|
KMID : 0604320060120010035
|
|
Journal of the Korean Society Picture Archiving and Communication System
2006 Volume.12 No. 1 p.35 ~ p.40
|
|
|
Comparison of Two Types of JPEG2000 Observation Programs for Clinical Evaluation
|
|
Hong Seong-Wan
Jeon Yong-Kyeong
Cha Soon-Joo
Kim Yong-Hoon
Hur Gham
|
|
|
Abstract
|
|
|
|
To apply JPEG2000 compression to long-term storage, reasonable compression rate within a certain range shall be obtained through clinical evaluation so that JPEG2000 compressed images do not affect interpretation. To do this, programs required for comparing and assessing between original image and compressed image were developed. This study is aimed to compare and assess between observation program using a monitor and observation program using two (2) monitors in terms of correctness and efficiency. The observation programs for image comparison were developed and implemented using Microsoft Visual C++ 6.0. The subject of this study was 10 losslessly compressed DICOM CR images. JPEG2000 images were compressed with Pegasus kit (PICTools 32-Bit Imaging Development Kits, Pegasus Imaging Corporation, Tampa, FL. U.S.A). PSNR (Peak Signal to Noise Ratio) value was calculated to quantify and compare the image quality of original images and compressed images. The monitors used for comparison were Portrait Type of color LCD monitors (Totoku, Tokyo Japan) whose resolution is 2-Megapixel (1200×1600, below 2M) used for diagnosis at PACS in the hospital to which the researchers belong. The observation program using a monitor and the program using two monitors interpreted the original image at 56% and 51% for 80:1 compression, 76% and 69% for 160:1 compression, and 100% and 96^ for 320:1 compression. Though insignificant difference, it was revealed that the observation program using a monitor could properly detect detailed variation of images. Assessment of images in the observation program using a monitor and two monitors took 659 seconds and 943 seconds for 80:1 compression, 527 seconds and 724 seconds for 160:1 compression, and 320 seconds and 370 seconds for 320:1 compression. It was found that the observation program using a monitor needs less time (26% in total) than the program using two monitors. The observation programs for image comparison developed by the present researchers allowed observers to use a monitor or two monitors, delivering them to an opportunity for selection at their discretion. In particular, as it was shown that the program using a monitor generates better results than using two monitors, it is expected that the result will facilitate experimental environment with high completion as well as correctness and convenience in the future image comparison studies.
|
|
|
KEYWORD
|
|
|
PACS, Image Compression, JPEG2000, PSNR
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|
|