基于个人计算机的皮瓣三维可视化研究
|
摘要
研究背景
针对显微外科领域皮瓣的血管系统进行的解剖学研究已广泛开展，例如大体解剖、血管铸型等。多层CT(MSCTA)在临床实践中的成功应用，为我们开展皮瓣血管形态学研究提供了新的手段和发展空间。无损伤、直观性强的多层CT，在心脑血管病变、椎管形态学研究方面取得了一些成功经验，同时也为我们在皮瓣血管形态学研究方面也提供了影像学技术手段，加之计算机处理技术手段的不断完善，使得重建皮瓣的三维解剖成为必须和可能，这为立体的观察皮瓣血供情况以及制定最佳的手术入路和治疗方案奠定了基础，实现了由平面解剖向立体解剖和数字解剖的发展。
目前在临床一线，皮瓣的设计主要依赖于医生的临床经验，要确定皮瓣的位置、大小、几何形状以及与周围血管组织之间的空间关系，只能凭医生“在头脑中进行重建”，无法准确地定量分析。现代皮瓣外科学的发展迫切需要显示其血供等复杂结构和功能的三维图像系统，为临床医生的培训、手术入路的设计、手术过程的虚拟和术中的精确导航提供支持，同时随着人体标本来源的日趋减少，有必要建立一整套包括常用皮瓣、肌瓣和骨瓣在内的三维解剖以供骨科临床教学和训练所用，随着“虚拟人”、计算机虚拟技术和现代医学影像技术的快速发展，使得虚拟皮瓣外科手术的研究与应用成为可能。数字化技术可以缩短学习过程、有效的掌握三维解剖、有利于手术的设计与模拟训练、降低手术风险、提高皮瓣外科治疗水平。本研究采用现代影像学、计算医学、现代临床解剖学和皮瓣外科学相结合进行跨学科、多学科交叉研究；利用动脉造影技术、氧化铅血管灌注技术和数字化图像处理技术建立常用皮瓣的三维数字化模型；建立数字化皮瓣外科多媒体教学视频系列光盘，以供临床教学、手术设计等所用。该项目的研究，将“虚拟皮瓣”技术与三维CT数据相结合作用于学科领域，将单调的二维式训练、手术模拟及手术训练由平面变为立体，由静态变动态，由单面变多面，对立体、数字化皮瓣外科解剖学的建立和临床教学训练、临床诊疗水平的提高、手术设计及手术的模拟均有较大的帮助。
在以往的医学可视化系统中，一般都要使用昂贵的专业图形工作站，甚至使用多CPU处理器的超级计算机或使用分布式计算的方法来实现。各个影像设备的医学图像工作站，虽然有高速高性能的专用医学图像处理芯片和多CPU并行处理能力与相应的支持诸如OpenGL和Direct 3D等图形加速标准的专业级显卡的支持，以及自带的软件体系，但是其系统过于庞大，价格昂贵，且数据在不同厂商的MOD及其驱动器在容量和格式上不兼容，软件加密捆绑，另外鉴于CT、MRI和DSA等控制台计算机一般忙于日常多科室临床病人检查的信息输入、照相等操作，不能为专科临床医生提供独立的工作环境等等诸多的弊端，能否在个人计算机上进行皮瓣三维图像处理成为未来的趋势。个人计算机图形处理的主要特点是能在相对较低的价位满足一般医学工程应用，并且一些软件厂商已经把大部分主流应用软件部分或全部移植到个人计算机平台上。因此在本次研究中，我们初步探索利用个人计算机及单机版重建软件对皮瓣动脉二维扫描图像进行二维或三维分析处理，如对感兴趣区域的分割提取、三维可视化重建，结果满意，可以辅助医生对皮瓣结构及其它感兴趣区域进行定性直至准确的定量分析。
目的
1．通过薄层连续CTA扫描显示皮瓣动脉三维形态；
2．通过观测氧化铅灌注动脉连续横断面的走行，建立皮瓣动脉的可视化数字模型。
3．提供皮瓣正常三维动态解剖以供临床教学，建立部分常用皮瓣的三维数据库，有助于手术入路的选择和手术设计。
材料和方法
1．皮瓣可视化研究个人计算机系统：系统硬件平台：本系统采用ACER的便携式个人计算机，CPU为Intel CoreTM Duo Processor T2050(1.6GHZ，533MHzFSB，2MB，L2，cache，内存1GB的DDR2 533MHz模块，硬盘80G，NVIDIA GeForce Go 7300图形卡带128MB GDDR2 VRAM显存，显示器15.4英寸WXGA200-nitAcer CrystalBriteTM高亮TFT LCD，显示器最佳分辨率1280×800像素，16ms反应时间。软件：操作系统：windows XP；应用软件Photoshop7.0、Geomagic Studio 5.0和RapidForm2004，包含图像处理在内的各种应用工具箱；图像可视化研究平台：Amira软件。
2．下肢CTA造影皮瓣动脉可视化初步研究：图像来源：数据来源为南方医科大学珠江医院CT室采集的5例下肢CTA扫描数据集。采用philips 64层螺旋CT机，增强后行容积扫描，扫描范围从腹主动脉至足趾，管电压120 kV，管电流250～300 mA，层厚1mm×64，床速47.5 mm／圈，螺距0.891，图像重建层厚1mm，重建间隔0.5mm。增强对比剂均选用非离子型对比剂碘比乐，用高压注射器，注射流率为3.5～4.0ml／s，经肘静脉(利用套管针)快速加压静脉注射，注射对比剂总量1.5ml／kg。选主动脉层面，使用智能触发技术，CT值设为200～250HU，智能触发扫描延迟时间5s。使用philips 64层螺旋CT三维工作站“Brilliance”和个人计算机进行图像后处理，并进行对比。
3．氧化铅动脉灌注皮瓣可视化初步研究：图像来源：新鲜捐献成人尸体1具，采用改良的明胶—氧化铅灌注术经腹主动脉灌注，灌注后对全身行连续螺旋CT扫描，扫描条件：120kV，110mAs，参数：层厚2mm，矩阵512×512。利用Amira提供的画笔(Brush)、套索(Lasso)、魔术棒(Magic wand)和Blow四种图像分割工具，对皮肤、血管及骨骼结构：进行轮廓提取、生成面片，根据计算机的配置情况对轮廓线进行适当的简化，进行表面重建和体重建。
结果
1．下肢CTA造影皮瓣动脉可视化初步研究：肢体CTA技术为临床皮瓣设计提供新的思路，不但可以了解肢体血管的变异和病变，明确患者是否适合行游离皮瓣移植，而且提供了常规血管物理检查无法提供的3D信息，通过CTA数据实施肢体动脉三维可视化重建，可追朔皮瓣深部的动脉干来源和行程，特别是对施术中调整血管蒂的蒂长和径粗两方面，均有较大的灵活性，增强了可靠性，但是由于管径细小的动脉显影欠佳，重建效果还待进一步提高。应用三维重建方面，通过对比，个人计算机具有丰富的手动分割工具，自动或交互式图像分割与提取等功能，三维重建效果更加优秀，可满足临床需求。
2．氧化铅动脉灌注皮瓣可视化初步研究：基于个人计算机为平台，运用Amira4.1图像处理软件，对明胶—氧化铅混悬液的灌注切片数据进行动脉皮瓣三维重建的实现方法。本研究重建了皮瓣主要构成，重建的三维结构可以多彩色、透明或任意组合显示，经不同角度观察，整体显示清晰、实体感强，皮肤、动脉的相互关系一目了然，在三维表面重建的图像中可清楚地观察各解剖结构的形态，特别是皮瓣动脉的分支情况和体表的投影情况得到了很好的显示。在表面三维重建的静态显示的基础上，利用软件的MovieMaker模块创建电影文件(Creating movie files)，将其制作为电影，利用你选择的电影播放器(Windows媒体播放器或其它类似工具)，来播放观看形成的电影文件，画面清晰流畅，直观、可完成360度的旋转，立体的显示了其形态特征。
结论
1．薄层连续CTA扫描可以显示部分皮瓣血管的形态，三维重建可以提高临床医师对皮瓣观察和手术设计水平。
2．依据氧化铅CT数据集能够提供完整精确的皮瓣血管断面解剖，其三维重建为术前诊断和外科治疗提供了良好的形态学依据。
3．基于个人计算机的皮瓣可视化研究系统的实现，为临床医师提供了个性化的研究平台。
Study background
Anatomic studies on vascular system of the skin flap in the field of microsurgery have carried out widely, for example, the success of multilayer CT(MSCTA) in the clinic application, provide us a new method and developing vacuity. Aiming at non-harm and strong direct-viewing multilayer CT, successful experiences have been made from the view of heart and cerebrovascular lesion and vertebral canal morphology. At the same time, these studies have given us some imaging techniques and means on the research of skin flap blood vessel morphology, additionally, with the unceasing development of computer processing tech, re-establishment of skin flap 3D dissect coming to be possible. This study established a foundation for observation on the state of skin flap blood supply stereoscopically and making best operative route and treatment plan, implementing the development from plane anatomy to stereoscopic and digital anatomy.
