植基於JPEG-LS之二維凝膠電泳影像無失真高動態壓縮技術

A JPEG-LS Based New Lossless Compression Method of 2D Gel Electrophoresis HDR Images

張景閔(Chung-Min Chang)

指導教授 ： 陳民枝

共同指導教授 ： 陳同孝(Tung-Shou Chen)

國立臺中科技大學/資訊學群/資訊科技與應用研究所/碩士(2010年)

https://doi.org/10.6826/NUTC.2010.00002

摘要

蛋白質體學(proteomics)在1995年第一次被澳洲學者V. Wasinger提出後，因為蛋白質體學技術能很快地在短時間內，有效地分析鑑定出各種蛋白質的種類，也因此有越來越來多的研究，開始利用蛋白質體學的技術，加以探討正常組織跟發病組織、致病病菌與非致病菌以及治療前與治療後，細胞內蛋白質表現的變化，蛋白質體學也開始逐漸廣泛地被應用在生物機制、疾病診斷以及藥物治療設計，而觀察這些蛋白質體學的工具最常使用的是蛋白質二維凝膠電泳影像。
二維凝膠電泳影像是目前醫學界最常使用的蛋白質分離工具，蛋白質混合物經過二維電泳分離之後，蛋白質會依據其特性分布在電泳片上，而蛋白質的數量多寡，呈現在電泳片上的顏色(依染劑)深淺也會有所不同，因為這些蛋白質呈現顏色的深淺不同，在拍攝二維凝膠電泳影像時，可能會因為不同的解析度、明亮對比的設定，造成蛋白質表現的變化及蛋白質數量取得的差異，遺失部分微量蛋白的資訊，有可能造成疾病診斷錯誤，或是無法針對關鍵的蛋白質設計藥物或是建立更有效的診斷方式。
在高動態範圍影像的技術中，合成不同曝光時間拍攝的影像，用來完整呈現
vi
出拍攝物每個細節的真實亮度，避免了相機無法真實呈現肉眼所看見的問題，因此我們運用此技術的概念，利用不同亮度所拍攝的影像，可以讓二維凝膠電泳影像，在不同亮度中細微的蛋白質，也可以完整呈現出來，而在論文研究方法中，還利用不同解析度的影像，讓不同解析度的影像資訊差異，可以相互彌補；所以儲存多張不同解析度跟明亮對比的影像，是可以表現出這些細微影像變化，避免微量蛋白的遺失，而如何更有效地將這些影像壓縮儲存，就顯得相當重要，目前市面常用的無失真壓縮方法區分有二種型式；壓縮軟體工具、影像壓縮方法；壓縮軟體工具包含有WINRAR、WINZIP、7-ZIP等等，影像壓縮方法則有JPEG2000、JPEG-LS、GIF等等，而本論文是以Jpeg-LS為基礎，利用其低複雜度、高效率壓縮的特性，提出改善式的HDRR Jpeg-LS的影像壓縮方法，能更有效的提高壓縮率，並利用現存影像模擬出各種不同解析度的影像，進而提高影像判斷的準確率。

關鍵字

JPEG2000 ； JPEG-LS ； 影像壓縮 ； 二維凝膠電泳影像 ； 蛋白質體學 ； 高動態範圍影像 ； 無失真

並列摘要

Proteomics, the study on proteins, was first introduced in 1995 by V. Wasinger. Techniques developed, thereafter, for the study of proteins can effectively analyze and identify the various types of protein in very short time. These resulted in increase protein researches using these techniques to study normal and diseased tissue organization; pathogenic and non-pathogenic bacteria; before and after treatment; and protein changes within cells. Proteomics is gradually becoming widely applied in bioinformatics, disease diagnosis and drug therapy design.
Currently, the 2DGE image is a popularly used method in the medical community for studying proteins via the electrophoresis separation of proteins. After electrophoresis the protein spots are distributed on a gel. The amount of proteins on the gel is different and their color gradient dependent on the staining. The photographed proteins of the 2DGE images are also dependent on image resolution and brightness contrast settings (controlled by shutter exposures). An inaccurate image could sometimes cause wrong disease diagnostic and also wrong administration of drugs to a patient.
viii
The High Dynamic Range (HDR) is a technique that renders images photographed in different exposures. This technique ensures the detailed information of an image by rendering images taken at different exposures so that details from different contrastive can preserved in the resulting image. In the proposed research, this concept is applied to the 2DGE images photographed in different brightness at different exposures so that all details on the protein information are retained. Therefore the need to store the different resolution images is important and essential. Currently, the two classes of commonly used lossless compression methods can be distinguished as compression software tools and image compression methods. Compression software tools include WINRAR, WINZIP, 7-ZIP and more. On the other hand, image compression methods include JPEG2000, JPEG-LS, GIF and more. In the paper, we proposed using the Jpeg-LS compression method with its low complexity and high compression efficiency and integrated it with HDR; the improved propose method is called the HDR JPEG-LS compression method. With HDR JPEG-LS, we can effectively increased compression rate and to retain more protein information that is essential for accurate medical diagnostics. Furthermore, the technique can be used to project images for a different exposure based on existent images to study for possible missing protein details to ensure accurate diagnostics.

並列關鍵字

JPEG2000 ； JPEG-LS ； image compression ； 2D gel electrophoresis image ； proteomics ； high dynamic range (HDR) image ； lossless

參考文獻

[19] 陳秉豐,植基於高動態範圍結構之二維凝膠電泳影像相關研究Researches of 2DEG Images based on High Dynamic Range，pp.36-40，Jane 2009.

[1] Ashikhmin .M., “A Tone Mapping Algorithm for High Contrast Images”, 13th Eurographics Workshop on Rendering, pp.1-11,2002.

[2] Bader, G. D., Heilbut, A. H., Andrews, B., Tyers, M., Hughes, T., and Boone, C. “Functional Genomics and Proteomics: Charting a Multidimensional Map of the Yeast Cell”, Trends Cell Biol, Vol. 13, No. 7, pp. 344-356, 2003.

[3] Blackstock. W.P., Weir M.P. Proteomics: Quantitative and Physical Mapping of Cellular Proteins, Trends Biotech,No.17, pp. 121-127,1999.

[6] DiCarlo, J. and Wandell, B., “Rendering High Dynamic Range Images”, Proc.SPIE, 2001.