠及周围结构数字模型及三维可视化的方法,简便、高效地建立直肠及周围结构三维模型。
2、研究CTA断层图像重建直肠肿瘤数字模型及三维可视化方法建立的模型为解剖教学、生理模拟、手术培训提供医疗教学平台。
3、寻求基于64层螺旋CT断层图像在个人电脑重建大肠三维数字模型及虚拟内镜检查方法。
4、探讨基于64层螺旋CT断层图像在个人电脑重建大肠三维数字模型及虚拟内镜检查方法,根据患者的数字化仿真模型,可制订个体化的诊断、术前规划和手术模拟,提高手术的安全性,降低手术并发症。
方法
1.研究对象:
第一部分基于CTA断层图像直肠及周围结构数字模型的重建及三维可视化研究研究对象为健康男性志愿者一名,27岁;
第二部分直肠肿瘤数字解剖模型构建及三维可视化研究对象为直肠肿瘤病人,男性,68岁;
第三部分基于64层螺旋CT断层图像的影像后处理:大肠三维重建及虚拟内镜研究对象为健康志愿者一名,男性,31岁;
第四部分基于64层CTA断层图像虚拟技术和CT仿真结肠镜临床应用研究研究对象:搜集2007年7～9月结肠癌和结肠息肉患者10例。其中结肠癌8例,男5例,女3例,年龄37～65岁,平均50.4岁;结肠息肉2例,男1例,女1例,年龄40～57岁。
2.检查前肠道准备:
检查前3日进流质饮食,检查前晚服蓖麻油30ml,当日禁食,检查前2小时作清洁灌肠,直到排出液澄清为止,术前15分钟肌注阿托品0.5mg。
3.CT扫描前空气灌肠使肠管充气膨胀:
使用1S-818E电脑遥控空气灌肠机灌肠。肛门及肛管涂抹液体石蜡,经肛门插入Foley氏18F气囊管(代肛管),插入深度为10 cm,注人20-30ml气体将气囊充起,防止气囊管从肛门脱出。X线透视观察下开始灌肠。以稳定低压4-5kPa持续向直肠内注入空气使肠管充气膨胀,固定维持压力在4kPa。
4.数据采集设备
64排螺旋CT-PHILIPS Brilliance64(荷兰)。高压注射器采用MEDRAD双筒高压注射器(美国),图像后处理工作站为PHILIPS Brilliance64层螺旋CT自带的Mxview工作站。
5.CT扫描
空气灌肠成功后行CTA动脉造影,亚毫米状态下高分辨力容积扫描,常规平扫时患者取仰卧位,头足方向,扫描范围从胸12至股骨中上部分,扫描条件120KV、300mAs;采用0.625×64排探测器组合,以层厚0.5mm、间隔0.5mm,螺距(Pitch)0.809mm,球管旋转一周时间0.5s,选择软组织窗,准直器宽41.40mm×0.625 mm,开始常规平扫。动态CT增强扫描:平扫完成后行动脉期及静脉期连续跟踪扫描,经肘静脉注射对比剂(应用套管针),使用双筒CT高压注射器,注射速率5ml/s,所用对比剂为高浓度非离子型碘必乐370(370mgI/ml)或优维显370(370mgI/ml),剂量1.5ml/kg体重,对比剂注射完毕后以生理盐水50ml冲管,扫描技术条件同平扫。扫描完成后,应用增强原始数据进行1mm的薄层重建,并将图像数据传至Mxview工作站。
6.CT薄层扫描数据的采集
在Mxview诊断工作站上,利用光盘刻录全部的数据,其中包括平扫期、动脉期、静脉期数据。格式为DICOM(Digital Imaging and Communications inMedicine)3.0。
7.建立直肠及周围结构数字模型
Mimics软件分别读入Dicom格式的动脉期及静脉期CT断层图像,软件自动设定CT原始扫描参数,层间距均为0.5mm,动脉期螺距0.809mm,静脉期螺距0.824mm,以.MCS文件格式保存。在动脉期数据集分别重建骨骼、肠道、动脉、皮肤、膀胱及前列腺三维模型,根据不同研究的需要还分别建立动脉系统及肿瘤供血动脉、直肠肿瘤等。在静脉期数据集重建输尿管三维模型,以STL格式导出在动脉期数据集行三维模型立体配对。
8.肠管面绘制和外显镜
Mimics软件读入Dicom格式CT断层图像,软件自动设定CT原始扫描参数,每层原始图像灰度值以内插值处理,达到亚体素的精度。调整窗位窗宽,对比度达到最佳视度。空气灌肠使肠管充气膨胀,肠道面绘制模型按空气阈值建模,界定阈值在-1024至-1024亨氏单位,并膨胀1—2个像素达到选择像素充满全部肠腔,区域增长工具分割出直肠及乙状结肠,升结肠,横结肠及部分小肠,去除冗余数据,经3D计算建立肠道的三维数字模型。面绘制模型以三角面片显示。在动脉期数据集重建骨骼系统,动脉系统及皮肤,前列腺三维模型。在110层及82层二维图像分别手动分割出膀胱及前列腺,3D计算建立膀胱及前列腺三维数字模型。在静脉期数据集根据残余造影剂浓度重建输尿管三维模型,阈值在220-1641亨氏单位,经区域增长3D计算建立输尿管三维模型,以STL格式导出在动脉期数据集根据输尿管和双肾及膀胱解剖结构关系行三维模型立体配对。
9.虚拟内镜
对基于CT断层图像建立的肠管面绘制模型,用虚拟内窥镜技术可对肠腔解剖和病变进行观察,观看三角面片内部信息能够获得有效肠道内解剖图像和诊断信息。可在二维平面图像进行大体观察来指导内镜重点观察目标,给合多平面裁剪功能可以综合分析。在肠腔内“漫游”可观察全结肠。对观察到的病变对发现病变后,变换视角、视距,甚至反向观察。返回到二维平面图像测量其最大径作为直径。
10.CT三维成像和仿真内镜检查
使用Philips公司Brilliance 64 CT用5 mm层厚完成全腹部容积扫描,在后处理工作站(Extended Brilliance workspace)重建层厚0.5mm,间隔0.5 mm,使用容积再现(volume rendering,VR)和内镜软件,通过合理调整阈值获取三维图像及内镜图像。对表面遮盖显示(shaded surface display,SSD)和透明显示(RaySum)图像进行大体观察来指导内镜的重点观察目标,结合轴面图像及多平面重建(MPR)综合分析。
结果
1.基于CTA断层图像直肠及周围结构数字模型的重建及三维可视化研究成功建立了直肠及周围结构三维数字模型,对腰椎、骨盆及股骨等骨骼结构,腹主动脉及髂内动脉及髂外动脉及分支、皮肤、直肠及乙状结肠、升结肠、横结肠及部分小肠、膀胱、前列腺、输尿管分别建立数字模型。
建立的数字模型几何外型逼真,具有较佳的视觉效果,可以多彩色、透明或任意组合显示,整体清晰、实体感强。通过不同平面的切面可以观察各组件的内部结构关系,能够任意角度的旋转及缩放观察,动态显示可制作成AVI格式电影存储、输出,画面清晰流畅。
2.直肠肿瘤数字解剖模型构建及三维可视化
建立直肠及直肠肿瘤结构三维数字模型,真实再现了肿瘤在直肠内的确切位置,对腰椎,骨盆及股骨等骨骼结构,腹主动脉,髂内动脉,髂外动脉及分支,皮肤,直肠,肛管,直肠肿瘤及乙状结肠,升结肠,横结肠及部分小肠,膀胱,前列腺,输尿管分别建立数字模型。肿瘤与周围其他组织的位置关系,浸润情况一目了然。
3.基于64层螺旋CT断层图像的影像后处理:大肠三维重建及虚拟内镜
基于64层螺旋CT断层图像用虚拟内窥镜技术对肠腔的解剖和病变进行了观察。能成功内镜样展现胃肠道的任何部位与转角处,并能从狭窄、梗阻处两端观察官腔的解剖和病变,可靠地显示了病变部位和大小。
4.基于64层CTA断层图像虚拟技术和CT仿真结肠镜临床应用研究
成功建立了结肠面绘制三维数字模型,效果良好。用虚拟内镜对大肠解剖和病理结构进行了观察,其肿瘤形态与术中所见几乎完全一致。
结论
1.直肠及周围结构的三维重建方法应用可视化技术利用人类的视觉特性,运用计算机图像学技术,将二维断层图像序列进行处理,重构出具有立体效果的空间三维形态结构图像。基于空气灌肠造影及CTA血管造影技术,可更精确还原肠管及动脉,输尿管在活体内的原有真实结构形态,可简便、高效地建立直肠及周围结构三维模型。其创新性及意义有(1)、制定手术方案:能够利用图像数据,帮助医生合理、定量地制定手术方案,对于选择最佳手术路径、减小手术损伤、减少对临近组织损害、提高肿瘤定位精度、使手术安全性增加,而手术的风险性降低、并发症减少,对执行复杂外科手术和提高手术成功率等具有十分重要的意义。(2)、增进医患交流,减少医疗纠纷:医生可以用重建的三维模型对病人及家属讲解病情及手术情况,使患者及家属对病情及风险有直观的了解。(3)、降低手术费用:由于医生有较好的手术预案,能够缩短病人的恢复周期、降低病人和医院的开支。
2.基于CT断层图像的虚拟内镜技术源于螺旋CT连续扫描获得的容积数据,和传统的内窥镜技术相比,虚拟内窥镜和CT工作站仿真内窥镜有着共同独特的优点:(1)一种理想非侵入性的检查方法,病人无不适感,检查过程中不会产生穿孔、出血或感染等副作用;(2)常规的纤维镜存在视野的局限和无法评价腔外解剖与病变等缺点。虚拟内窥镜能够充分显示结肠的解剖形态以及病变部位,并能从狭窄、梗阻处两端观察肠腔的解剖和病变。除了解腔内情况外,结合三维图像还可了解肠壁以及腔外的情况,更有利于肿瘤的定性以及分期诊断,为临床制定手术方案提供依据;(3)可以针对不同要求,制定不同的漫游计划,任意地重复检查过程;(4)降低了医疗检查复杂性、危险性和医疗成本;(5)对绝大多数结肠肿瘤性病变可做出定性诊断。同时能完整地保存原始数据,具有很好的可重复性,更加灵活,不受入路限制,可任意到达所需观察的解剖部位。反复多次观察有利于小病灶以及多发性病灶的检查,可避免因人为因素导致的漏诊。
3.基于64层CTA断层图像虚拟技术比CT仿真结肠镜技术优势:能更精确还原骨骼,肠管及动脉,输尿管在活体内的原有真实结构形态,简便、高效地建立大肠及周围结构三维模型。以三维可视化的方式从不同角度进行展示,结合传统教材的图像与实体解剖,能更好地理解真实的三维人体结构,是对传统教材中的完善和补充。三维重建图像利于整体直观地显示病变,帮助明确诊断并指导手术。在可视化技术的基础上可以进一步实现计算机模拟及手术规划。肠管柔软易变形,离体后不易定形,其在人体的真实外形不易理解,本研究以三维可视化的形式,通过任意角度的旋转,全方位显示大肠及其周围结构,便于对周围结构的观察理解,对理解掌握直肠周围结构的解剖关系有极大帮助。因此,与结肠三维成像和CT仿真内镜相比,基于64层CTA断层图像虚拟技术能达到与CT仿真内镜结合三维成像同样的敏感性和特异性,加上各种组织三维重建技术可以提供比仿真内镜更丰富的信息,其应用前景十分广阔。
建立的模型为解剖教学、生理模拟、手术培训提供医疗教学平台。临床医师根据患者的数字化仿真模型,可制订个体化的诊断、术前规划和手术模拟,提高手术的安全性,降低手术并发症,推动普通外科的发展。
Background
