近幾年來，電腦視覺系統在生醫影像上的應用日與遽增，應用電腦視覺系統在臨床上可以有助於醫師對病症的判斷，而肝臟的病變是醫界相當重視的課題。我們企圖從一張肝臟超音波影像中切割出疑似病變的區域，更進一步的，希望可以從中辨識出正常肝與疑似病變的肝組織，而我們採用的方式是先針對肝臟超音波影像擷取出特徵向量，進而對特徵向量進行分類。
特徵向量的截取方式是架構在空間-頻率的分解以及碎形幾何上，由於在肝臟超音波影像中，最明顯的特徵為粗糙度(roughness)，其中肝癌影像較肝硬化來的粗糙，因此在擷取特徵向量時，運用了碎形維度(fractal dimension)來描述此一特性，一個區域的碎形維度越高表示該區塊的強度變化越劇烈，結合空間-頻率的分解與碎形幾何，我們可以得到一組多重解析度的特徵向量，用以進行後續的分類。而空間-頻率的分解中，我們嘗試了兩種分解方式：(1)M-頻段小波轉換，(2)Gabor轉換，並比較兩種分解方式在非監督式影像切割上的表現與優劣。同時針對特徵向量的長度，我們也作了表現與計算複雜度上的比較，進而挑選出有效的特徵向量，在不影響表現的前提下，可以大幅的壓縮計算複雜度。
在擷取出有效的特徵向量後，我們根據特徵向量對影像進行分類。文獻中有許多的分類方法，其中，K-means與Fuzzy-c-means(FCM)是被廣泛運用的分類機制，在大多數的影像分類上，都有著卓越的表現，而類神經網路領域中的Self Organizing Maps可以描述出每筆資料間在特徵維度上的相對關係，進而被運用於許多的分類應用上。本篇論文中，我們針對兩種頻率分解分式，比較了上述三種分類機制在肝臟超音波多重解析度特徵向量上的表現。另外，我們提出了一種有效的初始化流程，用以改善K-means演算法主要的缺陷，也用以改善其他演算法的效果。
最後，我們提出一個改良式的FCM演算法，它提供了一個彈性的調整機制，該機制建立於機率圖的概念上，此一演算法在自然影像中得到卓越的成效。
我們將實驗延伸至超音波肝臟影像上，實驗結果顯示此一無監督式影像切割流程可以有效地切割出肝臟組織中粗糙的區塊，而我們提出的改良式FCM將可提供醫師在自動分類後可以進行手動的微調，此一彈性的機制，將可融入醫師臨床判定的經驗法則，彌補其他分類機制無法微調的缺陷，結合自動化分類機制與醫師經驗法則微調參數，可以建立一套可定量描述的自動化電腦輔助診斷分析系統。此外，為了提升年輕的臨床醫師對肝臟超音波影像的視覺辨識能力，我們可以利用提出的非監督式變式切割方式建立一套離線(off-line)的學習系統用以探討肝臟超音波影像的視覺辨識準則。

In the recent years, applications of computer vision system on the biological and
medical images are increasing. In clinics, the computer vision system is helpful for
visual interpretation. Among the diseases, the liver disease has received much
attention. We attempt to segment out the region which is likely to be ill from the
ultrasonic liver images. Furthermore, we may identify the normal region and
suspected ill region. At first, we extract the feature vector from the ultrasonic liver
images, and then the feature vector is applied into the clustering algorithm.
The feature extraction algorithm is based on the spatial-frequency decomposition
and fractal geometry. Such a multiresolution feature vector has been proved
trustworthy when handling such texture images as ultrasonic images. In the thesis, we
compare the performance of the two decomposition methods: (1) M-band wavelet
transform and (2) Gabor Transform. Besides, we also analyze the performance and
computational complexity versus different feature lengths. We attempt to reduce the
feature length to improve the execution speed but not to influence the performance
too much as the premise.
After getting efficient feature vector, the clustering algorithm is applied. Many
clustering algorithms have been proposed in the past. K-means and Fuzzy C-means
(FCM) have great performance in most images, and Self Organizing Maps (SOM),
which can describe the structure of feature dimension in the neural network filed, is
also applied to many clustering cases. In the thesis, the performances of the above
three clustering methods combined with the two decomposition methods are
compared. An efficient initialization procedure is proposed to solve the main
shortcoming of K-means algorithm and to improve the performances of other
clustering methods.
At last, a modified FCM method is proposed, which provides an elastic tuning
operation based on the probability maps to adjust the segmentation result. A great
performance is obtained in the cases of nature texture images via the mechanism of
this tuning operation.
The tests are extended to the cases of ultrasonic liver images. The experiments
show that the unsupervised segmentation framework can efficiently segment out the
rough area concerned and the modified FCM we proposed is furthermore expected to
provide doctors a probability tuning operation according to clinical experience.
Instead of hard-clustering, the modified FCM outperforms K-means, FCM and SOM
by its elasticity. A quantitative characterization based on the proposed unsupervised
segmentation algorithm not only can be utilized to establish an automatic
computer-aided diagnostic system. As well, to increase the visual interpretation
capacity of ultrasonic liver image for junior physicians, an off-line learning system
can be developed to investigate the visual criteria.

