A freeze-drying device was applied to t-butyl alcohol substituted nerve cells, fibers and synoptic terminals in the human cerebellar cortex for a scanning electron microscopic (SEM) study. The present SEM investigation revealed a small flat polygonal cell with many synoptic terminals using a new method. This cell was irregular shaped one, which had not been analyzed yet, but looked like a basket cell. The number of various-sized synapses on this cell wall was more than 32. And fibers were connected with other cells. The fuzziness of the results of measurements in the microfeatures of cells, fibers, and synapses was reported.

抄録全体を表示

This paper proposed a method to extract and measure the peritoneum automatically. In this algorithm, firstly, the input image is normalized and the edge between the empty area and the monolayer of the flat cells are detected. Then, the peritoneum area is extracted by the using a three layer neural network. And then measurement of the peritoneum is performed. We tested this algorithm by using the real-world image. The experiment results show that this method is very efficient and effective.

抄録全体を表示

Computer-aided diagnosis for pneumoconiosis using Neural Network is presented. The rounded opacities on the pneumoconiosis X-ray photo are picked up quickly through a back propagation (BP) neural network with several typical training patterns. The training patterns from 0.6mmφ to 4.0mmφ are made as simple circles. The neck problem for an automatic pneumoconiosis diagnosis has been to reject the unnecessary part like ribs and vessel's shades. In this paper such unnecessary parts are rejected well by the special technique called "moving normalization".. The new technique called moving normalization is developed here in order to made an appropriate bi-level ROI image. The total evaluation is done from the size and figure categorization. Many simulation examples show that the proposed method gives much reliable result than traditional ones.

抄録全体を表示

BP Learning is widely known to perform good classification for given training data. However, when there is a kind of noise or inconsistency of knowledge in training cases, a neural network may fail to converge. To avoid such a problem, we propose an adaptive evolutional neuro learning method to handle a subset of the complete set of training cases. This method has a sufficient adaptive ability similar to a living organism's evolutionary process based on Darwinian Genetic Inheritance. In this method, the network structure is determined by genetic search in each generation and the connection weights and learning parameters determined by BP learning are not inherited. To verify the validity and effectiveness of the proposed method, we developed a diagnostic system for hepatobiliary disorders.

抄録全体を表示

It is extremely complex to estimate the reference values of lung parameters obtained by spirometry due to the several potential sources of variability ranging from individual characteristics to regional, ethnical, environmental and technical aspects. Earlier works in estimating reference values provide predictive equations developed using multiple regression analysis on spirometric data of people from a particular region. In this paper, an artificial neural network technique has been applied to obtain the reference values of important lung parameters. This study is confined to South Indian population aged between 15 and 30. In comparison with the regression analysis, neural network model is robust and adaptive, and it does not require knowledge of the underlying relationship between input and output variables. A two layer feedforward neural network with 30 neurons in the hidden layer was trained with age, gender, height, weight, and body mass index of 39 normal subjects as inputs and their lung parameters as outputs. Backpropagation algorithm incorporating Levenburg-Marquardt optimisation technique was used for training the neural network. The trained neural network was verified using 5 sets of test data and observed that this technique provides better proximity to actual values than the conventional predictive equations.

抄録全体を表示

The purpose of this study is to develop a diagnosis and simulation tool for a circulatory system model. We use Coleman's dynamic model called Human, which consists of 25 physiological modules including equations of about 1000 lines, over 200 variables, and over 100 static parameters. In this paper, we present a diagnostic support method by obtaining structure of equations, variables, and parameters. The proposed method 1) generates cause and effect relations between variables, 2) visualizes hierarchical directed graph of variables, and 3) enables to track paths in the hierarchy of the circulatory model.

抄録全体を表示

It is well known that early diagnosis of many diseases assures in most cases therapeutic success. The relative great occurrence of neurological diseases that imply a neuro-motor disorder, like Parkinson or Parkinson-like syndromes, as well as their related risk factors, have determined us to introduce a method for objective assessment of the stage of that illness. In its essence, our method is based on the automatic processing and analysis of the image of the handwritten script belonging to a candidate of that disease. The method separates handwritten text into various segments delimited by endpoints and intersection points. Different geometric features, like size, shape and orientation, belonging to clinical Parkinsonians and control (healthy) subjects, are then statistically estimated and analyzed. The result is expressed as a prognostic (risk degree to contact that illness).

