醫用超音波影像因為斑點雜訊(speckle)而在低對比的腫瘤診斷上受到限制，傳統上常用利用空間統計特性來改變可適性濾波器的行為，以針對影像上不同的區域並進行不同的濾波，但實際上不同結構的區域也可能會得到相同的統計結果而造成誤判的可能。本文提出兩個影像處理上基於明亮差異值(brightness difference)之可適應性超音波影像斑點雜訊抑制的濾波器。方法一與傳統可適性濾波器架構相仿但利用最大明亮差異值(MBD)取代統計參數以控制濾波特性，在有特徵訊號的區域，有較高的明亮差異值，此時影像平滑化降低以維持特徵；而在斑點雜訊區，會有較低的明亮差異值，此時平滑化升高其達到濾除斑點雜訊的效果。而方法二是在沿著切割角度之遮罩(mask)分割下，將各方向上之明亮差異值所佔的比重結合該角度遮罩上的中間值來進行濾波。當影像上各切割角度間所佔的比重相似時可視為同性質區，會有類似平均濾波器的效果產生以達到最大雜訊抑制；而當某一角度上的比重較大時，也就是當遮罩含蓋可解析的輪廓(contour)時則沿著該結構做處理以呈現其輪廓。實驗結果証明，無論是在模擬圖或是在實際超音波影像上，本文所提的兩個方法相較於使用空間統計的傳統濾波器，不但有較佳運算效率也有較好的濾波效果。

Ultrasound imaging has become widely utilized for clinical diagnoses. Nevertheless, detection of low-contrast object in ultrasound images is significantly limited by inherent speckle artifacts. For speckle suppression using post-processing filtering, in this paper, we proposed two novel adaptive filters based on directional brightness differences (BD). The adaptive weighted median filter (AWMF) relies on statistic features of local image brightness. Though the spatial characteristics may significantly differ, a resolvable object could be erroneously blurred when it is statistically similar to speckle. The method 1 for median weighting is proposed to better separate resolvable objects from speckle background by the maximal brightness difference (MBD) of directional kernels. Since resolvable objects usually have distinct spatial orientation, a large brightness difference is expected among directional kernels with the same orientation. For speckles, the random fluctuation of brightness would result in low brightness difference for all directions. The method 2 of the median value in each direction is weighted by the BD of that angle. For a homogeneous region, the BD is similar in all directions and the median values are equally weighted for maximal smoothing. On the other hand, a large BD is detected in one specific angle when the mask covers a resolvable contour. The filter preserves the contour by giving the median value along that direction a larger weighting. The novel filters were examined using simulated and in-vivo ultrasound images. Results show that they are superior to the AWMF filter in computational efficiency and detail preserving with similar speckle suppression.

[1] 沈哲州，”醫用超音波影像上課講義”，國立台灣科技大學電機所，民國96年。
[2] 李嘉明、李玉華，”新超音波醫學-(1)醫用超音波的基礎”，合記圖書出版社，民國95年。
[3] 沈哲州，”超音波組織非線性影像分析”，國立台灣大學，碩士論文，民國89年。
[4] 李維寧，”高效能斑點追踨技術及其在乳房超音波影像之應用”，國立台灣大學，碩士論文，民國92年。
[5] 吳積霖，”超音波應變複合影像”，國立台灣大學，碩士論文，民國九十年。
[6] H.-C. Huang, J.-Y. Chen, S.-D. Wang, and C.-M. Chen, “Adaptive ultrasonic speckle reduction based on the slope-facet model,” Ultrasound Med. Biol., vol. 29, no.8, pp. 1161-1175, 2003.
[7] G. E. Trahey, S. W. Simith, and O. T. von Ramm, “Speckle pattern correlation with lateral aperture translation: Experimental results and implications for spatial compounding,” IEEE Trans. Ultrason., Ferroelect. Freq. Contr., vol. Uffc-33, no. 3, pp. 257-264, May 1986.
[8] D. Adam, S. Beilin-Nissan, Z. Firedman, V. Behar, “The combined effect of spatial compounding and nonlinear filtering on the speckle reduction in ultrasound images,” Ultrasonics, vol. 44, pp. 166-181, 2006.
