SITE LINK

계수공제영상의 비선형 대조도 증강효과에 관한 연구
NONLINEAR CONTRAST ENHANCEMENT ON SUBTRACTION IMAGES

치과방사선 1997년 27권 2호 p.83 ~ 90

이건일, 진연화,

소속 상세정보

이건일 ( ) - 원광대학교 치과대학 구강악안면방사선학교실
진연화 ( ) - 원광대학교 치과대학 구강악안면방사선학

KMID : 0378219970270020083

Abstract

계수공제영상에 대한 선형 및 비선형대조도증강의 효과를 알아보기 위해 하악구치부의 평
행 촬영법에 의한 표준구내사진을 얻어서 스캐너를 통해 모니터상에 띄운 후 세가지 크기와
세가지 회색조차이의 병소를 인위적으로 3가지 texture부위 (plain, bone 및 enamel)에 입
력시켰다. 정상 사진과의 계수공제를 통한 영상을 얻어서 학생 10명 및 전문가 10명에서 모
니터 상에서 판독하여 5가지 등급으로 분류한 후 ROC분석을 통해서 판독능을 비교하였고
다음과 같은 결론을 얻었다.
1. Enamel texture부위에서 선형대조도증강 영상을 제외하고는 전문가의 판독능이 학생에
비해서 유의하게 높았다.
2. 학생 및 전문가 모두에서 비선형 대조도증강 영상에서의 판독능이 선형대조도증강영상
및 증강이 되지 않은 영상보다 높았다.
#초록#
This study was performed to demonstrate the effect of linear or nonlinear contrast
enhancement on subtraction images. Three different textures were radiograped on dental
film. The first radiograph was taken without the presence of an object. the second,
which showed trabucular bone, was taken of the molar area of a human. the third
radiograph was taken of the coronal part of molars. Each film was digitized into a 1312
×1024 pixel x 8 bit depth matrix by matrix by means of a Nikon 35mm film
scanner(LS-3510AF, Japan) with fixed gain and intermal dark current correction to
maintain constant illumination. The scanner was interfaced to a Macintosh LC Ⅲ
computer(Apple Computer, Charlotte, N.C.). This resulted in three pairs of images,
including different textures-plain, bone and enamel. Digital regular, linearly and
nonlinearly enhanced subtraction was performed. Computer software was used to
simulate lesions in the shape of a 2D-Gaussian curve on each of a pair of images. The
each subtraction images were presented in a random sequence to two groups of 10
observers(students and dentists). ROC anatysis was used to compare observer
performance.
The following results were obtained:
1. All of LCE subtraction, equalized subtraction and regular subtraction images of
plain texture were diagnosed the best by far.
2. The data revealed a significant LCE effect in both the student group and the
expert group.
3. Clinical expertise was a helphul factor for the observers in this study.