Now, on the clinic, the design of skin flap mainly relies on the doctors experiences. The definition of the position, magnitude, geometry of the skin flap and the spatial relation of skin flap and peripheral vascular tissue can not be analyzed quantitatively exactly. The development of modern skin flap chirurgery needs badly the three- dimension picture system to display the blood supply and other profound structures and functions. This kind of system can also support the clinical training, designing of operative route, suppositional operation procedure and fidelity navigation during the operation. 3D imaging studies such as computed tomography (CT) and magnetic resonance (MR) provide the primary source of patient-specific data for such medical applications. The two-dimensional (2D) images produced by CT and MR imaging techniques represent a series of cross-sectional image slices through one's anatomy. When stacked on top of each other, these images represent the 3D structure of the one's anatomy. Two methods for reconstructing the 3D volume from the 2D images are surface rendering) and volume rendering. Surface rendering requires that the medical data be segmented, and that a geometric surface representation of the structures of interest be extracted. The advantage of this technique lies in the relatively small amount of contour data, resulting in fast rendering speeds. Shaded surface displays are useful if there is a need or desire to visualize specific 3D surfaces. Shaded surface displays are a 2D representation of a 3D surface. Volume rendering, by contrast, enables the visualization of 3D data without fitting geometric primitives to the data, but volume image data sets are characteristically large, taxing the computation abilities of volume-rendering techniques and the systems on which they are implemented. Of the methods available, SSD has been the most commonly used 3D technique, because it requires a small amount of data and can therefore be implemented on less powerful computers. This is why we chose SSD to display the outline of ALT flap and ADP flap. Our 3D reconstruction models perfectly displayed their anatomical characteristics. These models can be used in surgical simulation and medical training.
In the departed medical visualization techniques systems, expensive expert graphic
Workstations, super computers with more CPU or calculating distrubutedly are needed. It is a future trend that 3D image processing of the skin flap on PC. The main characteristics of PC image processing is that it can meet the need of general engineering application at a rather low price. And luckily, the manufactures of the software have transplanted main stream application software to the PC partly or totally. Accordingly, in this study, the 2D or 3D analyzes on the skin flap artery by use of PC and single rebuilding software. For example, the divisional extraction of the area-of-interest and 3D visualization, have come to satisfactory results, which can help the doctor make qualitative and quantitive analyses on the structure of skin flap and other area-of-interest.
Objectives
1. Three-dimensional computerized reconstructions of flap were conducted from these data by means of continuous thin-layer CTA
2.To observe the normal structure of flap artery in one adult fresh cadaver specimens perfused with lead oxide-gelatine mixture and to establish their digitized visible models.
3.The 3D models can providing morphologic data for clinical training, pre-operation designing and virtual reality operation procedure.
Materials and methods
1.The personal computer system ofthree-D visualization study: Platform: Intel CoreTM Duo processor T2050 (2 MB L2 cache, 1.6 GHz, 667MHz FSB); System memory: Up to 1GB of DDR2 533 MHz memory; Storage: 120 GB SATAJ100 hard disk drive; NVIDIA GeForce Go 7300 graphics with 128 MB of dedicated GDDR2 VRAM, supporting NVIDIA PowerMizer 6.0; 15.4" WXGA 200-nit Acer CrystalBriteTM highgraphics brightness TFT LCD, 1280 x 800 pixel resolution, 16 ms response time, supporting simultaneous multiwindow viewing via Acer GridVistaTM; Operating system Genuine Windows XP Home Edition (Service Pack 2); software:
Photoshop7.0、Geomagic Studio 5.0和RapidForm2004; the platform of three-D visualization study: Amira4.1.
