以蓋伯轉換用於去除色雜訊之運用

Using Gabor transform for colored noise denoising

袁啟銘(Chi-Ming Yuan)

指導教授 ： 曹建和

國立臺灣大學/電機資訊學院/電信工程學研究所/碩士(2005年)

https://doi.org/10.6342/NTU.2005.00592

摘要

肝硬化是在台灣是非常常見的一種疾病，傳統上是由醫師利用超音波診斷來加以發現，不過因為人體結構上的不同，以及醫師主觀上的判斷，導致誤診的比例大為上升。因此本研究以科學的方式加以分析，利用蓋柏轉換將從超音波儀器所取得的超音波訊號來計算出肝臟中平均散射顆粒間距的大小，來判斷組織是否為均勻(homogeneous)。不過因為組織的複雜性過大，常會有者過多的雜訊或衰減存在，所以我們使用雜訊估測的方式，並找出一組門檻值來將低雜訊成份去除。如此一來便可以得到較佳訊號成分，進而可以成為輔助醫生診斷肝硬化的一個依據，並且減輕人為觀察上的誤差。

關鍵字

蓋伯轉換 ； 色雜訊

並列摘要

Abstract

Liver Cirrhosis is a very popular disease in Taiwan. Utilizing ultrasound to diagnose Cirrhosis is one of the most frequently used methods by the doctors, but patients have different physiology structure and doctors also have some different subjective judgments, therefore contrived mistake is increased in diagnosis probability.
This research is to apply wavelet analysis to classify two cases between Cirrhosis and normal liver. In this research, I firstly introduce a summary to wavelet analysis, and secondly by means of wavelet analysis to calculate texture energy to proceed texture characterization in a supersonic diagnostic set to get ultrasound images. Finally, to use Multivariate Statistical Methods to effectively classify two kinds of different ultrasound liver images, furthermore to quantify the texture energy to distinguish Cirrhosis and normal liver.
This classified method is an assistance for doctors to diagnose Cirrhosis, so as to reduce man-made mistake in diagnosis probability.

並列關鍵字

Gabor transform ； colored-noise

參考文獻

[2] D. L. Donoho and I. M Johnstone. Adapting to unknown smoothness via wavelet shrinkage. J. Amer. Statist. ASSOC.,1995. to appear.

[3] David L. Donoho and Iain M. Johnstone“Threshold selection for wavelet shrinkage of noisy data”, 1994 IEEE.

[5] David L. Donoho, Iain M. Johnstone, “Neo-Classical Minimax Problems, Thresholding, and Adaptation”.

[7] L.Landini, L.Verrazzani, ”Spectral characterization of tiuuse microstructure by ultrasound : A stochastic approach”, IEEE Trans. Ultrason., Ferroelect., Freq.Contr., vol.37,pp.448-456,1990.

[8] S.G. Mallat, “A Theory for Multiresolution Signal Decomposition：The Wavelet Representation” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 11, no. 7, pp.674-693, 1989.