抄録全体を表示

The present study considered a fuzzy system in which vague colors can be expressed in words (usually adjectives). The words (color names) consist of a hue name with its modifier, or a hue name only. This system can translate input hue system data to color names (output). Triangular membership functions are applied to the hue angle, and conical membership functions are applied to the relationship between lightness and saturation. By treating the membership function of tone modifier as a fuzzy relation between lightness and saturation, we can easily obtain the membership grade of a target modifier as an output fuzzy set from projection of an input fuzzy set. Three modifiers associated with respective grades indicate vague colors. The present paper proposes a fuzzy system that can determine a few tone modifiers (e.g., dark, dull, deep, etc.) suitable for color names in common use.

抄録全体を表示

The effects of different color signal sequences with the same several colors in the human color impressions were examined. Whether a several-color cyclic sequence has a minimum distance or not in the RGB color space was applied for subjects in an analysis of color sensations. In this study the degrees of pairs of terms applied to color (or color sequence) such as natural-unnatural, pale-deep, and dark-light were investigated. The word natural as a human color impression is, for example, calm, flowing, relaxed, etc., and the word unnatural is intense, tight, unpleasant, etc. in this case. As the results, the well-ordered color signal sequences with the minimum distance (minimum sequences) showed natural degrees. The random-ordered color signal sequences not having the minimum distance (non-minimum sequences) showed unnatural degrees. In addition, for both non-minimum and minimum sequences, the majority of subjects are impressed with light rather than dark. For pale-deep impressions, they are impressed with deep rather than pale. It seems that the impressions of natural-unnatural, pale-deep, and dark-light are independent. And we proposed a human color impression model using the route area indicated by both hue and saturation.

抄録全体を表示

In this paper, we propose a method of personal authentication with keystroke timing patterns based on Self-Organizing Maps (SOM) and introduce an authentication front-end system of a DBMS for clinical cases. At first, we show a necessity of biometric authentication and general privacy policy on the Internet Service as WWW. Secondly, we show how to construct the front-end system, and explain an application of SOM theory. Finally we also show the system ability test to measure False Acceptance Rate (FAR) and False Rejection Rate (FRR) on the maps, which were trained with 40 input-vectors. For result of this ability test, our system keeps FAR at 62.0% when FRR is 20.0%. Thus it seems that our method is available for personal authentication with no hardware for biometric authentication.

抄録全体を表示

In this study, we have analyzed the relationship between the patients clinical backgrounds and the anaerobes from 1995 to 1998 in Osaka Prefectural General Hospital by Data Mining in infectious diseases database. As the result of the analysis, the significant rules (If-Then Rules) were discovered. The smaller usage of drugs resulted in the decrease of anaerobes detection. The overuse of drugs appeared to unbalance the normal bacterial flora. From this result, the Lactobacillus Therapy might be regarded as one of the causative factors for the reduction in the detection of anaerobes between 1995 and 1998.

抄録全体を表示

The problem of knowledge discovering in the form of rules from databases with incomplete information is studied. At first, the rule extraction from databases without incompleteness is surveyed according to the rough sets theory. Then, databases with incomplete information and the ride extraction from these databases are outlined. We briefly survey our previous research, and apply it for realizing some programs of the rule extraction. The implemented programs and the real execution of these programs are shown, too. In this way, a tool for extracting rides from databases with incomplete information is proposed.

抄録全体を表示

The diffusion equation, which describes the time development of diffusion, is applied to wave function problem in quantum physics and to social problem in human science. From the comparison between the fundamental equations in different fields, we make a generalization to the original form of diffusion equation. Starting from the fundamental equation we calculate the predicted distribution for age distribution of three-disease population. Starting from the same equation, we obtain the Schrodinger equation through Wick's rotation. We solve the Schrodinger equation with non-linear potentials by the Monte Carlo method and show the solved wave functions. It is shown that the comparison of two applications is important since we notice the meaning of the parameters and quantities in the equations by the comparison of two applications, and since we can obtain new ideas through the discussions with the researchers in other fields.

抄録全体を表示