2.Three-D visualization study on continuous thin-layer CTA: Five healthy adult (two men) underwent bilateral examination with thin-layer imager (philips 64CT) to assess the flap structures of the extremitas inferior in Zhujiang Hospital, Southern Medical University. They were subject to radiographic CT volume scanning. Extent was from abdominal aorta to toes. The date were tube tension 120 kV, tube current 250～300 mA, pachy 1mm×64, rapid 47.5 mm/circle, pitch 0.891, image reconstruction 1mm, 0.5mmo constrast medium 3.5～4.0ml/s, working station "Brilliance"
3.Three-D visualization study on CT section of the cadaver specimen: One man adult fresh cadaver specimen was perfused with lead oxide-gelatine mixture to be subject to radiographic CT scanning, scans covering both lower limbs with 1.0 mm collimation (120 kV, 110 mA, pitch 2mm, 512×512 matrix). The 2D images in Dicom format were correlated to the 3D models of the entire region by means of Amira software along with the cross-sectional images into standardized views. The segmented areas for skin, flap structures were displayed in different colors using tools of Brush, Lasso and Magic wand. Segmentation is performed using a custom-written discrete snake procedure. The 3D reconstruction via the Amira 4.1 software consisted of tracing the contours of the anatomical structures to be reconstructed; adjustment by geometrical alignment of the contours of stacked points; modeling of the surfaces by meshing the framework of the points transformed into polygons (wire-framed object) and smoothing the contours of the object reconstructed from points (surface rendering) . 3D models of these structures were saved as STL format, and then input Imageware 9.0 software to be re-touching and re-dividing.
Results
1.Three-D visualization study on continuous thin-layer CTA: The technique of continuous thin-layer CTA of limbs provided new way to clinicist who could get the message of variation and pathological changes. They identified that patients need free flap transplantation whether or not. They could get more 3D information than common blood vessel blood vessel. The source and trip of flap could be chased by means of three-D visualization. The length of blood vessel and diameter could be elected in operation. Because the leptos caliber could not be dev, three-D visualization was below the mark. Personal computer had affluent hand-motion tool, automatic and reciprocation extraction of diremption and extraction. Three-D visualization was very superiority.
2.Three-D visualization study on CT section of the cadaver specimen: The 3D images could display perfectly the main structures of flap and other adjacent structures by means of personal computer and software Amira4.1. Correlation of cross-sectional images from CT-scanning to 3D models is a more effective way of understanding of the flap anatomy. 3D reconstructed models of flap can be applied in clinical training, pre-operative designing and virtual operation procedures And also we transformed the images into movies with distinct and fluent frames by means of movie maker software.
Conclusions
1.Muti-silce helical CTA can display part of morphologic data of blood vessel. Three-D reconstruction can elevate observation of blood vessel and operative design of surgical flap for clinician
2.The cross-section anatomy of the normal structure of flap artery in one adult fresh cadaver specimens perfused with lead oxide-gelatine mixture can be observed. Their digitized visible models can provide upstanding morphologic evidence.
3.Three-D Visualization study on flap based on the operational platform of personal computer. It can provide the platform of the individual personality to clinicist.
针对显微外科领域皮瓣的血管系统进行的解剖学研究已广泛开展，例如大体解剖、血管铸型等。多层CT(MSCTA)在临床实践中的成功应用，为我们开展皮瓣血管形态学研究提供了新的手段和发展空间。无损伤、直观性强的多层CT，在心脑血管病变、椎管形态学研究方面取得了一些成功经验，同时也为我们在皮瓣血管形态学研究方面也提供了影像学技术手段，加之计算机处理技术手段的不断完善，使得重建皮瓣的三维解剖成为必须和可能，这为立体的观察皮瓣血供情况以及制定最佳的手术入路和治疗方案奠定了基础，实现了由平面解剖向立体解剖和数字解剖的发展。
