請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/64685
標題: | 雙重超音波影像之主動輪廓分割與影像追蹤進行肝臟三維輪廓重建 3D Liver Contour Reconstruction from Dual Ultrasound Slices Using Active Contour Model Segmentation and Image-Guided Tracking |
作者: | Feng-Chih Lin 林豐池 |
指導教授: | 連豊力(Feng-Li Lian) |
關鍵字: | 超音波,影像追蹤,影像分割,三維輪廓重建,影像對位, ultrasound,image tracking,image segmentation,3D contour reconstruction,image registration, |
出版年 : | 2012 |
學位: | 碩士 |
摘要: | 在這邊論文中提出一種以影像導引系統來追蹤肝臟,傳統的醫學影像導引系統包含斷層掃描、核磁共振和超音波,基於影像頻率與非入侵性的考量,以超音波影像導引系統來追蹤肝臟,其主要目的是以超音波影像來持續地追蹤肝臟運動並且治療肝腫瘤。傳統上的樣板比對既無法描述非鋼體運動又需要龐大的計算量,為了改善準確度與計算速度,額外測試了兩種追蹤方法包含光流法與類神經網路,並以兩組不同的場景進行實驗測試。在第一組場景中,目標物採取正常呼吸,在第二組場景中,目標物則採取變化式地呼吸,急促與緩慢的變化、屏住呼吸或氣喘般的呼氣。在第一組場景中,三種追蹤方法都能成功地追蹤到目標物,然而在第二組實驗中,三種方法都偶爾地會跟丟目標物。
現有上主要三維輪廓重建的方法,都是採取移動探頭來掃描靜態的器官,然而這個想法並不適用在重建動態器官,本篇裡提出了一種廉價且靈活的超音波影像系統,它結合輪廓對位與輪廓分割,從有限個數的二維輪廓來重建三維輪廓。它的概念是固定探頭,藉由呼吸下肝臟運動來採集不同部位的二維影像。為了達到可靠的重建結果,本篇裡發展了新的影像分割與輪廓對位的方法,一種結合紋理差異影像與主動輪廓模型能提供可靠的影像分割結果,接著,再利用已分割的輪廓,提出了一種雙重超音波對位的方法,這想法是利用多出來的探頭來追蹤被採集影像的位置,利用迭代最近的點來追蹤輪廓位置,並將它們對位,這套對位系統是可以從稀疏的二維輪廓來達成精確的三維輪廓。 An image-guided system for a liver tracking is presented in this thesis. Traditional medical image-guided systems include computed tomography (CT), magnetic resonance imaging (MRI) and ultrasound image (US). While considering frame rate and invasiveness, the ultrasound image-guided system is proposed for tracking the liver motion. The main objective is to use ultrasound image for continuously tracking liver motion and for tumor treatment. Traditional template matching cannot describe the non-rigid body motion, and consume a large amount of computational time. To improve the accuracy and computation speed, two tracking methods are tested, including optical flow and neural network. Two different scenarios are experimentally tested. In the first scenario, the “subject” breathes normally. In the second scenario, the “subject” varies between taking deep and slow breathes, holding his breath, or panting rapidly. For the first scenario, all three methods could track the target motion successfully, while, for the second scenario, all methods might lose the target occasionally. The primary methods of existing 3D contour reconstruction scan a static organ with moving probe. However, this idea is not suitable for dynamic organ. A low-cost and flexible ultrasound imaging system which combines contour registration with image segmentation for 3D reconstructions from limited numbers of 2D contours is presented. The proposed approach is based on a fixed ultrasound probe system that collect each partial 2D imaging through the liver motion due to respiration. For reliable reconstruction performance, a new method for image segmentation and contour registration is developed. A new hybrid approach that provides reliable segmentation performance with texture distance image and active contour model is presented. Second, using the segmented contour, a new dual contours registration method is introduced. The approach uses additional probe to track the position of acquisition of images during scanning. Then, the contour registration is performed using contour for iterative closest point (ICP) matching. This registration system allows for accurate 3D reconstructions from sparse 2D image slices. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/64685 |
全文授權: | 有償授權 |
顯示於系所單位: | 電機工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-101-1.pdf 目前未授權公開取用 | 9.22 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。