跳到主要內容

臺灣博碩士論文加值系統

(3.236.84.188) 您好!臺灣時間:2021/07/30 02:32
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:黃上然
研究生(外文):Shang-Ran Huang
論文名稱:利用最大期望值演算法分類內頸動脈狹窄病人腦血流灌注磁振影像之不同組織區域
論文名稱(外文):Using Expectation-Maximization Algorithm Initialized by Hierarchical Clustering on MR Dynamic Images from Patients with Unilateral Internal Carotid Artery Stenosis
指導教授:吳育德
指導教授(外文):Yu-Te Wu
學位類別:碩士
校院名稱:國立陽明大學
系所名稱:生物醫學影像暨放射科學系暨研究所
學門:醫藥衛生學門
學類:醫學技術及檢驗學類
論文種類:學術論文
論文出版年:2009
畢業學年度:97
語文別:英文
論文頁數:81
中文關鍵詞:磁振腦血流灌注影像最大期望值演算法動脈輸入函數腦血流動力參數
外文關鍵詞:MR brain perfusion imagesexpectation maximizationarterial input functioncerebral hemodynamics
相關次數:
  • 被引用被引用:0
  • 點閱點閱:249
  • 評分評分:
  • 下載下載:50
  • 收藏至我的研究室書目清單書目收藏:0
我們使用由階層式分群法起始化的最大期望值演算法(HC-EM-MoMG)來對單側頸動脈狹窄病人的磁振腦血流灌注影像做影像分割,依據不同組織形態的血流灌注特徵,將它們分到不同的群組。我們求得正常側與不正常側的兩類動脈區域,利用其隨時間變化濃度曲線為動脈輸入函數(AIF),各自計算大腦兩側的血流動力參數,尤其是計算腦血流灌注曲線有延遲及發散效應組織群組之血流動力參數,以減少延遲及發散效應所造成的誤差。在病人血流灌注的術前與術後的正常灰白質與不正常灰白質之間的不對稱性統計分析方面,我們提出下列三種方法:
第一種方法一是直接比較分割出來的正常灰質(GM)與白質(WM)組織與不正常灰質與白質(WM)組織之間的相對血流動力參數比值,包括相對腦血流容積比值(rCBV ratio)、相對腦血流流量比值(rCBF ratio)、平均穿流時間差值(dMTT)和尖峰時間差值(dTTP)。我們發現對大腦灰質而言,平均穿流時間差值(dMTT) 和尖峰時間差值(dTTP)這兩個參數最能表現病人在放置頸動脈支架手術前後的腦血流動力的改變(p值分別為0.040和0.008)。
第二種方法是再進一步定義出大腦兩側於前大腦動脈(ACA)、中大腦動脈(MCA)和後大腦動脈(PCA)三個流域,然後比較分割出來的正常灰質(GM)與白質(WM)組織在這三個流域正常側體素數量與頸動脈狹窄側體素數量的比值、相對腦血流容積比值(rCBV ratio)、相對腦血流流量比值(rCBF ratio)、平均穿流時間差值 (dMTT) 和尖峰時間差值(dTTP)等。結果顯示GM位於MCA流域的dTTP值和位於PCA流域的相對腦血流體積比值 (rCBV ratio) 在支架放置手術後有明顯的變化 (p值為0.002和0.027)。 此外,白質(WM)位於ACA和MCA流域的dTTP值和位於PCA流域的rCBV ratio值以及位於ACA和PCA流域的dMTT值於支架放置術後有顯著性的差異(p值依序為0.002、0.002、0.010、0.049和0.049 )。
第三種方法則是在大腦兩側於前大腦動脈(ACA)、中大腦動脈(MCA)和後大腦動脈(PCA)三個流域圈選局部感興趣正常區域(ROI)後分析該區域內灰質與鏡像不正常ROI內灰質之相對腦血流容積比值(rCBV ratio)、相對腦血流流量比值(rCBF ratio)、平均穿流時間差值 (dMTT) 和尖峰時間差值(dTTP)等。結果顯示無論是利用ROI內單純GM腦血流動力參數值或是利用整個ROI內混合了GM與其他組織的參數值,其dTTP在ACA和MCA流域於支架放置前後都有統計上的顯著差異。(對使用GM數值在來說,ACA和MCA的dTTP值於術後有顯著改變,p值是0.014和0.010,對使用混合在一起的參數而言,同樣地,dTTP值在ACA和MCA流域術前、術後差異的p值為0.002和0.002)。縱使混合參數的使用並不影響到統計檢定的結果,但確實會造成相對腦血流量與相對腦血流體積大約百分之十的低估。
除了對病人的腦血流灌注影像做分析之外,我們另外設計了蒙地卡羅模擬實驗來驗證HC-EM-MoMG在不同程度的延遲與發散效應下、受延遲與發散效應影響的不正常組織(腦血灌注延遲的動脈、灰質和白質)在擁有不同體素比率的狀況下與不同的訊雜比 (SNR) 的情況下,進行組織分群的正確率。實驗結果指出,在多數的情況下,組織分群的總體正確率都能達到85%以上。
Expectation-maximization (EM) algorithm with mixture of multivariate Gaussians (MoMG) initialized by hierarchical clustering (HC) was applied on dynamic susceptibility contrast (DSC) MR images from the patients with unilateral internal carotid artery stenosis (ICAs) to segment out different brain tissue clusters depending on distinct blood supply patterns. The segmented normal and delayed arterial compartments were used as arterial input functions, denoted by AIF and dAIF, to compute the hemodynamic parameters in the normal and stenotic sides, respectively, for the subsequent analysis. The use of delayed AIF provided the advantage to alleviate the underestimated relative cerebral blood flow (rCBF) caused by the delayed and dispersed concentration-time curves of morbid tissues. Based on the segmented results, we proposed three comparative approaches to assess the hemodynamic change of gray matter (GM) and white matter (WM) before and after stenting therapy. The first approach, referred to as component comparison, is to compare the ratio of relative cerebral blood volume (rCBV ratio), ratio of relative cerebral blood volume (rCBF ratio), difference in mean transit time (dMTT), and difference in time to peak (dTTP) between the delayed GM (dGM) and normal GM (nGM) compartments, and between the delayed WM (dWM) and normal WM (nWM) compartments. Results demonstrated