目前在临床一线，皮瓣的设计主要依赖于医生的临床经验，要确定皮瓣的位置、大小、几何形状以及与周围血管组织之间的空间关系，只能凭医生“在头脑中进行重建”，无法准确地定量分析。现代皮瓣外科学的发展迫切需要显示其血供等复杂结构和功能的三维图像系统，为临床医生的培训、手术入路的设计、手术过程的虚拟和术中的精确导航提供支持，同时随着人体标本来源的日趋减少，有必要建立一整套包括常用皮瓣、肌瓣和骨瓣在内的三维解剖以供骨科临床教学和训练所用，随着“虚拟人”、计算机虚拟技术和现代医学影像技术的快速发展，使得虚拟皮瓣外科手术的研究与应用成为可能。数字化技术可以缩短学习过程、有效的掌握三维解剖、有利于手术的设计与模拟训练、降低手术风险、提高皮瓣外科治疗水平。本研究采用现代影像学、计算医学、现代临床解剖学和皮瓣外科学相结合进行跨学科、多学科交叉研究；利用动脉造影技术、氧化铅血管灌注技术和数字化图像处理技术建立常用皮瓣的三维数字化模型；建立数字化皮瓣外科多媒体教学视频系列光盘，以供临床教学、手术设计等所用。该项目的研究，将“虚拟皮瓣”技术与三维CT数据相结合作用于学科领域，将单调的二维式训练、手术模拟及手术训练由平面变为立体，由静态变动态，由单面变多面，对立体、数字化皮瓣外科解剖学的建立和临床教学训练、临床诊疗水平的提高、手术设计及手术的模拟均有较大的帮助。
在以往的医学可视化系统中，一般都要使用昂贵的专业图形工作站，甚至使用多CPU处理器的超级计算机或使用分布式计算的方法来实现。各个影像设备的医学图像工作站，虽然有高速高性能的专用医学图像处理芯片和多CPU并行处理能力与相应的支持诸如OpenGL和Direct 3D等图形加速标准的专业级显卡的支持，以及自带的软件体系，但是其系统过于庞大，价格昂贵，且数据在不同厂商的MOD及其驱动器在容量和格式上不兼容，软件加密捆绑，另外鉴于CT、MRI和DSA等控制台计算机一般忙于日常多科室临床病人检查的信息输入、照相等操作，不能为专科临床医生提供独立的工作环境等等诸多的弊端，能否在个人计算机上进行皮瓣三维图像处理成为未来的趋势。个人计算机图形处理的主要特点是能在相对较低的价位满足一般医学工程应用，并且一些软件厂商已经把大部分主流应用软件部分或全部移植到个人计算机平台上。因此在本次研究中，我们初步探索利用个人计算机及单机版重建软件对皮瓣动脉二维扫描图像进行二维或三维分析处理，如对感兴趣区域的分割提取、三维可视化重建，结果满意，可以辅助医生对皮瓣结构及其它感兴趣区域进行定性直至准确的定量分析。
目的
1．通过薄层连续CTA扫描显示皮瓣动脉三维形态；
2．通过观测氧化铅灌注动脉连续横断面的走行，建立皮瓣动脉的可视化数字模型。
3．提供皮瓣正常三维动态解剖以供临床教学，建立部分常用皮瓣的三维数据库，有助于手术入路的选择和手术设计。
材料和方法
1．皮瓣可视化研究个人计算机系统：系统硬件平台：本系统采用ACER的便携式个人计算机，CPU为Intel CoreTM Duo Processor T2050(1.6GHZ，533MHzFSB，2MB，L2，cache，内存1GB的DDR2 533MHz模块，硬盘80G，NVIDIA GeForce Go 7300图形卡带128MB GDDR2 VRAM显存，显示器15.4英寸WXGA200-nitAcer CrystalBriteTM高亮TFT LCD，显示器最佳分辨率1280×800像素，16ms反应时间。软件：操作系统：windows XP；应用软件Photoshop7.0、Geomagic Studio 5.0和RapidForm2004，包含图像处理在内的各种应用工具箱；图像可视化研究平台：Amira软件。
2．下肢CTA造影皮瓣动脉可视化初步研究：图像来源：数据来源为南方医科大学珠江医院CT室采集的5例下肢CTA扫描数据集。采用philips 64层螺旋CT机，增强后行容积扫描，扫描范围从腹主动脉至足趾，管电压120 kV，管电流250～300 mA，层厚1mm×64，床速47.5 mm／圈，螺距0.891，图像重建层厚1mm，重建间隔0.5mm。增强对比剂均选用非离子型对比剂碘比乐，用高压注射器，注射流率为3.5～4.0ml／s，经肘静脉(利用套管针)快速加压静脉注射，注射对比剂总量1.5ml／kg。选主动脉层面，使用智能触发技术，CT值设为200～250HU，智能触发扫描延迟时间5s。使用philips 64层螺旋CT三维工作站“Brilliance”和个人计算机进行图像后处理，并进行对比。
3．氧化铅动脉灌注皮瓣可视化初步研究：图像来源：新鲜捐献成人尸体1具，采用改良的明胶—氧化铅灌注术经腹主动脉灌注，灌注后对全身行连续螺旋CT扫描，扫描条件：120kV，110mAs，参数：层厚2mm，矩阵512×512。利用Amira提供的画笔(Brush)、套索(Lasso)、魔术棒(Magic wand)和Blow四种图像分割工具，对皮肤、血管及骨骼结构：进行轮廓提取、生成面片，根据计算机的配置情况对轮廓线进行适当的简化，进行表面重建和体重建。
结果
1．下肢CTA造影皮瓣动脉可视化初步研究：肢体CTA技术为临床皮瓣设计提供新的思路，不但可以了解肢体血管的变异和病变，明确患者是否适合行游离皮瓣移植，而且提供了常规血管物理检查无法提供的3D信息，通过CTA数据实施肢体动脉三维可视化重建，可追朔皮瓣深部的动脉干来源和行程，特别是对施术中调整血管蒂的蒂长和径粗两方面，均有较大的灵活性，增强了可靠性，但是由于管径细小的动脉显影欠佳，重建效果还待进一步提高。应用三维重建方面，通过对比，个人计算机具有丰富的手动分割工具，自动或交互式图像分割与提取等功能，三维重建效果更加优秀，可满足临床需求。
2．氧化铅动脉灌注皮瓣可视化初步研究：基于个人计算机为平台，运用Amira4.1图像处理软件，对明胶—氧化铅混悬液的灌注切片数据进行动脉皮瓣三维重建的实现方法。本研究重建了皮瓣主要构成，重建的三维结构可以多彩色、透明或任意组合显示，经不同角度观察，整体显示清晰、实体感强，皮肤、动脉的相互关系一目了然，在三维表面重建的图像中可清楚地观察各解剖结构的形态，特别是皮瓣动脉的分支情况和体表的投影情况得到了很好的显示。在表面三维重建的静态显示的基础上，利用软件的MovieMaker模块创建电影文件(Creating movie files)，将其制作为电影，利用你选择的电影播放器(Windows媒体播放器或其它类似工具)，来播放观看形成的电影文件，画面清晰流畅，直观、可完成360度的旋转，立体的显示了其形态特征。
结论
1．薄层连续CTA扫描可以显示部分皮瓣血管的形态，三维重建可以提高临床医师对皮瓣观察和手术设计水平。
2．依据氧化铅CT数据集能够提供完整精确的皮瓣血管断面解剖，其三维重建为术前诊断和外科治疗提供了良好的形态学依据。
3．基于个人计算机的皮瓣可视化研究系统的实现，为临床医师提供了个性化的研究平台。
Study background
Anatomic studies on vascular system of the skin flap in the field of microsurgery have carried out widely, for example, the success of multilayer CT(MSCTA) in the clinic application, provide us a new method and developing vacuity. Aiming at non-harm and strong direct-viewing multilayer CT, successful experiences have been made from the view of heart and cerebrovascular lesion and vertebral canal morphology. At the same time, these studies have given us some imaging techniques and means on the research of skin flap blood vessel morphology, additionally, with the unceasing development of computer processing tech, re-establishment of skin flap 3D dissect coming to be possible. This study established a foundation for observation on the state of skin flap blood supply stereoscopically and making best operative route and treatment plan, implementing the development from plane anatomy to stereoscopic and digital anatomy.