The development of modern science and technology more and more reflects interdisciplinary intersection and penetration. The new interdisciplinary studies field-Medical Virtual Reality Technology (Medical Virtual Reality Technology, MVRT) which integrates medicine, biomechanics, mechanics, hylology, computer graphics, mathematical analysis, automatic control and the like is developing rapidly at present, bringing about rapid change on modern medical diagnosis and therapeutic methods and concepts. Computer Aided Surgery (Computer Aided Surgery, CAS) is a new technology which makes the surgery safer and more accurate through providing technical support for surgeons by virtual surgical environment. CAS is based on the high-speed processing and controlling capacity on a vast amount of data information of the computer. In recent years, with the development of image diagnostic apparatus, such as CT, MRI and the like, the application of computer virtual reality technology in medicine has developed rapidly. The computer provides objective, accurate, intuitive and scientific methods for surgery simulation, surgery navigator, surgery orientation and surgery planning through using the image information to reconstruct the three-dimensional images. Since Vining and so on have firstly reported the helical CT virtual endoscope in 1994, the helical CT virtual endoscope has entered in the stage of clinical trial. However, its value in clinical practical work needs to be estimated.
Colorectal cancer is the common malignant tumour in gastric bowel path. In recent years, the disease incidence has increased obviously. The disease incidence is only second to gastric cancer and esophageal cancer. The number of new cases of colorectal cancer every year around the world reaches 940,000. Nearly 500, 000 people die of colorectal cancer every year. Mortality is the third of cancer-related death. Surgical treatment is always the main treatment of rectum cancer. So far, the satisfactory fine three-dimensional anatomical structural model for clinical reference has not been established. The computer technology has been applied in imageology field, so slice image technology, such as CT, MRI and the like has realized digital image processing. Collection, storage, transmission, processing, display and so on of the medical image information have changed greatly. Thus, a new medical image processing and diagnostic mode is formed and the virtual imageology is born. The virtual imageology consists of various digital image three-dimensional reconstruction techniques and after-treatment techniques with virtual endoscopy (VE) as the core. The virtual imageology uses the two-dimensional slice images, such as CT, MRI and the like as the data source and utilizes the computer software reconstruction to directly display human organs, especially anatomical images of inner surfaces of lumen organs and three-dimensional virtual images of pathological changes. The virtual imageology can accurately and directly reflect the three-dimensional structures of human body and carry out disease diagnosis, stereoscopic localization, surgical simulation and so on. Large intestine and the surrounding structures are important anatomical structures of the general surgery. To slice image processing, CT and MRI workspace is mainly used for disease diagnosis, the purpose is single. Effective information mining of the slice image, such as CT, MRI and the like is not thorough. Colorectal cancer is the common malignant tumour in gastric bowel path, the regular barium enema and fibercoloscope are the main detection methods. Increasing researches on colonic pathological changes by three-dimensional imaging technology and CT virtual endoscopy (CT virtual endoscopy, CTVE) technology, especially 64-slice helical CT and the volume scanning with high-speed, thinner slice, wider range have largely promoted the development of this technology.
Thus, based on CTA slice data sets of different types, we (1) have established the three-dimensional visualization models of the anatomical structures of rectum and the surrounding structures; (2) have established the three-dimensional visualization models of rectal tumor and the surrounding anatomical structures; (3) have reconstructed the volume rendering model and the surface rendering model of large intestine, the spatial adjacent relation model of large intestine and the surrounding anatomical structures and large intestine virtual endoscope in personal computers and have tried to mine the effective information of the slice image and (4) have carried out the large intestine virtual endoscopy and reconstructed the structural models of large intestine and the surrounding structures on 10 cases of colon cancer patients and intestinal polyp patients in personal computers, viewed morphology through CT three-dimensional imaging and CT virtual colonoscope and discussed the clinical application value.
