Grant-in-Aid for Scientific Research (A)
|Research Institution||Osaka University|
INAMOTO Kazuo Osaka University, Faculty of Medicine, Professor, 医学部, 教授 (00068446)
HARAUCHI Hajime Osaka Univ., Faculty of Medicine, Research Associate, 医学部, 助手 (20243256)
TAMURA Shinichi Osaka Univ., Faculty of Medicine, Professor, 医学部, 教授 (30029540)
KIDO Yoshihiro Osaka Univ., Faculty of Medicine, Professor, 医学部, 教授 (20116023)
OGATA Yuji Osaka Univ., Faculty of Medicine, Research Associate, 医学部, 助手 (60281127)
MATSUMOTO Masao Osaka Univ., Faculty of Medicine, Associate Professor, 医学部, 助教授 (50149944)
森川 薫 大阪大学, 医学部, 助教授 (80030013)
稲邑 清也 大阪大学, 医学部, 教授 (90203207)
|Project Fiscal Year
1995 – 1997
Completed(Fiscal Year 1997)
|Budget Amount *help
¥6,700,000 (Direct Cost : ¥6,700,000)
Fiscal Year 1997 : ¥2,600,000 (Direct Cost : ¥2,600,000)
Fiscal Year 1996 : ¥4,100,000 (Direct Cost : ¥4,100,000)
|Keywords||Gastrointestinal Cancer / Esophagus Cancer / Detection of Lesion / Radiography / Digital Imaging / Image Processing / Image Enhancement / High Quality Image / 消化管癌 / 食道癌 / 病変検出 / X線画像 / デジタル画像 / 画像処理 / 画像強調 / 高精細画像|
a) Objective of the Study
In radiographic examinations, early detection of the malignancy is indispensable in the study of gastrointestinal tracts. We have evaluated computer image processing technologies are beneficial or not for visualization and detection of esophagus and stomach cancers on the routine radiographs.
b) Materials and Methods
Materials are the radiographs of 28 cases of the esophagus cancer, 18 cases of normal esophagus cases and GI phantom images. We used several computers and image processing machines with their softwares, including in Macintosh, SUN-workstation, CR-station 1 and Pictrography 3000.
We tried several technics of the image processing : gray - scale transformation, Emboss filtering, edge-enhancement, high-pass filtering and deformation. In esophagus cancer 83% of cases were succeeded. In video image analysis of esophagus radiograhs, we found the examination pattern of esophagus was fixed : X-ray exposure at 3.12(]SY.+-。[)0.95 sec during fluoroscopy 10.1(]SY.+-。[)6.4 sec in the right anterior position and also similar result in the left anterior position. It means an automatic exposure in the fluoroscopy is available by selected time interval.
As a result of image processing, we tried a new idea, in which we applied image processing techniques to improve image quality of CI radiography. Low density barium meal of 50 w/v% was used to make a worse image quality radiograph compared to routine 150 w/v% barium. We are succeeded to improve image quality by several techniques : high pass filtering, reverse, Emboss filtering and edge enhancement. This means thin barium meal easy to drink will be usable in routine examination.
As the good results of computer application after digitalization of the gastrointestinal radiographs, we concluded the image processings were beneficial to clarify early malignant change by enhancing pathologic lesion.