Now, on the clinic, the design of skin flap mainly relies on the doctors experiences. The definition of the position, magnitude, geometry of the skin flap and the spatial relation of skin flap and peripheral vascular tissue can not be analyzed quantitatively exactly. The development of modern skin flap chirurgery needs badly the three- dimension picture system to display the blood supply and other profound structures and functions. This kind of system can also support the clinical training, designing of operative route, suppositional operation procedure and fidelity navigation during the operation. 3D imaging studies such as computed tomography (CT) and magnetic resonance (MR) provide the primary source of patient-specific data for such medical applications. The two-dimensional (2D) images produced by CT and MR imaging techniques represent a series of cross-sectional image slices through one's anatomy. When stacked on top of each other, these images represent the 3D structure of the one's anatomy. Two methods for reconstructing the 3D volume from the 2D images are surface rendering) and volume rendering. Surface rendering requires that the medical data be segmented, and that a geometric surface representation of the structures of interest be extracted. The advantage of this technique lies in the relatively small amount of contour data, resulting in fast rendering speeds. Shaded surface displays are useful if there is a need or desire to visualize specific 3D surfaces. Shaded surface displays are a 2D representation of a 3D surface. Volume rendering, by contrast, enables the visualization of 3D data without fitting geometric primitives to the data, but volume image data sets are characteristically large, taxing the computation abilities of volume-rendering techniques and the systems on which they are implemented. Of the methods available, SSD has been the most commonly used 3D technique, because it requires a small amount of data and can therefore be implemented on less powerful computers. This is why we chose SSD to display the outline of ALT flap and ADP flap. Our 3D reconstruction models perfectly displayed their anatomical characteristics. These models can be used in surgical simulation and medical training.
In the departed medical visualization techniques systems, expensive expert graphic
Workstations, super computers with more CPU or calculating distrubutedly are needed. It is a future trend that 3D image processing of the skin flap on PC. The main characteristics of PC image processing is that it can meet the need of general engineering application at a rather low price. And luckily, the manufactures of the software have transplanted main stream application software to the PC partly or totally. Accordingly, in this study, the 2D or 3D analyzes on the skin flap artery by use of PC and single rebuilding software. For example, the divisional extraction of the area-of-interest and 3D visualization, have come to satisfactory results, which can help the doctor make qualitative and quantitive analyses on the structure of skin flap and other area-of-interest.
Objectives
1. Three-dimensional computerized reconstructions of flap were conducted from these data by means of continuous thin-layer CTA
2.To observe the normal structure of flap artery in one adult fresh cadaver specimens perfused with lead oxide-gelatine mixture and to establish their digitized visible models.
3.The 3D models can providing morphologic data for clinical training, pre-operation designing and virtual reality operation procedure.
Materials and methods
1.The personal computer system ofthree-D visualization study: Platform: Intel CoreTM Duo processor T2050 (2 MB L2 cache, 1.6 GHz, 667MHz FSB); System memory: Up to 1GB of DDR2 533 MHz memory; Storage: 120 GB SATAJ100 hard disk drive; NVIDIA GeForce Go 7300 graphics with 128 MB of dedicated GDDR2 VRAM, supporting NVIDIA PowerMizer 6.0; 15.4" WXGA 200-nit Acer CrystalBriteTM highgraphics brightness TFT LCD, 1280 x 800 pixel resolution, 16 ms response time, supporting simultaneous multiwindow viewing via Acer GridVistaTM; Operating system Genuine Windows XP Home Edition (Service Pack 2); software:
Photoshop7.0、Geomagic Studio 5.0和RapidForm2004; the platform of three-D visualization study: Amira4.1.
2.Three-D visualization study on continuous thin-layer CTA: Five healthy adult (two men) underwent bilateral examination with thin-layer imager (philips 64CT) to assess the flap structures of the extremitas inferior in Zhujiang Hospital, Southern Medical University. They were subject to radiographic CT volume scanning. Extent was from abdominal aorta to toes. The date were tube tension 120 kV, tube current 250～300 mA, pachy 1mm×64, rapid 47.5 mm/circle, pitch 0.891, image reconstruction 1mm, 0.5mmo constrast medium 3.5～4.0ml/s, working station "Brilliance"
3.Three-D visualization study on CT section of the cadaver specimen: One man adult fresh cadaver specimen was perfused with lead oxide-gelatine mixture to be subject to radiographic CT scanning, scans covering both lower limbs with 1.0 mm collimation (120 kV, 110 mA, pitch 2mm, 512×512 matrix). The 2D images in Dicom format were correlated to the 3D models of the entire region by means of Amira software along with the cross-sectional images into standardized views. The segmented areas for skin, flap structures were displayed in different colors using tools of Brush, Lasso and Magic wand. Segmentation is performed using a custom-written discrete snake procedure. The 3D reconstruction via the Amira 4.1 software consisted of tracing the contours of the anatomical structures to be reconstructed; adjustment by geometrical alignment of the contours of stacked points; modeling of the surfaces by meshing the framework of the points transformed into polygons (wire-framed object) and smoothing the contours of the object reconstructed from points (surface rendering) . 3D models of these structures were saved as STL format, and then input Imageware 9.0 software to be re-touching and re-dividing.
Results
1.Three-D visualization study on continuous thin-layer CTA: The technique of continuous thin-layer CTA of limbs provided new way to clinicist who could get the message of variation and pathological changes. They identified that patients need free flap transplantation whether or not. They could get more 3D information than common blood vessel blood vessel. The source and trip of flap could be chased by means of three-D visualization. The length of blood vessel and diameter could be elected in operation. Because the leptos caliber could not be dev, three-D visualization was below the mark. Personal computer had affluent hand-motion tool, automatic and reciprocation extraction of diremption and extraction. Three-D visualization was very superiority.
2.Three-D visualization study on CT section of the cadaver specimen: The 3D images could display perfectly the main structures of flap and other adjacent structures by means of personal computer and software Amira4.1. Correlation of cross-sectional images from CT-scanning to 3D models is a more effective way of understanding of the flap anatomy. 3D reconstructed models of flap can be applied in clinical training, pre-operative designing and virtual operation procedures And also we transformed the images into movies with distinct and fluent frames by means of movie maker software.
Conclusions
1.Muti-silce helical CTA can display part of morphologic data of blood vessel. Three-D reconstruction can elevate observation of blood vessel and operative design of surgical flap for clinician
2.The cross-section anatomy of the normal structure of flap artery in one adult fresh cadaver specimens perfused with lead oxide-gelatine mixture can be observed. Their digitized visible models can provide upstanding morphologic evidence.
3.Three-D Visualization study on flap based on the operational platform of personal computer. It can provide the platform of the individual personality to clinicist.