that difference in mean transit time (dMTT) and difference in time to peak (dTTP) between the dGM and nGM components can robustly reveal the hemodynamic change (p = 0.040 and 0.008, respectively). In the second approach, referred to as regional asymmetry comparison, we computed the asymmetry ratio of rCBV, asymmetry ratio of rCBF, dTTP and dMTT derived from the GM (or WM) in the anterior cerebral artery (ACA), middle cerebral artery (MCA) and posterior cerebral artery (PCA) territories of stenotic side relative to the GM (or WM) in the mirrored ACA, MCA and PCA territories of normal side. Results showed that GM dTTP in MCA territory (p = 0.002) and GM rCBV ratio in PCA territory (p = 0.027) presented significant change after the placement of stent in the stenotic artery. Moreover, WM dTTP in ACA and MCA areas, WM rCBV ratio in PCA areas and WM dMTT in ACA and PCA areas changed significantly (p = 0.002, 0.002, 0.010, 0.049 and 0.049 respectively). The third approach was ROI-based asymmetry comparison within bilateral ACA, MCA and PCA regions using the parameters rCBV ratio, rCBF ratio, dTTP and dMTT as defined in the previous approach. Results demonstrated significant difference in GM dTTP values within ACA and MCA territories (p = 0.014 and 0.010). In conclusion, this study proposed the HC-EM-MOMG segmentation method on which the component comparison, regional asymmetry comparison and ROI-based asymmetry comparison are promising assessment tools in assistance to diagnosis and therapeutic assessment for cerebrovascular diseases.
In addition to the analysis of MR perfusion images from patients with unilateral ICAs, we also performed Monte Carlo simulations, in which different SNR levels, different delays and dispersion effects, and different ratios of number of voxels of abnormal tissue components (delayed artery, delayed gray matter and delayed white matter) to the number of voxels of corresponding normal tissue component were devised, to assess the performance of the HC-EM-MoMG algorithm. Result showed that the classification rates reached 85% in most conditions.
中文摘要 I
ABSTRACT III
CONTENTS V
LIST OF FIGURES VII
LIST OF ACRONYMS XIV
1. INTRODUCTION - 1 -
1.1. MOTIVATION - 1 -
1.2. RELATED RESEARCHES - 2 -
1.3. CONTRIBUTIONS - 4 -
2. MATIRIALS AND METHODS - 5 -
2.1. DATA ACQUISITION - 5 -
2.2. IMAGE PREPROCESSING AND IMAGE SEGMENTATION USING HC-EM-MOMG - 7 -
2.3. DEFINITIONS OF RCBV RATIO, RCBF RATIO, DMTT AND DTTP USED IN THE COMPONENT COMPARISON METHOD - 8 -
2.4. THREE MAJOR CEREBRAL VASCULAR TERRITORIES AND SPATIAL ASYMMETRY RATIOS - 9 -
2.5. REGIONAL ASYMMETRY COMPARISON AND ROI-BASED ASYMMETRY COMPARISON IN THE ACA, MCA AND PCA TERRITORIES - 10 -
3. MONTE CARLO SIMULATION - 14 -
3.1. PARAMETERS FROM NORMAL SUBJECT - 14 -
3.2. SIMULATED AIF, NORMAL AND ABNORMAL TISSUE CONCENTRATION TIME CURVES - 14 -
3.3. SIMULATED DATA SET - 18 -
3.4. RESULTS OF MONTE CARLO SIMULATION - 20 -
4. RESULTS - 26 -
4.1. SEGMENTATION RESULTS - 26 -
4.2. RESULTS OF COMPONENT COMPARISON - 29 -
4.3. RESULTS OF REGIONAL ASYMMETRY COMPARISON AND ROI-BASED ASYMMETRY COMPARISON - 30 -
4.4. RESULTS OF SPATIAL ASYMMETRY IN THE THREE CEREBROVASCULAR TERRITORIES - 31 -
5. DISCUSSION - 35 -
6. CONCLUSION - 37 -
REFERENCES - 38 -
APPENDIX I - 43 -
APPENDIX Ⅱ - 44 -
Aronen HJ, Glass J, Pardo FS, Belliveau JW, Gruber ML, Buchbinder, BR, Gazit IE, Linggood RM, Fischman AJ, Rosen BR, 1995. Echo-planar MR cerebral blood volume mapping of gliomas. Clinical utility. Acta Radiologica 36(5):520-528.

