與其它醫(yī)學成像技術(shù)相比,超聲成像具有安全性高、時間分辨率好等重要優(yōu)勢。由超聲圖像引導的介入式手術(shù)導航技術(shù)正日益成為臨床醫(yī)學的重要技術(shù)手段之一。在介入式手術(shù)過程中,超聲導航技術(shù)跟蹤超聲圖像探頭的坐標位置,用于引導手術(shù)器械使之快速準確地達到目標位置。研究并實現(xiàn)基于雙目視覺的超聲引導介入是手術(shù)導航系統(tǒng),需要面對三個主要技術(shù)挑戰(zhàn):第一,要保證導航系統(tǒng)的高度精確性,并能夠準確地將身體內(nèi)部狀態(tài)與變化展現(xiàn)給屏幕旁的外科醫(yī)生;第二,要在模糊的和固有斑點噪聲的醫(yī)學超聲圖像中準確分割出腫瘤部位;最后,要結(jié)合利用數(shù)據(jù)實現(xiàn)導航過程的可視化,對手術(shù)醫(yī)生實時顯示合適的信息。為了解決這些問題,本文進行了以下研究工作:首先,開發(fā)了一種手術(shù)器械尖端的標定方法。針對傳統(tǒng)的基于最小二乘法(LS)標定方法對異常值敏感問題,提出了一種基于遞推最大熵(RMC)準則的自適應標定方法。分析和實驗表明,該方法對異常值具有較好魯棒性,且標定精度較高。其次,通過結(jié)合局部相關(guān)熵和全局相關(guān)熵,提出了一種基于全局和局部相關(guān)熵的k均值(GLCK)圖像分割算法。局部相關(guān)熵力在目標邊界附近起主導作用,用來吸引水平集函數(shù)曲線達到目標邊界;而全局相關(guān)熵力在遠離目標邊界時起主導作用。結(jié)果表明,GLCK模型提高了斑點噪聲和邊界模糊的圖像在分割上的魯棒性和精度,且計算時間也明顯減少。最后,提出了一個綜合的多模態(tài)可視化場景。我們設計了可以從任何位置觀看的虛擬3D導航場景,包括跟蹤的手術(shù)器械和實時2D超聲圖像平面的偽影表示。
【學位單位】：大連理工大學
【學位級別】：碩士
【學位年份】：2018
【中圖分類】：R319
【文章目錄】：
Abstract
摘要
1 Introduction
    1.1 Research Background and Significance
    1.2 The Key Technologies in Ultrasound-Based Surgery Navigation
        1.2.1 Ultrasound Imaging
        1.2.2 Tracking Technologies
        1.2.3 Calibration of the Ultrasound Probe and Surgery Instrument
        1.2.4 Registration and Ultrasound Segmentation
        1.2.5 Visualization in Ultrasound-based Surgery Navigation
    1.3 Research Progress
        1.3.1 Literature Review
        1.3.2 The Challenge of Ultrasound-based Surgical Navigation
    1.4 Outline
2 Workflow and Application of Ultrasound-based Surgery Navigation
    2.1 Workflow of Ultrasound-based Surgery Navigation
    2.2 Applications of Ultrasound-Based Surgical Navigation
        2.2.1 Ultrasound-based Surgery Navigation in Neurosurgery
        2.2.2 Ultrasound-based Surgery Navigation in Laparoscopic
    2.3 Chapter Summary
3 The Principle and Method of the Surgical Instrument’s Tip and Ultrasound Probe Calibration
    3.1 Recursive Maximum Correntropy based Calibration of the Surgical Instrument’s Tip
        3.1.1 Principle Method of Surgical Instrument Tip Calibration
        3.1.2 Recursive Maximum Correntropy Theory
        3.1.3 Surgical Instrument Tip Calibration using the Recursive Maximum Correntropy Algorithm
        3.1.4 Experimental Setup
        3.1.5 Experimental Results
        3.1.6 Conclusion and Discussion
    3.2 Ultrasound Probe Calibration
        3.2.1 Principle Method of Ultrasound Probe Calibration
        3.2.2 Experimental Process and Results
        3.2.3 Conclusion and Discussion
    3.3 Chapter Summary
4 The Principle and Method of Tumor Segmentation in the Ultrasound Image
    4.1 Image Segmentation Background
        4.1.1 CV Method
        4.1.2 LBF Method
    4.2 Active Contour Driven by Global and Local Correntropy Energies （GLCK） for Segmentation Ultrasound Image
        4.2.1 GLCK Methodt
        4.2.2 GLCK Algorithm
        4.2.3 Experimental Results
        4.2.4 Discussion and Conclusion
    4.3 Chapter Summary
5 Assembly of Ultrasound-based Navigation Systems
    5.1 System Material
    5.2 System Calibration
    5.3 System Visualization
        5.3.1 Main Part of the Program
        5.3.2 Experimental Results
    5.4 Chapter Summary
Conclusions
References
Research Projects and Publications in Master Study
Acknowledgement

【參考文獻】

相關(guān)期刊論文 前2條

1 羅楊宇;徐靜;魯通;陳懇;李成榮;;基于磁定位器的手動三維超聲圖像標定[J];中國生物醫(yī)學工程學報;2008年02期

2 劉潔,高上凱,李錚;三維Freehand超聲血流測量系統(tǒng)中的單平面定標方法[J];北京生物醫(yī)學工程;2001年03期

相關(guān)博士學位論文 前1條

1 馬文娟;紅外手術(shù)導航儀關(guān)鍵技術(shù)研究[D];上海交通大學;2010年

本文編號：2858624