引文
1 Kass M, Witkin A, Terzopoulos D. Snakes: Active Contour Models, International Journal of Computer Vision 1988, 1:321-331
2 Gibson S, Fyock C, Grimson E, Kanade T, Kikinis R, Lauer H, McKenzie N, Mor A, Nakajima S, Ohkami H, Osborne R, Samosky J, Sawada A .Volumetric object modeling for surgical simulation. Medical Image Analysis 1998, 2: 121-132
3 Lorensen WE, Cline HE. Marching cubes: A high-resolution 3-D surface construction algorithm. Computer Graphics 1987, 21:163-169
4 张元智，李严兵，唐茂林，等．数字化与虚拟现实技术在皮瓣移植中的应用．中华创伤骨科杂志，2006，8(6)：501-504
5 张志浩，李严斌，梅劲，等．应用放射造影术进行血管3D可视化研究初探．临床应用解剖学杂志，2006，24(2)：255-258
6 Choplin RH, Buckwalter KA, Rydberg F, Farbe FM. CT with 3D rendering of the tendons of the foot and ankle: technique, normal anatomy, and disease. Radiographics 2004, 24: 343-356
7 Tang M, Geddes CR, Yang D, et al. Modified lead oxide-gelatin injection technique for vascular studies [J]. J Clin Anat, 2002, 1(1) : 73-78
8 楼新法，梅劲，Christopher R．Geddes2，等．明胶-氧化铅血管造影术的优化．中国临床解剖学杂志，2006，24(2)：259-262
9 高宏建，吴水才，任新颖．利用Amira进行肝脏及其管道系统的三维重建．北京生物医学工程，2006，25(5)：534-537
10 Amira 4.1 Manual (PDF). 2005. MERCURY (Amira 用户指南)
11 Levoy M .Efficient Ray Tracing of Volume Data. ACM Transactions on Graphics 1990, 9: 245-261
12 徐达传，钟世镇，刘牧之，等．股前外侧皮瓣的解剖学．临床应用解剖学杂志，1984， 23 (3)：158
13 罗庚，郭现辉，欧阳杰，等．股前外侧皮瓣在修复踝部创面中的应用．实用手外科杂志，2005，19(3)：169
14 孙宝，东乔群．股前外侧皮瓣的临床应用．整形再造外科杂志，2005，2(2)：111-114
15 Young DM, Trabulsy PP, Anthony JP.The need for preoperative leg angiography in fibula free flaps·J ReconstrMicrosurg, 1994, 10: 283-287
16 Margiotta MS, Markowitz B, ShawW-Routine angiography in free fibula flap reconstruction·Plast SurgForum, 1997, 20: 102-107
17 Carroll WR, Esclamado R·Preoperative vascular imaging for the fibular osteocuatneous flap·Arch OtolaryngolHead Neck Surg, 1996, 122: 708-712
18 钟世镇．数字人和数字解剖学[M]．济南：山东科学技术出版社，2004．163．
19 钟世镇．“虚拟中国人”(VCH)切片建模研究进展．中国临床解剖学杂志，2002，20：323．
20 Warrick PA, Funnell WR. A VRML-based anatomical visualization tool for medical education. IEEE Trans Inf Technol Biomed 1998, 2:55-61
21 Haluck RS, Krummel TM. Computers and virtual reality for surgical education in the 21 st century. Arch Surg 2000, 135: 786-792
22 Zheng YJ. A survey on evaluation methods for image segmentation[J]. Pattern Recognition, 1996, 29 (8) : 1335-1346
23 候春林，顾玉东，等．皮瓣外科学[M]．上海：上海科学技术出版社，2006．393
24 秦笃烈，原林．数字化虚拟中国人女性-1VCH F-1实验数据集血管标识的突破进展[J]．科学中国人，2003，4：4-8．
25 毕思文．数字人体与医学[J]．生物医学工程与临床，2005，9(3)：179-184．
26 张志浩，李严斌，梅劲，等．应用放射造影术进行血管3D可视化研究初探[J]．中国 临床解剖学杂志，2006，24(2)：255-258
27 钟世镇．“虚拟中国人”(VCH)切片建模研究进展．中国临床解剖学杂志，2002，20：323．
28 张元智，顾立强，原林，等．腰骶丛神经的断层解剖学及可视化初步研究．中华创伤骨科杂志，2004，6(12)：1362-1364
29 张元智，顾立强，原林，等．虚拟中国人女1号臂丛神经可视化初步研究．中华手外科杂志，2005，21(5)：277-279
30 Tang M (唐茂林), Geddes CR, Yang D (杨大平), et al. Modified lead oxide-gelatin injection technique for vascular studies [J]. J Clin Anat, 2002, 1(1) : 73-78.
31 邝石峰，陈彦震，黄广香，等．以足背皮瓣为基础的复合组织瓣的应用．广东医学．2001，22(2)：149-150
32 刘黎星，李钟峰，林建生，等．足背皮瓣的解剖学基础和临床应用．齐齐哈尔医学院学报．2001，22(9)：984-985
33 Mairesse JL, Mestdagh H, Bailleul JP, Letendart J, Dabrowski A. The arterial blood supply of the skin flap of the dorsal foot. ActaAnat (Basel) 1983, 115: 296-301
34 杨大平，唐茂林，Christopher R．Geddes，等．皮肤穿支血管的解剖学研究 中国临床解剖学杂志 2006，24(2)：232-235
1 张元智，李严兵，唐茂林，等．数字化与虚拟现实技术在皮瓣移植中的应用．中华创伤骨科杂志，2006，8(6)：501-504
2 张志浩，李严斌，梅劲，等．应用放射造影术进行血管3D可视化研究初探．临床应用解剖学杂志，2006，24(2)：255-258
3 毕思文．数字人体——人体系统数字学总论，中国医学影像技术，2003，19(204)：1-8
4 毕思文、王秀利．数字人体原型——人体系统，中国医学影像技术，2003，19(2)：140-144．
5 钟世镇．“虚拟中国人”(VCH)切片建模研究进展．中国临床解剖学杂志，2002，20：323．
6 钟世镇，原林，黄文华．数字化虚拟人体为临床解剖学开拓研究新领域．中国临床解剖学杂志，2002，20：3-4
7 原林，黄文华，唐雷，等．可视虚拟人研究概况．中国临床解剖学杂志，2002，20：341-342．
8 毕思文、王秀利．数字人体研究的方法论，中国医学影像技术，2003，19(3)：269—272．
9 毕思文等．数字人体(人体系统数字学)的概念、框架、内涵和应用，北京：第九届全国医药信息学大会论文集，2002．7．
10 郭森慧，王极，刘建毅，数字人体与人工智能—生命科学与信息科学的融合中国医学影像技术，2003，19(204)：22-24．
11 Levoy M .Efficient Ray Tracing of Volume Data. ACM Transactions on Graphics 1990; 9:245-261
12 Robb, RA. Three-Dimensional Visualization in Medicine and Biology. Book Chapter in: Handbook of Medical Imaging: Processing and Analysis, ed. Isaac N. Bankman, Academic Press, San Diego, CA, Chapter 2002; 42:685-712
13 Levoy M. Display of surfaces from volume data. IEEE Computer Graphics and Applications 1988; 8:29-37
14 Gibson S, Fyock C, Grimson E, Kanade T, Kikinis R, Lauer H, McKenzie N, Mor-A, Nakajima S, Ohkami H, Osborne R, Samosky J, Sawada A .Volumetric object modeling for surgical simulation. Medical Image Analysis 1998; 2:121-132
15 毕思文．数字人体简单巨系统的几种数学模型，中国医学影像技术．2003，19(11)：1490—1493．
16 Kass M, Witkin A, Terzopoulos D. Snakes: Active Contour Models, International Journal of Computer Vision 1988; 1:321-331
17 Warrick PA, Funnell WR. A VRML-based anatomical visualization tool for medical education. IEEE Trans Inf Technol Biomed 1998; 2:55-61
18 Haluck RS, Krummel TM. Computers and virtual reality for surgical education in the 21st century. Arch Surg 2000;135:786-792
19 Friedl R, Preisack MB, Klas W, Rose T, Stracke S, Quast KJ, Hannekum A, Godje O .Virtual Reality and 3D Visualizations in Heart Surgery Education. The Heart Surgery Forum 2002; 5:E17-E21
20 Smith DM, Aston SJ, Cutting CB, et al. Designing a virtualreality model for aesthetic surgery. Plant Reconstr Surg. 2005, 116:839-897
21 Chou B, Handa VL. Simulators and virtual reality in surgical education. Obstet Gynecol Clin North Am, 2006, 33:283-296.