Study objectives and methods Purpose
1. Seek for the method for reconstructing the digital models of rectum and the surrounding structures and for three-dimensional visualizing based on CTA slice images.
2. Research the method for reconstructing the digital model of rectum cancer tumour and for three-dimensional visualizing based on CTA slice images.
3. Seek for the method for reconstructing the three-dimensional digital models and for virtual endoscope detecting in personal computers based on 64-slice helical CT slice images.
4. Discuss the method for reconstructing the three-dimensional digital models and for virtual endoscope detecting in personal computers based on 64-slice helical CT slice images.
Method
1. Study object:
The first part The study object of reconstruction of the digital models of rectumand the surrounding structures and three-dimensional visualization based on CTAslice images is a healthy male volunteer who is 27-year-old.
The second part The study object of the construction of the digital model ofrectum cancer tumour and three-dimensional visualization is a male rectum tumourpatient who is 68-year-old.
The third part Image after-treatment based on 64-slice helical CT slice images: thestudy object of the three-dimensional reconstruction of large intestine and virtualendoscopy is a healthy male volunteer who is 31-year-old.
The fourth part The study object of the virtual technology based on 64-slice helicalCT slice images and the CT virtual colonoscope clinical application research: collect10 cases of colon cancer patients and intestinal polyp patients during July toSeptember in 2007. The colon cancer patients are 8 cases, wherein, 5 cases are maleand 3 cases are female, who are 36 to 65 years old. The intestinal polyp patients are 2cases, wherein, 1 case is male and 1 case is female, who are 40 to 57 years old.
2. Bowel preparation before testing
Eat liquid food 3 days before testing. Eat 30ml caster oil the night before testing. Fast on the day of testing. Carry out cleansing enema 2 hours before testing till the discharge liquor is clear. Intramuscularly inject 0.5mg atropine 15 minutes before operation.
3. Carrying out air enema to inflate intestinal canal before CT scanning
Use 1S-818E computer to remote-control an air enemator to carry out enema. Smear liquid paraffin on anus and anal canal. Insert Foley 18F sac duct (anal tube) per anus. The depth of insertion is 10cm. Inject 20-30ml gas to inflate the sac in order to prevent the sac duct from falling out. Begin to carry out enema under X-ray perspective observation. Continuously inject air into rectum with the stable 4-5kPa low pressure to inflate intestinal canal. Keep the pressure at 4kPa.
4. Data acquisition equipment:
64-slice helical CT-PHILIPS Brilliance64 (Holland) is used. The high-pressure injector uses MEDRAD dual-tube high-pressure injector (America). The image after-treatment processing workspace is attached Mxview workspace of PHILIPS Brilliance64-slice helical CT.
5. CT scanning:
Carry out CTA arteriography and high-resolution volume scanning under submillimeter state after the air enema is successful. When carry out routine scanning, the patient is in supine position. The scanning area is from twelfth thoracic vertebra to the middle upper part of thighbone along head-foot direction. The scanning condition is 120KV, 300mAs. 0.625×64-slice detector combination is used. Slice thickness is 0.5mm, interval is 0.5mm and pitch is 0.809mm. Time of a pipet rotating one circle is 0.5s. Soft tissue window is selected. Width of the collimator is 41.40mm×0.625 mm. Begin to routine scan. Dynamic CT enhanced scan: carry out arterial phase and venous phase continuous tracking scanning after the routine scanning is finished. Inject a contrast agent (using trocar) per cubital vein. Use a dual-tube CT high-pressure injector. The injection rate is 5ml/s. The used contrast agent is high-concentration non-ionic Iopamiro 370 (370mgI/ml) or Ultravist 370 (370mgI/ml). The dosage is 1.5 ml/kg weight. Use 50ml normal saline solution to rinse the tube after injecting the contrast agent. The scanning condition is the same with the routine scanning. Use the enhanced original data to reconstruct 1mm lamel after finishing scanning and transmit the image data to Mxview workspace.
6. Collection of CT lamel scanning data: at Mxview diagnosis workspace, use a compact disk to burn all data, comprising routine scanning phase, arterial phase and venous phase data. The format is DICOM (Digital Imaging and Communications in Medicine)3.0.
7. Construct digital models of rectum and the surrounding structures
Mimics software respectively reads in arterial phase and venous phase CT slice images of Dicom format. The software automatically sets CT original scanning parameters. The interlamellar spacing is 0.5 mm, the arterial phase pitch is 0.809mm and the venous phase is 0.824mm. The parameters are stored by .MCS file format. Respectively reconstruct the three-dimensional models of bone, intestinal canal, artery, skin, bladder and prostate at the arterial phase data set. Respectively reconstruct arterial system, tumour artery, rectum tumour and so on according to different research needs. Reconstruct the three-dimensional model of ureter at the venous phase data set. Lead out the three-dimensional models in STL format and carry out the three-dimensional model stereoscopic pairing at the arterial phase data set.
8. Intestinal canal surface rendering and external microscope Mimics software reads in CT slice images of Dicom format. The software automatically sets CT original scanning parameters. Gray value of each original image is processed by interpolated value in order to reach the precision of subvoxel. Adjust the window position and the window width to make the contrast reach the optimum visibility. Carry out air enema to inflate the intestinal canal. Construct the intestinal canal surface rendering model according to the air threshold value. The threshold value is defined between -1024 and -1024 Hounsfield unit. Inflate 1-2 pixels to make intestinal cavity be full with the selected pixels. Regional growth tools divide rectum and sigmoid colon, ascending colon, transverse colon and part of small intestine. Remove the redundant data. Construct the three-dimensional digital model of rectum through 3D computation. Surface rendering model is displayed by tri-patch. Reconstruct the three-dimensional models of bone system, arterial system, skin and prostate at the arterial phase data set. Manually divide bladder and prostate at 110-layer and 82-layer two-dimensional images. Construct the three-dimensional digital models of bladder and prostate through 3D computation. Reconstruct the three-dimensional model of ureter at the venous phase data set according to the concentration of the residual contrast agent. The threshold value is between 220 and 1641 Hounsfield unit. Reconstruct the three-dimensional model of ureter through regional growth 3D computation. Lead out the three-dimensional models in STL format and carry out the three-dimensional model stereoscopic pairing according to the anatomical structural relation of ureter, double kidneys and bladder at the arterial phase data set.
9. Virtual endoscope
To the intestinal canal surface rendering model constructed based on CT slice image, virtual endoscopy can be used to anatomize intestinal cavity and observe pathological changes. Effective intestinal canal anatomical image and diagnosis information can be obtained through observing the internal information of the tri-patch. Generally observe the two-dimensional images to guide the important observing target of the endoscope. Comprehensively analyze the target through multi-plane cutting function. Roam in intestinal cavity to observe the whole colon. Change the visual angle and the viewing distance, even reversely observe when the observed pathological changes have occurred. Return to the two-dimensional plane image to measure the largest diameter as the diameter.
10. CT three-dimensional imaging and virtual endoscopy checking
Use Philips Brilliance 64 CT and 5mm thickness to finish the whole abdomen volume scanning. Reconstruct 0.5mm thickness and 0.5mm interval at the after-treatment workspace (Extended Brilliance workspace) to make volume rendering (volume rendering, VR) and endoscope software. Obtain the three-dimensional image and the endoscope image through rationally adjusting the threshold value. Generally observe shaded surface display (shaded surface display, SSD) and Raysum (RaySum) images to guide the important observing target of the endoscope. Comprehensively analyze the target through combining the axial plane image and multi-plane reconstruction (MRP).
