A New Algorithm for Medical Image Window Transformation

10.3966/199115992017102805018

Zhen-Huan Zhou；Xiao-Jun Wen

grayscale image ； medical image ； shift operation ； window transformation

電腦學刊

28卷5期（2017 / 10 / 01）

202 - 209

英文

Most medical images are 12-bit grayscale images, Window transformation is an algorithm that can map 12-bits raw data(0-4095) to 8-bits displaying data (0-255). This paper presents a new mapping table, first, suppose ww representatives window width, wl representatives window level, then the windows top (top) = wl+ww/2, the window bottom (bottom) = wl-ww/2. we use the window level (wl) divided raw data (0-4095) into low zone (0-wl), the upper zone (wl-4095). The low zone (bottom-wl) in raw data is mapped to the low zone (0-128) in the displaying data, the upper zone (wl-top) is mapped to the upper zone (128-255), in this way, we may make displaying image equalization. Secondly, we use shift instead of the division in mapping operations, have overcome the division caused by loss of accuracy, and improve images clarity significantly. Finally, without knowing the value of the window, we can calculate the histogram of the image to get a more ideal window value. The experimental results show that these methods can improve the clarity significantly in window transformation..

1. 菲力普克勞茲(2017)。殖民性交易與性經驗作為生命權力的部署。交通大學社會與文化研究所學位論文。2017。1-204。