22 Tang M, Geddes CR., Yang D, et al. Modified lead oxide-gelatin injection technique for vascular studies [J]. J Clin Anat, 2002, 1(1): 73-78
23 黄文华，原林，唐雷，等．“虚拟中国人体Ⅰ号”铣削前尸体材料的预处理．中国临床解剖学杂志，2002，20：336-337．
24 漆光平，姜平，徐焕俐．人体断层标本制作法。湖南医科大学学报，1999，31(1)：封三
25 Amira 4.1 Manual (PDF). 2005. MERCURY
26 原林，黄文华，唐雷，等．数字化虚拟中国人女性1号数据图像处理．中国临床解剖学杂志，2003，21(3)：193-96
27 高宏建，吴水才，任新颖．利用Amira进行肝脏及其管道系统的三维重建．北京生物医学工程，2006，25(5)：534-537
28 Zheng YJ. A survey on evaluation methods for image segmentation[J]. Pattern Recognition, 1996, 29 (8): 1335-1346
29 Kass M, Witkin A, Terzopoulos D. Snakes: Active Contour Models, International Journal of Computer Vision 1988; 1:321-331
30 Pflesser B, Tiede U, HohneKH. Specification, modelling and visualization of arbitrarily shaped cut surfaces in the volume model. Medical Image Computing and Computer-Assisted Intervention - MICCAI '98 1998; 1496:
2 Gibson S, Fyock C, Grimson E, Kanade T, Kikinis R, Lauer H, McKenzie N, Mor A, Nakajima S, Ohkami H, Osborne R, Samosky J, Sawada A .Volumetric object modeling for surgical simulation. Medical Image Analysis 1998, 2: 121-132
3 Lorensen WE, Cline HE. Marching cubes: A high-resolution 3-D surface construction algorithm. Computer Graphics 1987, 21:163-169
4 张元智，李严兵，唐茂林，等．数字化与虚拟现实技术在皮瓣移植中的应用．中华创伤骨科杂志，2006，8(6)：501-504
5 张志浩，李严斌，梅劲，等．应用放射造影术进行血管3D可视化研究初探．临床应用解剖学杂志，2006，24(2)：255-258
6 Choplin RH, Buckwalter KA, Rydberg F, Farbe FM. CT with 3D rendering of the tendons of the foot and ankle: technique, normal anatomy, and disease. Radiographics 2004, 24: 343-356
7 Tang M, Geddes CR, Yang D, et al. Modified lead oxide-gelatin injection technique for vascular studies [J]. J Clin Anat, 2002, 1(1) : 73-78
8 楼新法，梅劲，Christopher R．Geddes2，等．明胶-氧化铅血管造影术的优化．中国临床解剖学杂志，2006，24(2)：259-262
9 高宏建，吴水才，任新颖．利用Amira进行肝脏及其管道系统的三维重建．北京生物医学工程，2006，25(5)：534-537
10 Amira 4.1 Manual (PDF). 2005. MERCURY (Amira 用户指南)
11 Levoy M .Efficient Ray Tracing of Volume Data. ACM Transactions on Graphics 1990, 9: 245-261
12 徐达传，钟世镇，刘牧之，等．股前外侧皮瓣的解剖学．临床应用解剖学杂志，1984， 23 (3)：158
13 罗庚，郭现辉，欧阳杰，等．股前外侧皮瓣在修复踝部创面中的应用．实用手外科杂志，2005，19(3)：169
14 孙宝，东乔群．股前外侧皮瓣的临床应用．整形再造外科杂志，2005，2(2)：111-114
15 Young DM, Trabulsy PP, Anthony JP.The need for preoperative leg angiography in fibula free flaps·J ReconstrMicrosurg, 1994, 10: 283-287
16 Margiotta MS, Markowitz B, ShawW-Routine angiography in free fibula flap reconstruction·Plast SurgForum, 1997, 20: 102-107
17 Carroll WR, Esclamado R·Preoperative vascular imaging for the fibular osteocuatneous flap·Arch OtolaryngolHead Neck Surg, 1996, 122: 708-712
18 钟世镇．数字人和数字解剖学[M]．济南：山东科学技术出版社，2004．163．
19 钟世镇．“虚拟中国人”(VCH)切片建模研究进展．中国临床解剖学杂志，2002，20：323．
20 Warrick PA, Funnell WR. A VRML-based anatomical visualization tool for medical education. IEEE Trans Inf Technol Biomed 1998, 2:55-61
21 Haluck RS, Krummel TM. Computers and virtual reality for surgical education in the 21 st century. Arch Surg 2000, 135: 786-792
22 Zheng YJ. A survey on evaluation methods for image segmentation[J]. Pattern Recognition, 1996, 29 (8) : 1335-1346
23 候春林，顾玉东，等．皮瓣外科学[M]．上海：上海科学技术出版社，2006．393
24 秦笃烈，原林．数字化虚拟中国人女性-1VCH F-1实验数据集血管标识的突破进展[J]．科学中国人，2003，4：4-8．
25 毕思文．数字人体与医学[J]．生物医学工程与临床，2005，9(3)：179-184．
26 张志浩，李严斌，梅劲，等．应用放射造影术进行血管3D可视化研究初探[J]．中国 临床解剖学杂志，2006，24(2)：255-258
27 钟世镇．“虚拟中国人”(VCH)切片建模研究进展．中国临床解剖学杂志，2002，20：323．
28 张元智，顾立强，原林，等．腰骶丛神经的断层解剖学及可视化初步研究．中华创伤骨科杂志，2004，6(12)：1362-1364
29 张元智，顾立强，原林，等．虚拟中国人女1号臂丛神经可视化初步研究．中华手外科杂志，2005，21(5)：277-279
30 Tang M (唐茂林), Geddes CR, Yang D (杨大平), et al. Modified lead oxide-gelatin injection technique for vascular studies [J]. J Clin Anat, 2002, 1(1) : 73-78.