Result
1. Reconstruction of the digital models of rectum and the surrounding structures and three-dimensional visualization research based on CTA slice images
Successfully construct the digital models of rectum and the surrounding structures. Respectively construct the digital models of bone structures, such as lumber, pelvis, thighbone and so on, abdominal aorta, internal iliac artery, external iliac artery, branch, skin, rectum, sigmoid colon, ascending colon, transverse colon, part of small intestine, bladder, prostate and ureter.
The constructed models have vivid geometrical appearance, better visualization effect, clear view and strong reality. The models can be displayed through several colours, tranparancy or arbitrary combinations. The internal structural relation between each component can be observed through the sections of different planes. The models can be rotated or observed at any angles. Dynamic display can be made into AVI format film to store and output. The pictures are clear and smooth.
2. Construction of the digital anatomical model of rectum tumour and three-dimensional visualization
Construct the three-dimensional digital models of rectum and rectum tumour, truly rendering the accurate position of the tumour in rectum. Respectively construct the digital models of bone structures, such as lumber, pelvis, thighbone and so on, abdominal aorta, internal iliac artery, external iliac artery, branch, skin, rectum, anal canal, rectum tumour, sigmoid colon, ascending colon, transverse colon, part of small intestine, bladder, prostate and ureter. The position relation between tumour and the surrounding tissues and the infiltration situation are clear.
3. Image after-treatment based on 64-slice helical CT slice image: large intestine three-dimensional reconstruction and virtual endoscopy
Observe anatomization and pathological changes of intestinal cavity through using the virtual endoscopy based on 64-slice helical CT slice image. Any part and corner in gastrointestinal tract can be successfully rendered. Anatomization and pathological changes of intestinal cavity can be observed from narrow and block part. Position and size of pathological change can be reliably displayed.
4. Virtual technology and CT virtual colonoscope clinical application research based on 64-slice helical CT slice image.
Successfully construct the colon surface rendering three-dimensional digital model.
Observe anatomization and pathological changes of large intestine through using thevirtual endoscopy. The tumour form is exactly the same with the form seen in theoperation.
Conclusion
1. The three-dimensional reconstruction method of rectum and the surrounding structures reconstructs the spatial three-dimensional images with stereoscopic effect through using visualization technology, human visual characteristic and computer graphics to process the two-dimensional slice image series. The original and actual structures of intestinal canal, artery and ureter in vivo can be more accurately reconstructed and the three-dimensional models of rectum and the surrounding structures can be simply and effectively constructed based on air enema and CTA angiography.
2. The virtual endoscopy based on CT slice image derives from the volume data obtained through helical CT continuous scanning. Compared with the traditional endoscopy, the virtual endoscope and CT workspace virtual endoscope have the same special advantages. (1) It is a perfect non-invasively check method. Patients are free of discomfort. There are no side effects of perforating, bleeding or infecting during the checking process. (2) The regular fiber scope has limited view and can not evaluate outer cavity anatomization and pathological changes. The virtual endoscope can fully display anatomization and pathological change position of colon and can observe anatomization and pathological changes of intestinal cavity from narrow and block part. Besides conditions in the cavity, the conditions of intestinal wall and out of the cavity can be known through combining with the three-dimensional images. It is better for determining tumour and stage diagnosis. The virtual endoscope provides evidence for clinic operation plan. (3) Different roaming plans can be planned out according to different needs and checking process can be arbitrarily repeated. (4) Medical check difficulty, danger and medical cost are reduced. (5) Most colon tumour pathological changes can be determined. At the same time, original data can be completely stored. The virtual endoscopy has better repeatability and flexibility and it is not limited by approaches and can reach the part which is needed to be observed arbitrarily. Repeated observation is good for checking small lesions and multiple lesions, avoiding missed diagnosis caused by human factors.
3. Compared with CT virtual colonoscope technology, the advantages of the virtual technology based on 64-slice CTA slice image are: it can more accurately reconstruct bone, intestinal canal, artery and ureter in vivo and simply and effectively construct the three-dimensional models of large intestine and the surrounding structures. The virtual technology based on 64-slice CTA slice image can display from different angles through the three-dimensional visualization method. It is good for learning the actual three-dimensional human body through combining with the images in the traditional teaching material and substantial anatomization. The virtual technology based on 64-slice CTA slice image is supplement to the traditional teaching material. The three-dimensional reconstructed images are good for directly displaying the pathological changes, determining diagnosis and guiding operation. Computer simulation and operation planning can be further realized on the basis of visualization technology. The intestinal canal is soft and easy to deform, and is hard to shape in vitro. The actual shape of the intestinal canal in the human body is hard to understand. The study displays large intestine and the surrounding structure in an all-round manner through the three-dimensional visualization and rotation at any angle, which is good for observing and understanding the surrounding structures and helpful for understanding the anatomic relation of the surrounding structures of rectum. Compared with colon three-dimensional imaging and CT virtual endoscopy, the virtual technology based on 64-slice CTA slice image can realize the same sensitivity and speciality of the CT virtual endoscopy and three-dimensional imaging. Additionally, various tissue three-dimensional reconstruction technologies can provide richer information than the virtual endoscopy, its application future is wide.
The constructed models provide a medical teaching platform for anatomic teaching, physical simulation and operation training. Clinic doctors can work out individual diagnosis, plan before operation and carry out operation simulation according to the digital simulated models of patients. Thus, operation safety is enhanced and complications of surgery are reduced. Development of general surgery is promoted.
引文
[1]李万冬,徐惠绵,凌光烈,等.中国数字化可视人体腹腔内脏器官三维重建[J].中国医科大学学报,2006.35(5):484-486
[2]贾克斌.数字医学图像处理、存档及传输技术.[M].北京:科学出版社,2006.14-16.
[3]傅栋,靳安民.CT断层图像快速构建人体骨骼有限元几何模型的方法[J].中国组织工程研究与临床康复,2007.11(9):1620-1623
[4]http://www.irfanview.com[Z]
[5]原林,唐雷,黄文华,等.虚拟中国人男性一号(VCH—M1)数据集研究(英文)[J].第一军医大学学报,2003.23(6):520-523
[6]胡小春,郭松青,等.基于CT图像建立人体足部骨骼三维有限元模型的研究[J].合肥工业大学学报(自然科学版),2005,28(9):1188-1191
[7]Phillips CL,Arend LJ,Filson AJ,et al.Three-dimensional imaging of Embryonic mouse kidney by two-photon microscopy[J].Am J Pathol,2001,158(1):49-55.
[8]Vlaik T,Yazdi HM,Burns BE et at.Pattem of distribution of inlraduclal and infiltrating duclal cancinoma:A three dimensional stundy using serialcoronal giant sections of the breast[J].Hum Patrol,2000,31(4):464-474.
[9]李文刚,黄志强,李滨等.大鼠门静脉动脉化肝门部胆管微血管的三维重建观察[J].消化外科,2004,3(6):411-414.
[10]Fenster A,DowneyDB,CardinalNH.Three dimension ultrasound imaging[J]. Physics in Medicine and Biology,2001,46(1):67-99.
[11]Devore GR.Three-dimensional and four-dimensional fetal echocardiography:a new frontier[J].Curr Opin Pediatr.2005 Oct;17(5):592-604.
[12]何爱军,汪天富,尹立雪等.多普勒血流超声医学图像的动态三维重建[J].航天医学与医学工程.2002,15(2):112-116.
[13]Hsu CY,Chiba Y,Fukui O.et al.Counterclockwise barber-pole sign on prenatal three-dimensional power Doppler sonography in a case of duodenal obstruction without intestinal malrotation[J].J Clin Ultrasound.2004,32(2):86-90.
[1]孟庆玲.结直肠癌发病率明显上升[N].中国医药报,2007-06-25,B版,081.
[2]顾晋.直肠癌手术治疗的热点问题[J].中国实用外科杂志,2007,28(6):
[3]Fasel JH,Gingins P,Kalra P,et al.liver of the“visibleman”[J].Clin Anat, 1997,10(6):389.