Axel L, 1980. Cerebral blood flow determination by rapid-sequence computed tomography. Radiology 137(3):679-686.

Barber, D.C., 1980. The use of principal components in the quantitative analysis of gamma camera dynamic studies. Physics in Medicine and Biology 25(2):283-292.

Boxerman JL, Rosen BR, Weisskoff RM, 1997. Signal-to-noise analysis of cerebral blood volume maps from dynamic NMR imaging studies. Journal of Magnetic Resonance Imaging 7(3):528-537.

Calamante F, Gadian DG, Connelly A, 2000. Delay and dispersion effects in dynamic susceptibility contrast MRI: simulations using singular value decomposition. Magnetic Resonance in Medicine 44(3):466-473.

Calamnate F, Gadian DG, Connelly A, 2002. Quantification of perfusion using bolus tracking magnetic resonance imaging in stroke. Assumptions, limitations and potential implications for clinical use. Stroke 33(4):1146-1151.

Calamante F, Morup M, Hansen LK, 2004. Defining a local arterial input function for perfusion MRI using independent component analysis. Magnetic Resonance in Medicine 52(4):789-797.

Calamante F, 2005. Bolus dispersion issues related to the quantification of perfusion MRI data. Journal of Magnetic Resonance Imaging 22(6):718-722.

Carroll TJ, Haughton VM, Rowley HA, Cordes D, 2002. Confounding effect of large vessels on MR perfusion images analyzed with independent component analysis. American Journal of Neuroradiology 23(6):1007-1012.

Chou YC, 2007. Tissue Classification from Brain Perfusion MR Images Using Expectation-Maximization Algorithm Initialized by Hierarchical Clustering on Whitened Data. Doctoral Dissertation. Department of Biomedical Imaging and Radiological Sciences, National Yang-Ming University, Taipei, Taiwan.

Damasio H, 1983. A computed tomographic guide to the identification of cerebral vascular territories. Archives of Neurology 40(3):138-142.

Di Paola, R., Bazin, J.P., Aubry, F., Aurengo, A., Cavailloles, F., Herry, J.Y., Kahn, E., 1982. Handling of dynamic sequences in nuclear medicine. IEEE Transaction on Nuclear Science 29 (4):1310-1321.

