All issues
Volume 11 (1994)
- Issue 3 Pages 87-
- Issue 2 Pages 35-
- Issue 1 Pages 1-
Volume 11, Issue 2
Displaying 1-5 of 5 articles from this issue
- |<
- <
- 1
- >
- >|
-
Akira FUJII1994 Volume 11 Issue 2 Pages 35-41
Published: 1994
Released on J-STAGE: August 27, 2012
JOURNAL FREE ACCESSThe 2048×2048 matrix I. I-TV digital radiography (DR) system has been introduced for GI-tract examinations since January 1990. Furthermore, we applied to mass screening for upper GI-tract since November 1992. All images are displayed and diagnosed on the CRT monitor. The benefits and problems, basic imaging properties as well as clinical evaluation of the DR system comparing with screen/film system and computed radiography (CR) will be present for upper GI-tract examinations.View full abstractDownload PDF (4151K) -
Nader RIYAHI-ALAM, Hiroshi FUJITA, Tokiko ENDO, Takeshi HARA, Katsuhei ...1994 Volume 11 Issue 2 Pages 46-56
Published: 1994
Released on J-STAGE: August 27, 2012
JOURNAL FREE ACCESSWe studied the effects of data digitization on the detection accuracy of subtle microcalcifications on mammograms. Thereby, the required spatial resolution for providing enough detectability of mammographic microcalcifications was determined. Radiographs of a breast phantom, contained four sizes of simulated microcalcifications ranging from 120μm to 230μm, were digitized by five pixel sizes from 25μm to 500μm with 12-bit gray levels, by a drum scanner. Then the images were evaluated by physical image quality index, calculated from displayed amplitude model (peak detection strategy) in detection process, and were also assessed by the visual image quality rank in a human observer performance study. The results in the present work, for the first time of employing a spatial resolution smaller than 100μm pixel size, showed high or enough detectability of subtle microcalcifications on mammograms. In our experimental conditions, approximate 50μm pixel size digitization, with almost the same detectability as 25μm pixel size, provided required detectability of subtle microcalcifications ranging from 120μm to 230μm such that could be observed clearly by radiologists.View full abstractDownload PDF (2507K) -
Du-Yih TSAI, Hiroshi FUJITA, Katsuhei HORITA, Tokiko ENDO, Choichiro K ...1994 Volume 11 Issue 2 Pages 57-63
Published: 1994
Released on J-STAGE: August 27, 2012
JOURNAL FREE ACCESSTwo artificial neural network techniques for classifying possible tumors into benign and malignant ones in mammograms were reported earlier by us. In this paper we describe a revised method for further improving the recognition rate, by inputting two selected features, the entropy and the standard deviation of an image, to our neural network. The result is achieved with 100% accuracy, an improvement of 15% over our previous result. The performance of the proposed method is considered to be comparable to several-year experienced radiologists. This encouraging result indicates that the revised method may be useful for classification of benign and malignant tumors in mammograms, although further tests with larger data set are required.View full abstractDownload PDF (2332K) -
Du-Yih TSAI, Nobutaka TANAHASHI1994 Volume 11 Issue 2 Pages 64-72
Published: 1994
Released on J-STAGE: August 27, 2012
JOURNAL FREE ACCESSThis paper describes an automated segmentation method of liver structure from abdominal CT images using an artificial neural network (NN), together with a prior information about liver location and area in the abdomen cross section and with digital imaging processing techniques. This approach based on the NN is to classify each pixel on an image into one of three categories: boundary, liver, and non-liver. Supervised training technique is used in our experiments. The training data set is obtained from any one of the given set of images by creating gray level histograms for the three categories. The histograms are considered as the respective feature values. Prior to NN classification, preprocessing is employed to locally enhance the contrast of the region of interest. To evaluate the performance of our method, NN-determined boundaries are compared with those traced by two human experts. Our preliminary results show that the proposed method has potential utility in automated segmentation of liver structure and other organ in the human body.View full abstractDownload PDF (3125K) -
Yoshiharu HIGASHIDA, Yasuhiro HIAI, Takao TAKADA, Mutsumasa TAKAHASHI, ...1994 Volume 11 Issue 2 Pages 73-80
Published: 1994
Released on J-STAGE: August 27, 2012
JOURNAL FREE ACCESSModulation transfer functions (MTFs) of the screen-film systems were compared among the results obtained at ten cooperating facilities. Three kinds of green screen-film systems were purchased at the Kumamoto University. Each facility prepared sensitometric strips and slit images. and analyzed its own samples. Results indicated that the MTF data measured at ten facilities showed significantly large fructuation.View full abstractDownload PDF (1130K)
- |<
- <
- 1
- >
- >|