[4]贾克斌.数字医学图像处理、存档及传输技术.[M].北京:科学出版社,2006.14-16.
[5]傅栋,靳安民.CT断层图像快速构建人体骨骼有限元几何模型的方法[J].中国组织工程研究与临床康复,2007.11(9):1620-1623
[6]原林,唐雷,黄文华,等.虚拟中国人男性一号(VCH—M1)数据集研究(英文)[J].第一军医大学学报,2003.23(6):520-523
[7]胡小春,郭松青,等.基于CT图像建立人体足部骨骼三维有限元模型的研究[J].合肥工业大学学报(自然科学版),2005,28(9):1188-1191
[1]罗立民.仿真医学影像学[J].中国医疗器械杂志,2002.26(2):79-81
[2]闵翔,谢德明.CT仿真内镜的应用[J].现代医学仪器与应用,2007.11(3):28-31
[3]Dusan SY,Zhang DT,Lin QC,Wu YH.clinical value of CT three-dimensional imaging in diagnosing gastrointestinal tract diseases.World Gastroenterol 2006;12:2945-2948
[4]Horton KM,FishmanEK.Current role of CT in imaging of the stomach.Radiographics 2003;23:75-87
[5]Ba-Ssalamah A,ProkopM,Uffmann M,Pokieser P,Teleky B,Lechner G.Dedicated multidetector CT of the stomach:spectyum of diseases.Rediographics 2003;23:625-644
[6]Patak MA,Mortele KJ,Ros PR.Multidetector row CT of the small bowel.Rodial Clin North Am 2005;43:1063-1077,viii
[7]Qalbani A,Paushter D,Dachman AH.Multidetector row CT of the small bowel obstruction.Rodial Clin North Am 2007;45:499-512,viii
[8]Ramachandran I,Sinha R,Rajesh A,Verma R,Maglinte DD.Multidetector row CT of the small bowel tumours.Clin Rodial 2007;62:607-614
[9]张崴,牛玮.医学数字图像和通讯标准——DICOM[J].现代医药卫生,2005.21(1):111-112
[10]骆建珍,林财兴,孟令旗,杨安荣.DICOM医学图像三维可视化系统的研究.机电一体化2007;13:45-48
[11]Liu S,Liao W,Yu Q,Cheng X,Dai N,Zhang X.The development of a system for 3D reconstruction from DICOM data and collaborative visualization.Sheng wu yi xue Gong cheng xue Za zhi 2007;24:1152-1156
[12]Kling-petersen T,Rydmark M.Modeling and modification of medical 3D objects.The benefit of using haptic modeling tool,Stud health technol inform2000;70:162-167
[13]Durand lg,Garcia D,Sakr F,Sava H,Cimon R,Pibarot P,Fenster A,Dumesnil JG.A new foiw model for Doppler ultrasound study of prosthetic heart valves.J Heart Valve Dis 1999;8:85-95
[14]张娟,王昌元,鲁雯,侯庆峰,张会如.医学图像三维重建的算法研究.中国医学装备2008;5:19-21
[15]王晓敏,张艳,王鹏程,谢晋东,韩彬,唐峰.医学三维可视化方法及算法平台的研究进展.中国医学装备2007;4:5-9
[16]王福昌,胡顺田,张宝雷.基于体绘制法的头部CT图像的三维重建.计算机与现代化2007;23:86-88
[17]宋卫东,雷万军.基于一种改进的光线投影体绘制算法的三维重建.医疗设备信息2007;22:1-3
[18]Escott EJ,Rubinstein D.Free DICOM image viewing and processing software for your desktop computer:what's available and what it can do for you.Radiogiaphics 2003;23:1341-1357
[19]Zhu XY,Fang CH,Bao SS,Quan XY,Zhong SZ.Image segmentation and three-dimensional reconstruction of the liver based on 64-slice spiral CT scanning datal.Nanfang Yike Daxue Xuebao 2008;28:345-347
[20]Cline HE,Dumomlin CL,Hart HR Jr,Lorensen WE,Ludke S.3D reconstruction of the brain from magnetic resonance images using a connectivity algorithm.Magn Reson Imanging 1987;5:345-352
[21]Lee TY,Lin CH,Growing-club isosurface extraction algorithm for medical volume data.Computer Medimaging Graph 2001;25:405-415
[22]Wang k,denney t,Morrison E,Vodyanoy V.Construction of volume meshes from computed tomography data.Conf ProcIEEE Eng Med Boil Soc2005;5:5168-5171
[23]Delibasis KS,Matsopoulos GK,Mouravliansky NA,Nikita KS,A novel and efficient implementation of the marching cubes algorithm.Comput Med Imaging Graph 2001;25:343-352
[24]Lesk VI,Sternberg MJ,3D-garden:a system for modeling protein-protein complexes based on conformational refinement of ensembles generated with the marching cubes algorithm.Bioinformatics 2008;24:1137-1144
[25]Teferi S,Visualization,volume and surface area calculation of there dimensional(3D)ultrasound images.Ethiop Med J 2007;45:383-390
[26]Niu Q,Chi X,Leu MC,Ochoa J.Image processing geometric modeling and data management for development of a virtual bone surgery system.Computer Aided Surg 2008;13:30-40
[27]Schreiner J,Scheidegger CE,Sivla CT.High-quality extraction of isosurface from regular and irregular grids.IEEE Trans Vis Comput Graph 2006;12:1205-1212
[28]傅栋,靳安民.CT断层图像快速构建人体骨骼有限元几何模型的方法[J].中国组织工程研究与临床康复,2007.11(9):1620-1623
[29]陈新,陈良琴,徐伟.三维可视化技术及其在现代医学诊断中的应用[J].中国高校科技与产业化,2006.10(S1):306-309
[30]姚德民,宋志坚.虚拟内窥镜关键技术的研究与临床应用.生物医学工程学杂志,2008;25:18-22.
[31]高向军,田联房,王立非,毛宗源.虚拟内镜系统的关键技术分析.计算机应用研 究2008;25:1956-1959
[32]彭延军,王元红,郑永果.虚拟内镜系统研究进展.计算机工程与应用2005;41:203-205
[33]姚德民,宋志坚.交互式虚拟内镜成像及应用.解剖学杂志2005;28:192-194
[1]刘秋红.纤维结肠镜诊断结肠息肉及结肠癌231例[J].大理学院学报,2002.1(3):49-50
[2]罗立民.仿真医学影像学[J].中国医疗器械杂志,2002.26(2):79-81
[3]闵翔,谢德明.CT仿真内镜的应用[J].现代医学仪器与应用,2007.11(3):28-31
[4]王冠民,郭存英,张仰奇,等.CT仿真结肠镜与三维重建相结合对结肠病变的诊断价值[J].实用医学影像杂志,2004.5(3):151-153
[5]王岚.CT仿真内镜与纤维内镜及钡灌肠造影对结肠疾病的比较研究[J].黑龙江医学,2005.29(9):659-660
[6]张太发.基于CT医学图像的三维表面重建算法的研究[D].西安理工大学,2007.
[7]陈新,陈良琴,徐伟.三维可视化技术及其在现代医学诊断中的应用[J].中国高校科技与产业化,2006.10(S1):306-309
[8]傅栋,靳安民.CT断层图像快速构建人体骨骼有限元几何模型的方法[J].中国组织工程研究与临床康复,2007.11(9):1620-1623
[1]Phillips CL,Arend LJ,Filson AJ,et al.Three-dimensional imaging of Embryonic mouse kidney by two-photon microscopy[J].Am J Pathol,2001,158(1):49-55.
[2]Vlaik T,Yazdi HM,Burns BF,et at.Pattern of distribution of inlraduclal and infiltrating duclal cancinoma:A three dimensional stundy using serialcoronal giant sections of the breast[J].Hum Patrol,2000,31(4):464-474.