Gauvrit JY, Delmaire C, Henon H, Koussa MA, Leys D, Pruvo JP, Leclerc X, 2004. Diffusion /perfusion-weighted magnetic resonance imaging after carotid angioplasty and stenting. Journal of Neurology. 251(9):1060-1067.

Gückel FJ, Brix G, Schmiedek P, Piepgras Z, Becker G, Kopke J, Gross H, Georgi M, 1996. Cerebrovascular reserve capacity in patients with occlusive cerebrovascular disease: assessment with dynamic susceptibility contrast-enhanced MR imaging and the acetazolamide stimulation test. Radiology 201 (2), 405-412.

Houston AS, 1984. The effect of apex-finding errors on factor images obtained from factor analysis and oblique transformation. Physics in Medicine and Biology 29(9):1109-1116.

Kao YH, Guo WY, Wu YT, Liu KC, Chai WY, Lin CY, Hwang YS, Liou AJK, Wu HM, Cheng HC, Yeh TC, Hsieh JC, Teng MMH, 2003. Hemodynamic segmentation of MR brain perfusion images using independent component analysis, thresholding, and Bayesian estimation. Magnetic Resonance in Medicine. 49 (5):885-894.

Kao YH, Teng MMH, Liu KC, Lam IP, Lin YC, 2008. Hemodynamic segmentation of MR perfusion images in patients with unilateral carotid stenosis using independent component analysis. Journal of Magnetic Resonance Imaging 28(4):1125-1132.

Kiselev, VG, 2001. On the theoretical basis of perfusion measurements by dynamic susceptibility contrast MRI. Magnetic Resonance in Medicine 46(6):1113-1122.
Knutsson L, Larsson EM, Thilmann O, Stahlberg F, Wirestam R, 2006. Calculation of cerebral perfusion parameters using regional arterial input functions identified by factor analysis, 2006. Journal of Magnetic Resonance Imaging 23(4):444-453.

Lorenz C, Benner T, Lopez CJ, Ay H, Zhu MW, Aronen H, Karonen J, Liu Y, Nuutien J, Sorensen AG, 2006. Effect of using local arterial input functions on cerebral blood flow estimation. Journal of Magnetic Resonance Imaging 24(1):57-65.

Mai JK, Assheuer J, Paxinos G, 1997. Atlas of the human brain. Academic Press, 33-35.

Martel AL, Moody AR, Allder SJ, Delay GS, Morgan PS, 2001. Extracting parametric images from dynamic contrast-enhanced MRI studies of the brain using factor analysis. Medical Image Analysis 5(1):29-39.

Moftakhar R, Turk AS, Niemann DB, Hussain S, Rajpal S, Cook T, Geraghty M, Kienitz BA, Turski PA, Newman GC, 2005. Effects of Carotid or Vertebrobasilar Stent Placement on Cerebral Perfusion and Cognition. American Journal of Neuroradiology 26(7):1772-1780.

Nolte J, 2002. The human brain-an introduction to its functional anatomy. 5th edition.

Ostergaard L, R.M. Weisskoff, D.A. Chesler et al, 1996a. High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. Part I: Mathematical approach and statistical analysis, Magnetic Resonance in Medicine 36(5):715-725.

Ostergaard L, Sorensen AG, Kwong KK, Weisskoff, RM, Gyldensted C, Rosen BR, 1996b. High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. Part II: Experimental comparison and preliminary results. Magn. Resonance in Medicine. 36 (5), 726-736.

Otsu N, 1979. A threshold selection method from gray-level histograms. IEEE Transaction on Systems, Man and Cybernetics. 9(1):62-66

Rempp KA, Brix G, Wenz F, Becker CR, Guckel F, Lorenz WJ, 1994. Quantification
of regional cerebral blood flow and volume with dynamic susceptibility contrast-enhanced MR imaging. Radiology 193(3):637–641.

Rosen BR, Belliveau JW, Vevea JM, Brady TJ, 1990. Perfusion imaging with NMR contrast agents. Magnetic Resonance Medicine 14(2):249-265.

Rogowska J, Preston JK, Aronen HJ, Wolf GL, 1994. A comparative analysis of similarity mapping and eigenimaging as applied to dynamic MR imaging of a low grade astrocytoma. Acta Radiol 35(4):371-377.

Sorensen AG, Tievsky AL, Østergaard L, Weisskoff RM, Rosen BR, 1997. Contrast agents in functional MR imaging. Journal of Magnetic Resonance Imaging 7(1): 47-55.

Sorensen AG, Copen WA, Ostergaard L, Buonanno FS, Gonzales RG, Rordorf G, Rosen BR, Schwamm LH, Weisskoff RM, Koroshetz WJ, 1999. Hyperacute stroke: simultaneous measurement of relative cerebral blood volume, relative cerebral blood flow, and mean tissue transit time. Radiology 210(2):519-527.

Starmer CF, Clark DO. Computer computations of cardiac output using the gamma function. Journal of Applied Physiology 1970(2);28:219-220.

Sobesky J, Weber OZ, Lehnhardt FG, Hesselmann V, Neveling M, Jacobs A, Heiss WD, 2005. Does the mismatch match the penumbra? Magnetic resonance imaging and positron emission tomography in early ischemic stroke. Stroke 36(5):980-985.

Takasawa M, Jones PS, Guadagno JV, Christensen S, Fryer TD, Harding S, Gillard JH, Williams GB, Aigbirhio FI, Warburton EA, Ostergaard L, Baron JC, 2008. How reliable is perfusion MR in acute stroke? Validation and determination of the penumbra threshold against quantitative PET. Stroke 39(3):870-877.

Waaijer A, Leeuwen MSV, Osch MJPV, Worp BHVD, Moll FL, Lo RTH, Mali WPTM, Prokop M, 2007a. Changes in cerebral perfusion after revascularization of symptomatic carotid artery stenosis:CT measurement. Radiology 245(2):541-548.

Waaijer A, van der Schaaf IC, Velthuis BK, Quist M, van Osch MJP, Vonken EPA, van Leeuwen MS, Prokop M, 2007. Reproducibility of quantitative CT brain perfusion measurements in patients with symptomatic unilateral carotid artery stenosis. American Journal of Neuroradiology 28(5):927-932.

Willats L, Connelly A, Calamante F, 2006. Improved deconvolution of perfusion MRI data in the presence of bolus delay and dispersion. Magnetic Resonance in Medicine 56(1):146-156.

Willats L, Connelly A, Calamante F, 2008. Minimising the effects of bolus dispersion in bolus-tracking MRI. NMR in Biomedicine 21(10):1126-1137.

Wiart M, Rognin N, Berthezene Y, Nighoghossian N, Froment JC, Baskurt A, 2001. Perfusion-based segmentation of the human brain using similarity mapping. Magnetic Resonance in Medicine 45 (2):261-268.

Wishart D., 1969. An algorithm for hierarchical classifications. Biometrics 25(1):165-170.

Wismuller A, Meyer-Baese A, Lange O, Reiser MF, Leinsinger G, 2006. Cluster analysis of dynamic cerebral contrast-enhanced perfusion MRI time-series. IEEE Transactions on Medical Imaging. 25(1):62-73.

Wong KK, Tam CP, Ng M.,Wong S TC,Young GS, 2009. Improved residue function and reduced flow dependence in MR perfusion using least-absolute-deviation regularization. Magnetic Resonance in Medicine 61(2):418-428.

Wu O, Ostergaard L, Weisskoff RM, Benner T, Rosen BR, Sorensen AG, 2003a. Tracer arrival timing-insensitive technique for estimating flow in MR perfusion-weighted imaging using singular value decomposition with a block-circulant deconvolution matrix. Magnetic Resonance in Medicine 50(1):164-174.

Wu O, Ostergaard L, Koroshetz WJ,Schwamm LH,O’Donnell J,Schaefer PW, Rosen BR,Weisskoff RM, Sorensen AG, 2003b. Effects of tracer arrival time on flow estimates in MR perfusion-weighted imaging. Magnetic Resonance in Medicine 50(4):856-864.

Wu YT, Chou YC, Guo WY, Yeh TC, Hsieh JC, 2007. Classification of spatiotemporal hemodynamics from brain perfusion MR images using expectation-maximization estimation with finite mixture of multivariate gaussian distributions. Magnetic Resonance in Medicine 57(1):181-191.

Wu YT, Chou YC, Lu CF, Huang SR, WY G, 2008. Tissue classification from brain perfusion MR images using expectation-maximization algorithm initialized by hierarchical clustering on whitened data, 13th ICBME, Singapore, 714-717.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top