[9] G. E. Trahey, J. W. Allison, S. W. Smith, and O. T. von Ramm. “A quantitative approach to speckle reduction via frequency compounding,” Ultrason. Imag., vol. 8, no. 3, pp. 151-164, 1986.
[10] T. Loupas, W. N. McDicken, and P. L. Allan, “An adaptive weighted median filter for speckle suppression in medical ultrasonic images,” IEEE Trans. Circuits Syst., vol. 36, no. 1, pp. 129-135, Jan. 1989.
[11] Ping Yang and Otman A.Basir, “Adaptive weighted median filter using local entropy for ultrasonic image de-noising,” in Proc. of 3rd Int. Symp. on Image and Signal Process. and Anal., vol. 2, pp. 799-803, Sept. 2003.
[12] J. I. Koo and S. B. Park, “Speckle reduction with edge preservation in medical ultrasonic images using a homogeneous region growing mean filter (HRGMF),” Ultrason. Imag., vol. 13, no. 3, pp. 211-237, 1991.
[13] Y. Chen, R. M. Yin, P. Flynn, and S. Broschat, “Aggressive region growing for speckle reduction in ultrasound images”, Pattern Recognition Lett., vol. 24, no. 4-5, pp. 677-691, 2003.
[14] P. Perona and J. Malik, ”Scale-space and edge detection using anisotropic diffusion,” IEEE Trans. Pattern Anal. Machine Intell., vol. 12, no. 7, pp. 629-639, Jul. 1990.
[15] Y. Yu and S. T. Acton, “Speckle reducing anisotropic diffusion,” IEEE Trans. Image Process., vol. 11, no. 11, pp. 1260-1270, Nov. 2002.
[16] K. Z. Abd-Elmoniem, A. M. Youssef, and Y. M. Kadah, “Real-time speckle reduction and coherence enhancement in ultrasound imaging via nonlinear anisotropic diffusion,” IEEE Trans. Biomed. Eng., vol. 49, no. 9, pp. 997-1014, Sep. 2002.
[17] C.-Y. Xiao, S. Zhang, and Y.-Z. Chen, “A diffusion stick method for speckle suppression in ultrasonic images,” Pattern Recognition Lett., vol. 25, no. 16, pp. 1867-1877, Dec. 2004.
[18] Q. Sun, J. A. Hossack, J. Tang, and S. T. Acton, “Speckle reducing anisotropic diffusion for 3D ultrasound images,” Comput. Med. Imag. Grap., vol. 28, no. 8 pp. 461-470, 2004.
[19] R. G. Dantas and E. T. Costa, “Ultrasound speckle reduction using modified Gabor filters,” IEEE Trans. Ultrason., Ferroelect., Freq. Contr., vol. 54, no. 3, pp. 530-538, Mar. 2007.
[20] D. L. Donoho, “De-noising by soft-thresholding,” IEEE Trans. Inf. Theory, vol. 41, no. 3, pp. 613-627, May 1995.
[21] A. Thakur and R. S. Anand, “Image quality based comparative evaluation of wavelet filters in ultrasound speckle reduction,” Digital Singnal Process., vol. 15, no. 5, pp. 455-465, 2005.
[22] F. Zhang, Y. M. Yoo, L. M. Koh, and Y. Kim, “Nonlinear Diffusion in Laplacian Pyramid Domain for Ultrasonic Speckle Reduction,” IEEE Trans. Med. Imag., vol. 26, no. 2, pp. 200-211, Feb. 2007.
[23] J. Meunier, and M. Bertrand, “Ultrasonic Texture Motion Analysis: Theory and Simulation,” IEEE Trans. Med. Image., vol. 14, no. 2, pp. 293-300, June 1995.
[24] F. Sattar, L. Floreby, G. Salomonsson, and B. Lovstrom, “Image Enhancement Based on a Nonlinear Multiscale Method,” IEEE Trans. Image Process., vol. 6, no. 6, pp. 888-895, June 1997.
[25] C. Xu and J. L. Prince, “Sankes, shapes, and gradient vector flow,” IEEE Trans. Image Process., vol. 7, no. 3, pp. 359-369, Mar. 1998.