[3]李文刚,黄志强,李滨等.大鼠肝门部胆管周围血管丛的研究[J].中国普外基础与临床杂志,2005;12,(4)335-337.
[4]李文刚,黄志强,李滨等.大鼠门静脉动脉化肝门部胆管微血管的三维重建观察[J].消化外科,2004,3(6):411-414.
[5]Fenster A,Downey DB,Cardinal NH.Three dimension ultrasound imaging[J].Physics in Medicine and Biology,2001,46(1):67-99.
[6]Devore GR.Three-dimensional and four-dimensional fetal echocardiography:a new frontier[J].Curr Opin Pediatr.2005 Oct;17(5):592-604.
[7]沈学东,洪涛,周京敏等.体元模型动态三维重建超声显像的定量诊断价值[J].中国临床医学,1999,6(3)197--199.
[8]何爱军,汪天富,尹立雪等.多普勒血流超声医学图像的动态三维重建[J].航天医学与医学工程.2002,15(2):112-116.
[9]Kimura B J,Bhargara V,DeMarin AN.Value and limitations of intravascular ultrasound imaging in characterizing coronary atherosclerotic plaque[J].Am Heart J,1995,130:386-396.
[10]王新房,黄润青.实时三维超声心动图的应用现状及前景展望[J],中国医疗器械信息.2004,11(1):12-14.
[11]Hsu CY,Chiba Y,Fukui O.et al.Counterclockwise barber-pole sign on prenatal three-dimensional power Doppler sonography in a case of duodenal obstruction without intestinal malrotation[J].J Clin Ultrasound.2004,32(2):86-90.
[12]Bonsanto MM,Metzner R,Aschoff A.et al.3D ultrasound navigation in syrinx surgery--a feasibility study[J].Acta Neurochir(Wien).2005,147(5):533-540.
[13]Uchida T,Takada T,Ammori BJ,et al.Three-dimensional reconstruction of the ventral and dorsal pancreas:a new insight into anatomy and embryonic development[J].J Hepatobiliary Pancreat Surg.1999;6(2):176-180.
[14]苏焕群,林万莲,唐丹,等.概述数字化虚拟人体的研究[J].医学信息,2003,16(7):358-359
[15]Spitzer VM,Whitlock DG..The visible human dataset:the anatomical platform of human simulation[J].Anat Rec,1998,253:49-57.
[16]Spitzer VM,Ackerman MJ,Scherringer AL,et al.The visible human male:a technical report[J].J An Med Inform Assoc,1996,3:118-130.
[17]Ackeman NJ.The visible human project:a resource for education[J].Acad Med,1999,74:667-670.
[18]王旭,桂业英.可视人体(VH)数据集及使用[J].生物医学工程学杂志,1999,16(1:109-111.
[19]罗述谦.中国虚拟人医学图像方法学考虑[J].中国基础科学,2002,6:20-24.
[20]张绍祥,王平安,刘正津,等.首套中国男、女数字化可视人体结构数据的可视化研究[J].第三军医大学学报,2003,25(5):394-396.
[21]奇云.数字化虚拟中国人的研究及战略意义[J].物理通报,2003,4:1-3.
[22]罗述谦.虚拟人体研究与医学图像处理方法学[J].医疗设备信息,2002,17(5):1-3.
[23]方驰华,周五一,黄立伟等.虚拟中国人女性一号肝脏图像三维重建和虚拟手术的切割[J].中华外科杂志,2005,43(11):682-686.
[24]单锦露,张绍祥,刘正津等.女性盆腔的可视化研究及其意义[J].中国医学影像技术,2005,21(3):478-480.
[25]钟世镇.数字化虚拟人体的科学意义及应用前景[J].第一军医大学学 报,2003,23(3):193-195.
[26]李七渝,张绍祥,王平安等.人体大脑数字化解剖模型的构建及可视化[J].解剖学报,2005,36(6):638-641.
[27]Lutz C,Takagi A,Janecka IP,et al.Three dimensional computer reconstruction of a temporal bone[J].Otol aryngol Head Neck Surg,1989,101(5):522-256.
[28]邱明国,张绍祥,刘正李等.中国人侧颅底区可视化研究[J].解剖学报,2004,35(6):598-601.
[29]张伟国,陈现红,张绍祥等.枕下海绵窦可视化成像与MRI对照研究[J].中国医学影像技术,2005,21(5):809-811.
[30]张元智,顾立强,尹博等.虚拟中国人女号(VCH-F1)臂丛神经可视化初步研究[J].中华手外科杂志,2005,21(5):277-279.
[31]Guo YL,Heng PA,Zhang SX.et al.Thin sectional anatomy,three-dimensional reconstruction and visualization of the heart from the Chinese Visible Human[J].Surg Radiol Anat.2005,27(2):113-118.
[32]周泽民,方驰华,钟世镇等.基于数字化虚拟中国人女性一号的胰腺图像三维重建及可视化研究[J].中华外科杂志,2005,43(21):1401-1404.
[33]周泽民,方驰华,钟世镇等.数字化虚拟胰腺三维重建图像与传统解剖学图像的比较[J].第四军医大学学报,2005,26(18):1653-1656.
[34]方驰华,唐云强,鲁朝敏,等.胆囊切开取石和胆囊切除术的可视化研究[J].南方医科大学学报,2008,28(03):356
[35]方驰华,鲁朝敏,黄燕鹏,等.数字医学技术在肝癌外科诊治中的应用价值研究.中华外科杂志,2009,47(7):1-4.
[36]方驰华,鲁朝敏,黄燕鹏,等.仿真手术在肝移植2型肝动脉变异动脉重建中应用价值的研究.南方医科大学学报,2009,29(1):16-19.
[37]周泽民,方驰华,钟世镇等.数字化虚拟胰腺三维重建图像与传统解剖学图像的比较[J].第四军医大学学报,2005,26(18):1653-1656.
[38]邱明国,张绍祥,刘正津,等.耳三维重建及虚拟内窥镜[J].第三军医大学学报,2003,25(7):572-574.
[39]林鸿国,杨时鸿.虚拟人对中医药学发展的影响[].中医药学刊,2003,21(6):953-954.
[40]余安胜,张海东,李凤梅等.穴位标本断面图相配准方法研究[J].中国针灸,2003,23(2):89-91.
[41]钟沁,罗述谦.基于首例中国女性数字人的经穴三维重建[J].首都医科大学学报,Journal of Capital University of Medical Sciences,2004,25(3):281-283.
[42]刘红菊,赵静,严振国等.基于虚拟人技术的风池穴可视化研究[J].针刺研究,2004,29(4):282-285.
[2]贾克斌.数字医学图像处理、存档及传输技术.[M].北京:科学出版社,2006.14-16.
[3]傅栋,靳安民.CT断层图像快速构建人体骨骼有限元几何模型的方法[J].中国组织工程研究与临床康复,2007.11(9):1620-1623
[4]http://www.irfanview.com[Z]
[5]原林,唐雷,黄文华,等.虚拟中国人男性一号(VCH—M1)数据集研究(英文)[J].第一军医大学学报,2003.23(6):520-523
[6]胡小春,郭松青,等.基于CT图像建立人体足部骨骼三维有限元模型的研究[J].合肥工业大学学报(自然科学版),2005,28(9):1188-1191
[7]Phillips CL,Arend LJ,Filson AJ,et al.Three-dimensional imaging of Embryonic mouse kidney by two-photon microscopy[J].Am J Pathol,2001,158(1):49-55.
[8]Vlaik T,Yazdi HM,Burns BE et at.Pattem of distribution of inlraduclal and infiltrating duclal cancinoma:A three dimensional stundy using serialcoronal giant sections of the breast[J].Hum Patrol,2000,31(4):464-474.