31 邝石峰，陈彦震，黄广香，等．以足背皮瓣为基础的复合组织瓣的应用．广东医学．2001，22(2)：149-150
32 刘黎星，李钟峰，林建生，等．足背皮瓣的解剖学基础和临床应用．齐齐哈尔医学院学报．2001，22(9)：984-985
33 Mairesse JL, Mestdagh H, Bailleul JP, Letendart J, Dabrowski A. The arterial blood supply of the skin flap of the dorsal foot. ActaAnat (Basel) 1983, 115: 296-301
34 杨大平，唐茂林，Christopher R．Geddes，等．皮肤穿支血管的解剖学研究 中国临床解剖学杂志 2006，24(2)：232-235
1 张元智，李严兵，唐茂林，等．数字化与虚拟现实技术在皮瓣移植中的应用．中华创伤骨科杂志，2006，8(6)：501-504
2 张志浩，李严斌，梅劲，等．应用放射造影术进行血管3D可视化研究初探．临床应用解剖学杂志，2006，24(2)：255-258
3 毕思文．数字人体——人体系统数字学总论，中国医学影像技术，2003，19(204)：1-8
4 毕思文、王秀利．数字人体原型——人体系统，中国医学影像技术，2003，19(2)：140-144．
5 钟世镇．“虚拟中国人”(VCH)切片建模研究进展．中国临床解剖学杂志，2002，20：323．
6 钟世镇，原林，黄文华．数字化虚拟人体为临床解剖学开拓研究新领域．中国临床解剖学杂志，2002，20：3-4
7 原林，黄文华，唐雷，等．可视虚拟人研究概况．中国临床解剖学杂志，2002，20：341-342．
8 毕思文、王秀利．数字人体研究的方法论，中国医学影像技术，2003，19(3)：269—272．
9 毕思文等．数字人体(人体系统数字学)的概念、框架、内涵和应用，北京：第九届全国医药信息学大会论文集，2002．7．
10 郭森慧，王极，刘建毅，数字人体与人工智能—生命科学与信息科学的融合中国医学影像技术，2003，19(204)：22-24．
11 Levoy M .Efficient Ray Tracing of Volume Data. ACM Transactions on Graphics 1990; 9:245-261
12 Robb, RA. Three-Dimensional Visualization in Medicine and Biology. Book Chapter in: Handbook of Medical Imaging: Processing and Analysis, ed. Isaac N. Bankman, Academic Press, San Diego, CA, Chapter 2002; 42:685-712
13 Levoy M. Display of surfaces from volume data. IEEE Computer Graphics and Applications 1988; 8:29-37
14 Gibson S, Fyock C, Grimson E, Kanade T, Kikinis R, Lauer H, McKenzie N, Mor-A, Nakajima S, Ohkami H, Osborne R, Samosky J, Sawada A .Volumetric object modeling for surgical simulation. Medical Image Analysis 1998; 2:121-132
15 毕思文．数字人体简单巨系统的几种数学模型，中国医学影像技术．2003，19(11)：1490—1493．
16 Kass M, Witkin A, Terzopoulos D. Snakes: Active Contour Models, International Journal of Computer Vision 1988; 1:321-331
17 Warrick PA, Funnell WR. A VRML-based anatomical visualization tool for medical education. IEEE Trans Inf Technol Biomed 1998; 2:55-61
18 Haluck RS, Krummel TM. Computers and virtual reality for surgical education in the 21st century. Arch Surg 2000;135:786-792
19 Friedl R, Preisack MB, Klas W, Rose T, Stracke S, Quast KJ, Hannekum A, Godje O .Virtual Reality and 3D Visualizations in Heart Surgery Education. The Heart Surgery Forum 2002; 5:E17-E21
20 Smith DM, Aston SJ, Cutting CB, et al. Designing a virtualreality model for aesthetic surgery. Plant Reconstr Surg. 2005, 116:839-897
21 Chou B, Handa VL. Simulators and virtual reality in surgical education. Obstet Gynecol Clin North Am, 2006, 33:283-296.
22 Tang M, Geddes CR., Yang D, et al. Modified lead oxide-gelatin injection technique for vascular studies [J]. J Clin Anat, 2002, 1(1): 73-78
23 黄文华，原林，唐雷，等．“虚拟中国人体Ⅰ号”铣削前尸体材料的预处理．中国临床解剖学杂志，2002，20：336-337．
24 漆光平，姜平，徐焕俐．人体断层标本制作法。湖南医科大学学报，1999，31(1)：封三
25 Amira 4.1 Manual (PDF). 2005. MERCURY
26 原林，黄文华，唐雷，等．数字化虚拟中国人女性1号数据图像处理．中国临床解剖学杂志，2003，21(3)：193-96
27 高宏建，吴水才，任新颖．利用Amira进行肝脏及其管道系统的三维重建．北京生物医学工程，2006，25(5)：534-537
28 Zheng YJ. A survey on evaluation methods for image segmentation[J]. Pattern Recognition, 1996, 29 (8): 1335-1346
29 Kass M, Witkin A, Terzopoulos D. Snakes: Active Contour Models, International Journal of Computer Vision 1988; 1:321-331
30 Pflesser B, Tiede U, HohneKH. Specification, modelling and visualization of arbitrarily shaped cut surfaces in the volume model. Medical Image Computing and Computer-Assisted Intervention - MICCAI '98 1998; 1496: