超音波動態影像分析技術與臨床醫學應用

Dynamic Analyses of Ultrasound Imaging and Their Clinical Applications

包舜華(Sun-Hua Pao)

指導教授 ： 邵耀華

國立臺灣大學/工學院/應用力學研究所/博士(2004年)

https://doi.org/10.6342/NTU.2004.02307

摘要

超音波動態影像提供一種即時方便的醫療檢測工具。 本文主要目的在結合超音波動態影像傳輸、生物力學與視窗化軟體設計，以呈現便捷與客觀的臨床檢測應用。經由不同超音波影像原理良莠互補的情況，進行血管硬度（E）、能量消散比例（EDR）分析，並與脈波速度（PWV）進行比較。為了減少血管邊緣影像擷取的人為誤差，本文採用自動化深度補償，修正的邊緣偵測。透過自動化的新技術結合M-mode影像、序列B-mode影像的多軸向檢測，可提供客觀的血管機械特性偵測。在相同的硬體架構下，發展單點測量壓力之PWV理論，利用 Pulsed Doppler的影像亦可估測局部區域的PWV與血流量，作為血管硬度的重複驗證。最後利用相同的分析技術進行更細微的Intima –Media 厚度（IMT）變化量測，以及提出非侵襲性 Intima -Media與Adventitia硬度比值測量的力學架構。
此外，在超音波Doppler模式下，我們發展 Power Doppler對於腫瘤血液循環的評估指標（DFVI）。藉由甲狀腺良性與惡性腫瘤的雛形，證實該指標相較於傳統使用的Vascular Index（VI）有明顯的改善。在43位病患（16為良性腫瘤、27位惡性腫瘤裡），惡性腫瘤的 VI數值明顯高於良性腫瘤，但是統計上卻無明顯差異（P=0.055）。 本文透過動態影像處理後的DFVI具有良好的檢測能力（P=0.004 Sensitivity 84% Specificity 64%），其中主要的關鍵在於DFVI能夠凸顯較末端微循環特性。另外 Power Doppler的功率強度在整個實驗的比較結果，雖然並不能證明與血流量、壓力波有直接的關連，但是惡性腫瘤的功率強度的變異量（PW Var）均明顯高於良性腫瘤（P=0.014 Sensitivity=80% Specificity 64%），不過惡性腫瘤卻與正常組織無法區分。值得一提的是，PW Var是與腫瘤的框選面積（ROI）無關的指標，僅與血管本身的脈動特性有關，未來可以提供ROI難以確定的腫瘤檢測的另一種指標。

關鍵字

血管指數 ； 脈波速度 ； 血管彈性 ； 血管密度

並列摘要

Dynamic imaging of ultrasonography provides a convenient real-time diagnostic tool for clinical research. The objective of this study is to develop the interactive image analysis software for clinical biomechanics of blood vessels and tumor vascularity. Automatic edge-detection techniques of vascular walls and the depth compensation of ultrasound image have been implemented. The M-mode images have been used to evaluate vascular energy dissipation ratio (EDR) with a higher spatial resolution. Furthermore, sequential B-mode images offer the vascular stiffness in different radial direction. In addition, the single-point pulsed Doppler has been shown to be capable of measuring the pulsed wave velocity (PWV) locally. We also developed a non-invasive technique detect the change of intima-media thickness (IMT) with the lumen pressure, and the relative stiffness between the intima-media and the adventitia could be calculated.

Using power Doppler a novel tumor differentiation parameter, DFVI (Differential Vascular Index), has been proposed to quantify the blood circulation in tumors. DFVI is affirmed to be more efficient in the classification of malignant tumor (thyroid carcinoma) from the benign tumor than the traditional VI (Vascular Index). Among 43 patients (16 benignancy and 27 malignancy), the VI of malignant thyroid tumor is higher than those of benignancy but with only a minor statistical significance (p=0.055). On the other hand, the difference between the benignancy and the malignancy is successfully identified by DFVI（p=0.004, sensitivity 84%, specificity 64%） because of the protruded microcirculation among thyroid tumor tissue. In this study, although the quantity of the scatter power can not reflect the content of blood flow, the scattered power variance (PW Var) is statistically significant between benignancy and malignancy （p=0.014 , sensitivity=80% , specificity 64%）. However, the PW Var does not have sufficient statistical significance in differentiate the normal tissue perfusion from the malignant one. Because the PW var is independent from region of interest (ROI), it may have the potential to investigate those tumors where the ROI is ambiguous.

並列關鍵字

Vascular Index ； MVD ； Vascular elasticity ； pulsed wave velocity

參考文獻

[1] Adler, R.S.,Rubin, J.M.,Fowlkes, J.B.: Ultrasound Estimation of Tissue Perfusion: A Stochastic Approach, Ultrasound Med. & Biol. 1995;21(4): 493-500.

[2] Arbelle, P.,Berson, M.,Achaibou, F.: Vascular Resistence Quantification In High Flow Resistence Areas Using The Doppler Method , Ultrasound Med. & Biol.1995;21(3):321-328.

[3] Benthin M., Dahl P., Ruzicka R., Lindstro¨m K.: Calculation of Pulsewave Velocity Using Cross Correlations—Effects Of Reflexes in The Arterial Tree. Ultrasound Med & Biol. 1991;17:461–469.

[4] Bosari, S.,Lee, A.K.C.,Delellis, R.A.: Microvessel Quantification and Prognosis in Invasive Breast Carcinoma, Human Pathology, 1992;23(7):755-761.

[5] Brown, J.M.Quedens-Case, C.,Alderman, J.L..: Contrast-Enhanced Sonography Of Tumor Neovascularity in A Rabbit Model, Ultrasound Med. & Biol. 1998;24(4):495-501.