[9]李文刚,黄志强,李滨等.大鼠门静脉动脉化肝门部胆管微血管的三维重建观察[J].消化外科,2004,3(6):411-414.
[10]Fenster A,DowneyDB,CardinalNH.Three dimension ultrasound imaging[J]. Physics in Medicine and Biology,2001,46(1):67-99.
[11]Devore GR.Three-dimensional and four-dimensional fetal echocardiography:a new frontier[J].Curr Opin Pediatr.2005 Oct;17(5):592-604.
[12]何爱军,汪天富,尹立雪等.多普勒血流超声医学图像的动态三维重建[J].航天医学与医学工程.2002,15(2):112-116.
[13]Hsu CY,Chiba Y,Fukui O.et al.Counterclockwise barber-pole sign on prenatal three-dimensional power Doppler sonography in a case of duodenal obstruction without intestinal malrotation[J].J Clin Ultrasound.2004,32(2):86-90.
[1]孟庆玲.结直肠癌发病率明显上升[N].中国医药报,2007-06-25,B版,081.
[2]顾晋.直肠癌手术治疗的热点问题[J].中国实用外科杂志,2007,28(6):
[3]Fasel JH,Gingins P,Kalra P,et al.liver of the“visibleman”[J].Clin Anat, 1997,10(6):389.
[4]贾克斌.数字医学图像处理、存档及传输技术.[M].北京:科学出版社,2006.14-16.
[5]傅栋,靳安民.CT断层图像快速构建人体骨骼有限元几何模型的方法[J].中国组织工程研究与临床康复,2007.11(9):1620-1623
[6]原林,唐雷,黄文华,等.虚拟中国人男性一号(VCH—M1)数据集研究(英文)[J].第一军医大学学报,2003.23(6):520-523
[7]胡小春,郭松青,等.基于CT图像建立人体足部骨骼三维有限元模型的研究[J].合肥工业大学学报(自然科学版),2005,28(9):1188-1191
[1]罗立民.仿真医学影像学[J].中国医疗器械杂志,2002.26(2):79-81
[2]闵翔,谢德明.CT仿真内镜的应用[J].现代医学仪器与应用,2007.11(3):28-31
[3]Dusan SY,Zhang DT,Lin QC,Wu YH.clinical value of CT three-dimensional imaging in diagnosing gastrointestinal tract diseases.World Gastroenterol 2006;12:2945-2948
[4]Horton KM,FishmanEK.Current role of CT in imaging of the stomach.Radiographics 2003;23:75-87
[5]Ba-Ssalamah A,ProkopM,Uffmann M,Pokieser P,Teleky B,Lechner G.Dedicated multidetector CT of the stomach:spectyum of diseases.Rediographics 2003;23:625-644
[6]Patak MA,Mortele KJ,Ros PR.Multidetector row CT of the small bowel.Rodial Clin North Am 2005;43:1063-1077,viii
[7]Qalbani A,Paushter D,Dachman AH.Multidetector row CT of the small bowel obstruction.Rodial Clin North Am 2007;45:499-512,viii
[8]Ramachandran I,Sinha R,Rajesh A,Verma R,Maglinte DD.Multidetector row CT of the small bowel tumours.Clin Rodial 2007;62:607-614
[9]张崴,牛玮.医学数字图像和通讯标准——DICOM[J].现代医药卫生,2005.21(1):111-112
[10]骆建珍,林财兴,孟令旗,杨安荣.DICOM医学图像三维可视化系统的研究.机电一体化2007;13:45-48
[11]Liu S,Liao W,Yu Q,Cheng X,Dai N,Zhang X.The development of a system for 3D reconstruction from DICOM data and collaborative visualization.Sheng wu yi xue Gong cheng xue Za zhi 2007;24:1152-1156
[12]Kling-petersen T,Rydmark M.Modeling and modification of medical 3D objects.The benefit of using haptic modeling tool,Stud health technol inform2000;70:162-167
[13]Durand lg,Garcia D,Sakr F,Sava H,Cimon R,Pibarot P,Fenster A,Dumesnil JG.A new foiw model for Doppler ultrasound study of prosthetic heart valves.J Heart Valve Dis 1999;8:85-95
[14]张娟,王昌元,鲁雯,侯庆峰,张会如.医学图像三维重建的算法研究.中国医学装备2008;5:19-21
[15]王晓敏,张艳,王鹏程,谢晋东,韩彬,唐峰.医学三维可视化方法及算法平台的研究进展.中国医学装备2007;4:5-9
[16]王福昌,胡顺田,张宝雷.基于体绘制法的头部CT图像的三维重建.计算机与现代化2007;23:86-88
[17]宋卫东,雷万军.基于一种改进的光线投影体绘制算法的三维重建.医疗设备信息2007;22:1-3
[18]Escott EJ,Rubinstein D.Free DICOM image viewing and processing software for your desktop computer:what's available and what it can do for you.Radiogiaphics 2003;23:1341-1357
[19]Zhu XY,Fang CH,Bao SS,Quan XY,Zhong SZ.Image segmentation and three-dimensional reconstruction of the liver based on 64-slice spiral CT scanning datal.Nanfang Yike Daxue Xuebao 2008;28:345-347
[20]Cline HE,Dumomlin CL,Hart HR Jr,Lorensen WE,Ludke S.3D reconstruction of the brain from magnetic resonance images using a connectivity algorithm.Magn Reson Imanging 1987;5:345-352
[21]Lee TY,Lin CH,Growing-club isosurface extraction algorithm for medical volume data.Computer Medimaging Graph 2001;25:405-415
[22]Wang k,denney t,Morrison E,Vodyanoy V.Construction of volume meshes from computed tomography data.Conf ProcIEEE Eng Med Boil Soc2005;5:5168-5171
[23]Delibasis KS,Matsopoulos GK,Mouravliansky NA,Nikita KS,A novel and efficient implementation of the marching cubes algorithm.Comput Med Imaging Graph 2001;25:343-352
[24]Lesk VI,Sternberg MJ,3D-garden:a system for modeling protein-protein complexes based on conformational refinement of ensembles generated with the marching cubes algorithm.Bioinformatics 2008;24:1137-1144
[25]Teferi S,Visualization,volume and surface area calculation of there dimensional(3D)ultrasound images.Ethiop Med J 2007;45:383-390
[26]Niu Q,Chi X,Leu MC,Ochoa J.Image processing geometric modeling and data management for development of a virtual bone surgery system.Computer Aided Surg 2008;13:30-40
[27]Schreiner J,Scheidegger CE,Sivla CT.High-quality extraction of isosurface from regular and irregular grids.IEEE Trans Vis Comput Graph 2006;12:1205-1212
[28]傅栋,靳安民.CT断层图像快速构建人体骨骼有限元几何模型的方法[J].中国组织工程研究与临床康复,2007.11(9):1620-1623
[29]陈新,陈良琴,徐伟.三维可视化技术及其在现代医学诊断中的应用[J].中国高校科技与产业化,2006.10(S1):306-309
[30]姚德民,宋志坚.虚拟内窥镜关键技术的研究与临床应用.生物医学工程学杂志,2008;25:18-22.
[31]高向军,田联房,王立非,毛宗源.虚拟内镜系统的关键技术分析.计算机应用研 究2008;25:1956-1959
[32]彭延军,王元红,郑永果.虚拟内镜系统研究进展.计算机工程与应用2005;41:203-205
[33]姚德民,宋志坚.交互式虚拟内镜成像及应用.解剖学杂志2005;28:192-194
[1]刘秋红.纤维结肠镜诊断结肠息肉及结肠癌231例[J].大理学院学报,2002.1(3):49-50
[2]罗立民.仿真医学影像学[J].中国医疗器械杂志,2002.26(2):79-81
[3]闵翔,谢德明.CT仿真内镜的应用[J].现代医学仪器与应用,2007.11(3):28-31
[4]王冠民,郭存英,张仰奇,等.CT仿真结肠镜与三维重建相结合对结肠病变的诊断价值[J].实用医学影像杂志,2004.5(3):151-153
[5]王岚.CT仿真内镜与纤维内镜及钡灌肠造影对结肠疾病的比较研究[J].黑龙江医学,2005.29(9):659-660
[6]张太发.基于CT医学图像的三维表面重建算法的研究[D].西安理工大学,2007.