Reference

[1] R. Momenan, M. H. Loew, M. F. Insana, R. F. Wagner, and B. S. Garra, “Application of pattern recognition techniques in ultrasonic tissue characterization,” 10¬th Int. Conf. Pattern Recognition, vol. 1, pp608-612, 1990
[2] R. Momenan, R. F. Wagner, B. S. Garra, M. H. Loew, and M. F. Insana, “Image staining and differential diagnosis of ultrasound scans based on the Mahalanobis distance,” IEEE Trans. Medical Imaging, vol. 11, pp37-47, June 1994
[3] K. Ogawa, M. Fukushima, K. Kubota, and N. Hisa, “Computer-aided Diagnostic System for Diffuse Liver Diseases with Ultrasonography by Neural Network,” IEEE Trans. Nuclear Science, vol. 45, no. 6, pp 3069-3074, Dec 1998
[4] G.F. Vawter and H. Shwachman, “Cystic fibrosis in adults: an autopsy study,” Pathol. Annu., vol. 14. pp357-382, 1979
[5] N. I. Standford, P. Walsh, C. Matis, H. Baddeley and L. W. Powell, “Is ultrasonography useful in the development of diffuse parenchyma liver disease,” Gastroenterolohy, vol. 9, pp. 186-191, 1985
[6] B. B, Mandelbrot, Fractal Geometry of Nature.” Freeman Press, San Francisco, 1982
[7] Pentland, A. P., “Fractal based description of natural scenes,” IEEE Trans. Pattern Anal. And Machine Intell., vol. PAMI-6, pp.661-674, 1984.
[8] Pentland, A. P. “Shading into texture,” Artificial Intelligence, vol. 29, pp.147-170, 1986

[9] Wen-Li Lee, Yung-Chang Chen, Ying-Cheng Chen and Kai-Sheng Hsieh (2005). Unsupervised Segmentation of Ultrasonic Liver Images by Multiresolution Fractal Feature Vector. Information Sciences (SCI), 175, 177-199.
[10] "A texture descriptor for browsing and similarity retrieval", P. Wu, B.S. Manjinath, S. Newsam, H.D. Shin, Signal Processing: Image Communication 16 (2000) 33-43.
[11] "Texture Features for Browsing and Retrieval of Image Data", P. Wu, B.S. Manjinath, S. Newsam, H.D. Shin, IEEE Tarns. on PAMI, vol.18, no.8, Aug.1996
[12] "MPEG-7 Homogeneous Texture Descriptor", Yong Man Ro, M. Kim, H.K. Kang, B.S. Manjinath, and J. Kim, ETRI Journal, Volume 23, Number 2, June 2001 41-51
[13] "Color and Texture Descriptors", B.S. Manjinath, J. Ohm, V. V. Vasudevan, and A. Yamada, IEEE Tarns. on Circuits and Systems for Video Technology, VOL.11, NO.,6, June.2001
[14] ."Object detection using Gabor filters", A.K. Jain, N.R. Ratha, S. Lakhsmanan, Pattern Recognition 30(2)(1997) 1167-1186
[15] C. M. Wu, Y. C. Chen and K. S. Hsieh, “Texture features for classification of ultrasonic liver images,” IEEE Trans. Medical Imaging, vol. 11, pp. 141-152, June 1992.
[16] C. M. Wu and Y. C. Chen, “Multi-threshold dimension vector for texture analysis and its application to liver tissue classification,” Pattern Recognition, vol. 26, no. 1, pp. 137-144, Jan. 1993

[17] Y. C. Chen and W. L. Lee, “Texture Classification Using Multiresolution Fractal Feature Vector,” Proc. 4th Asian Conf. On Computer Vision, pp.204-209, 2000
[18] W. L. Lee, Y. C. Chen and K. S. Hsieh, ”Ultrasonic liver tissues classification by fractal feature vector based on M-band wavelet transform,” The 2001 IEEE International Symposium on Circuits and Systems, Vol. 2, pp. 1-4, 2001
[19] W. L. Lee, Y. C. Chen and K. S. Hsieh, “Robust calculation of fractal dimension of images and its applications to classification of ultrasonic liver images and texture images,” The 2002 IEEE International Symposium on Circuits and Systems. Vol. 2, pp. 656-659, 2002
[20] I. Fogel and D. Sagi, “Gabor filters as texture discriminator,” Bio-logical Cybernetics, Vol. 61, 1989, pp.103-113.
[21] N. Sarkar and B. B. Chaudhuri, “Ab efficient differential box-counting approach to compute fractal dimension of image,” IEEE Trans. Systems Man and Cybernetics, Vol. 24, pp. 115-120, Jan. 1994
[22] “The Self-Organizing Map,” Kohonen, T, Proceedings of the IEEE, Vol. 78, Issue 9, Sept. 1990 Page(s): 1464 - 1480