Human bodies work as important media for communication between humans and between humans and machines. Especially nonverbal messages by means of the human body convey more information than verbal ones, and besides, they have functions to control dialogues. Computer graphics (CG) representation of human bodies are, therefore, indispensable in order to make network - mediated communication and communication with computers natural to participants. In this paper types of CG representation of the human body is classified into four categories; 1) precise representation of human body shape,2) human body representation as communication media,3) human body representation as feedback information for human interface, and 4) multivariate data repersentation using human body appearance. The former three types are described by using examples.

View full abstract

The purpose of this work presents a method of subjective evaluation for radiographs of RMI -Breast Phantom [Model 156] by using the fuzzy measure model. Furthermore, we also examied the Quality Control (QC) method by using the standard value based on the mammography guideline. For this purpose, a fuzzy measure model which is one of fuzzy theory and a fuzzy integral applying the fuzzy measure were adopted.
We produced three breast phantom radiographs from each of the three screen - film systems respectively, and carried out an examination by eight observers. The observers evaluated the limit of sight for each sham shadow, and gave an evaluation value on the basis of the method of score summation. With the total score average (the average of evaluation value for image quality factor) and the fuzzy measure allocated for each image quality factor, the fuzzy integral is calculated to determine the total evaluation value (fuzzy integral value) of breast phantom radiographs. Fuzzy measure was determined as the contribution ratio of the image quality factor regarding the precision of the breast radiograph by film readers. It was concluded that evaluation of image quality for breast phantom radiographs by using the fuzzy measure based on the fuzzy measure model and its fuzzy integral became possible, and a method for QC was proposed.

View full abstract

Theoretical and experimentally investigations on the physical quantity for measuring the Wiener spectra and the standard density of film were performed.
Both the density and its transmittance are used for the measurement of Wiener spectra. The density, generally used, has some problem that calculation of its arithmetic mean can not be made and that it depends on the aperture size. The fluctuation of the density depends on the avarage density. On the other hand, the transmittance is more suitable than the density as the physical quantity for the measurment of very small area of film. At the density 0.3, the fluctuation of the transmittance becomes the maximum in any type of films.
It is reasonable that, for the measurement Wiener spectra, the density of film is standardized to be 0.3(transmittance is 0.5) and its transmittance is used as the physical quantity.

View full abstract

The DICOM standard is about to be in wide use. The authors have completed development of the software, "GAZO NO TATSUJINTTM(expert of Images for DICOM)" set on the DICOM standard, which runs on PC, a universal apparatus. This paper describes the outline of the DICOM standard which is required to develop the software. It also describes the introduction of CTN which is a software kit to grasp the DICOM standard, the concept of the software, some problems and their solutions in using the software, and summary, in this order. The DICOM standard is expected to be further developed to form a core element of PACS. This paper is announced to serve as a foothold toward further development of DICOM.

View full abstract

This paper presents a method for automated classification of cardiomyopathies in ultrasonic heart(echocardiographic) images using genetic algorithms(GA). The model of this GA based classifier is built with reference to a neural network (NN). The proposed system is composed of an addition part and a non - linear processing part. The performance of the GA based classifier is compared to that of a decision - theoretic(DT) classifier and a NN classifier. Our preliminary results show that the performance of the proposed GA - based classifier is comparable to that of the DT classifier but slightly poorer than that of the NN classifier.

View full abstract