[7]陈新,陈良琴,徐伟.三维可视化技术及其在现代医学诊断中的应用[J].中国高校科技与产业化,2006.10(S1):306-309
[8]傅栋,靳安民.CT断层图像快速构建人体骨骼有限元几何模型的方法[J].中国组织工程研究与临床康复,2007.11(9):1620-1623
[1]Phillips CL,Arend LJ,Filson AJ,et al.Three-dimensional imaging of Embryonic mouse kidney by two-photon microscopy[J].Am J Pathol,2001,158(1):49-55.
[2]Vlaik T,Yazdi HM,Burns BF,et at.Pattern of distribution of inlraduclal and infiltrating duclal cancinoma:A three dimensional stundy using serialcoronal giant sections of the breast[J].Hum Patrol,2000,31(4):464-474.
[3]李文刚,黄志强,李滨等.大鼠肝门部胆管周围血管丛的研究[J].中国普外基础与临床杂志,2005;12,(4)335-337.
[4]李文刚,黄志强,李滨等.大鼠门静脉动脉化肝门部胆管微血管的三维重建观察[J].消化外科,2004,3(6):411-414.
[5]Fenster A,Downey DB,Cardinal NH.Three dimension ultrasound imaging[J].Physics in Medicine and Biology,2001,46(1):67-99.
[6]Devore GR.Three-dimensional and four-dimensional fetal echocardiography:a new frontier[J].Curr Opin Pediatr.2005 Oct;17(5):592-604.
[7]沈学东,洪涛,周京敏等.体元模型动态三维重建超声显像的定量诊断价值[J].中国临床医学,1999,6(3)197--199.
[8]何爱军,汪天富,尹立雪等.多普勒血流超声医学图像的动态三维重建[J].航天医学与医学工程.2002,15(2):112-116.
[9]Kimura B J,Bhargara V,DeMarin AN.Value and limitations of intravascular ultrasound imaging in characterizing coronary atherosclerotic plaque[J].Am Heart J,1995,130:386-396.
[10]王新房,黄润青.实时三维超声心动图的应用现状及前景展望[J],中国医疗器械信息.2004,11(1):12-14.
[11]Hsu CY,Chiba Y,Fukui O.et al.Counterclockwise barber-pole sign on prenatal three-dimensional power Doppler sonography in a case of duodenal obstruction without intestinal malrotation[J].J Clin Ultrasound.2004,32(2):86-90.
[12]Bonsanto MM,Metzner R,Aschoff A.et al.3D ultrasound navigation in syrinx surgery--a feasibility study[J].Acta Neurochir(Wien).2005,147(5):533-540.
[13]Uchida T,Takada T,Ammori BJ,et al.Three-dimensional reconstruction of the ventral and dorsal pancreas:a new insight into anatomy and embryonic development[J].J Hepatobiliary Pancreat Surg.1999;6(2):176-180.
[14]苏焕群,林万莲,唐丹,等.概述数字化虚拟人体的研究[J].医学信息,2003,16(7):358-359
[15]Spitzer VM,Whitlock DG..The visible human dataset:the anatomical platform of human simulation[J].Anat Rec,1998,253:49-57.
[16]Spitzer VM,Ackerman MJ,Scherringer AL,et al.The visible human male:a technical report[J].J An Med Inform Assoc,1996,3:118-130.
[17]Ackeman NJ.The visible human project:a resource for education[J].Acad Med,1999,74:667-670.
[18]王旭,桂业英.可视人体(VH)数据集及使用[J].生物医学工程学杂志,1999,16(1:109-111.
[19]罗述谦.中国虚拟人医学图像方法学考虑[J].中国基础科学,2002,6:20-24.
[20]张绍祥,王平安,刘正津,等.首套中国男、女数字化可视人体结构数据的可视化研究[J].第三军医大学学报,2003,25(5):394-396.
[21]奇云.数字化虚拟中国人的研究及战略意义[J].物理通报,2003,4:1-3.
[22]罗述谦.虚拟人体研究与医学图像处理方法学[J].医疗设备信息,2002,17(5):1-3.
[23]方驰华,周五一,黄立伟等.虚拟中国人女性一号肝脏图像三维重建和虚拟手术的切割[J].中华外科杂志,2005,43(11):682-686.
[24]单锦露,张绍祥,刘正津等.女性盆腔的可视化研究及其意义[J].中国医学影像技术,2005,21(3):478-480.
[25]钟世镇.数字化虚拟人体的科学意义及应用前景[J].第一军医大学学 报,2003,23(3):193-195.
[26]李七渝,张绍祥,王平安等.人体大脑数字化解剖模型的构建及可视化[J].解剖学报,2005,36(6):638-641.
[27]Lutz C,Takagi A,Janecka IP,et al.Three dimensional computer reconstruction of a temporal bone[J].Otol aryngol Head Neck Surg,1989,101(5):522-256.
[28]邱明国,张绍祥,刘正李等.中国人侧颅底区可视化研究[J].解剖学报,2004,35(6):598-601.
[29]张伟国,陈现红,张绍祥等.枕下海绵窦可视化成像与MRI对照研究[J].中国医学影像技术,2005,21(5):809-811.
[30]张元智,顾立强,尹博等.虚拟中国人女号(VCH-F1)臂丛神经可视化初步研究[J].中华手外科杂志,2005,21(5):277-279.
[31]Guo YL,Heng PA,Zhang SX.et al.Thin sectional anatomy,three-dimensional reconstruction and visualization of the heart from the Chinese Visible Human[J].Surg Radiol Anat.2005,27(2):113-118.
[32]周泽民,方驰华,钟世镇等.基于数字化虚拟中国人女性一号的胰腺图像三维重建及可视化研究[J].中华外科杂志,2005,43(21):1401-1404.
[33]周泽民,方驰华,钟世镇等.数字化虚拟胰腺三维重建图像与传统解剖学图像的比较[J].第四军医大学学报,2005,26(18):1653-1656.
[34]方驰华,唐云强,鲁朝敏,等.胆囊切开取石和胆囊切除术的可视化研究[J].南方医科大学学报,2008,28(03):356
[35]方驰华,鲁朝敏,黄燕鹏,等.数字医学技术在肝癌外科诊治中的应用价值研究.中华外科杂志,2009,47(7):1-4.
[36]方驰华,鲁朝敏,黄燕鹏,等.仿真手术在肝移植2型肝动脉变异动脉重建中应用价值的研究.南方医科大学学报,2009,29(1):16-19.
[37]周泽民,方驰华,钟世镇等.数字化虚拟胰腺三维重建图像与传统解剖学图像的比较[J].第四军医大学学报,2005,26(18):1653-1656.
[38]邱明国,张绍祥,刘正津,等.耳三维重建及虚拟内窥镜[J].第三军医大学学报,2003,25(7):572-574.
[39]林鸿国,杨时鸿.虚拟人对中医药学发展的影响[].中医药学刊,2003,21(6):953-954.
[40]余安胜,张海东,李凤梅等.穴位标本断面图相配准方法研究[J].中国针灸,2003,23(2):89-91.
[41]钟沁,罗述谦.基于首例中国女性数字人的经穴三维重建[J].首都医科大学学报,Journal of Capital University of Medical Sciences,2004,25(3):281-283.
[42]刘红菊,赵静,严振国等.基于虚拟人技术的风池穴可视化研究[J].针刺研究,2004,29(4):282-285.