By this report, the significant degrees determined by between the cluster classification and the diagnostic results of the doctor were compared. Comparison was performed in 50 genes extracted from 831genes by the significance method. The result by the cluster classification for each gene is 15 points score higher than the other. In terms of the deviation value, there is also a difference of 15 points.

It succeeded in choosing significant components in the data from the data-constellation which consists of labels and components. Let us consider the herb medicine as a label and the herb as acomponent in the section 3. In the case, clear data groups cannot be seen in the data. In such a case, theproposal method could be applied in the validity. Next, the cases with and without the groups (section 4) arealso considered in this text. Finally, like the body-fat which is shown in the section 5, there are alreadyseparated as large, middle, small. There significant degrees are examined among these 3 groups. Thus, anew significant working-out method was developed about these data-constellations.

This paper shows filtering effects on performances for multi scale and orientation of filter bank in the image analysis using subspace classifier. Based on three kinds of intensity distributions, the feature extraction was firstly conducted, and the feature vectors and subspace dimension for recognition were examined.

Afterwards, a series of the analysis concerning accuracies were conducted in the cases of single filter and filter bank. The effects on classification performances for filter bank were compared with the performances for single’s type filter. The deference of filtering effect with color spaces was clarified using the significance degree by SOM.

A medical checkup comprehensive health support tool "Dr. Ningen Dock" has been produced using the previously reported SOM visualization plane map．Here, as for the metabolic syndrome, the personal and the mass medical examination data are applied to this tool.

Conventional fuzzy information theory have an essential problem. A membership value 1/2 should be recognized as a meaningful value of membership function although Shannon's probabilistic value 1/2 is not. Fuzzy set theory based on fuzzy logic which is essentially 3-valued ({0, 1/2, 1}). 1/2 means "Unknown" explicitly.

In the paper, we try to define amount of fuzzy information, which is represented in ratio of cardinalities of "don't care"s. Also we define fuzzy entropy based on fuzzy logical sentences.

Inclusive spectra of protons, deuterons, tritons, ³He, and ⁴He particles from alpha induced reactions have been investigated through model calculations, mainly to improve a theoretical model for understanding the mechanism of cluster-induced reactions. In this report, the intranuclear cascade (INC) model, developed at Kyushu University, has been extended for alpha induced reactions. The incident alpha particles are assumed to break up into fragments stochastically at the initial or the final interaction with the target nucleus. The momentum of a fragment is assumed to be the vector sum of constituent nucleons. Then fragments and particles may undergo the cascade process. Model calculations show reasonable agreement with the experimental data.

The swallowing function is achieved with the tongue, supra-hyoid muscles, pharynx and larynx muscles. The supra-hyoid muscles attached to the mandible pulls the hyoid bone up, the epiglottis covers the trachea during swallowing so that food does not enter the lungs. It is supposed that the maxillofacial skeleton might be related to swallowing performance. The aim of this study is to clarify the skeletal factors that affect swallowing with cephalometric and statistics analysis. The results showed that the mandibular anteroposterior position may be related to swallowing performance.

This paper presents a method of object detection using a moving camera. In order to perform background subtraction with a moving camera case, the proposed method estimates camera motion and creates a current background model from a previous background model using the camera motion. The background model is a set of a single Gaussian distribution. The method employs the upper and the lower bound of a threshold used for the judgment of background/foreground and color information to achieve precise detection of an object. We have confirmed the effectiveness of the proposed method by the experiments performed using some human motion videos.

In order to provide independent living support for lower limb disabilities, an indoor mobile assistance robot is under development in our laboratory. In the preceding study, a remote operation method of indoor mobile assistance robot was developed based on DP matching method according to motion information of hand sign. In order to promote the recognition rate of DP matching, the parameters are necessary to be optimized. In this study, a method of optimizing multiple parameters of DP matching by genetic algorithm was proposed．Hand sign recognition experiments were implemented to verify the effectiveness of the proposed method．

Nowadays the population of elderly people who live alone has been growing continuously. When they encounter an accident at home, they won’t be able to ask for help in the worst case. Under these circumstances, development of an automatic system to care elderly people living alone becomes increasingly important. We have been developing a computer vision system which finds abnormal motions of elderly people among their daily activities indoors. In this paper, we propose a new method of human motion representation called 3D-MHIs (3-dimensional-Motion History Image) based on FoE (Focus of Expansion) and a method of human motion recognition employing 3-D Hu moments by use of a single camera view. The method can recognize not only those motions parallel to a camera lens but also those motions toward depth direction. The effectiveness of the proposed method was experimentally shown.

In recent years, the development of Intelligent Transportation Systems (ITS) has become more active than ever and many methods of detecting people and automobiles using an in-vehicle camera have been proposed to date. However, a method of predicting a pedestrian's rush-out into a road still faces a mountain of challenges. In this paper, we propose a method of detecting a pedestrian and judging his/her direction of walk employing MSC-HOG, an improved version of the HOG, and AdaBoost. The prediction of a pedestrian's rush-out is carried out by tracking each pedestrian and judging the change of his/her walk directions. The effectiveness of the proposed method is verified by the experiments employing car videos.

This paper proposes a method of recognizing finger-spelled letters using the MY VISION system which contains a camera mounted on the chest of a human and a PC. A finger-speller performs the spelling to the camera. The images of the speller's hand are captured and hand areas are detected employing skin color.

For the detection, skin color calibration is performed in advance to cope with various illumination conditions and cluttered backgrounds. The shape and motion of the hand area are analyzed to recognize the finger-spelled letters. The PCA and the nearest neighbor method are employed for the recognition. The high recognition rate with the finger-spelled letters was achieved by the proposed system along with the high speed recognition.

In recent years, wearable devices have been developing rapidly as intelligent tools. According to MIC, the wearable devices market is expected to be 15 times larger in 2020 than 2013. Among them, a wearable camera with a PC has been drawing much attention as a substitute for a human real time visual processing system. However, because such a camera is used by being mounted on a user, the vibration generated by the movement of the user affects captured video images and hence resulting in the difficulty in the video processing. This paper proposes a method of compensating the vibration of a video images captured by a wearable camera by detecting the background movement obtained from optical flow analysis. Experimental results are shown and the method is evaluated.

In previous study, a compact omni-directional wheel was developed with built in motor and reduction gear. The omni-wheel has been applied in the mobile robot’s development.

In order to extend the application scenery for this wheel, the redesign is necessary to make the wheel more compact and powerful.

In this report, a flat motor and a reduction gear built-in free roller nested type omni-wheel is proposed. With this design, the proposed omni-wheel is more powerful and compact. Prototype model was made to verify the utility of the proposed design.

An intelligent walking support machine has been developed for the people with lower limb disabilities. In previous study, the standing up supporting experiment while user catching the handrail of the walking support machine has been conducted. The results showed that, when the robot did not supply an active supporting action, standing up for user is difficult by catching the handrail. Therefore, in this study, based on the accurate identification of user’s intent of standing up action, we develop a standing up support method for the robot to reduce the user’s burden. Specifically, a standing up support method for the robot is developed based on supporting action of caregivers measured and recognized by the acceleration information of their waist. Finally, the effectiveness of the method is valid by experiment.

In the preceding study, an omni-directional walking training machine which can train muscle of lower limb disabilities evenly developed and the effectiveness of the early recovery is recognized by clinical trial. On the other hand, at the stage of the finish, in other words the stage of voluntary training, this robot is required to recognize direction intention of trains’ walk.

Therefore, I’m developing a discrimination method of walking intention using the signal of leg sEMS. Then in this report, I extract the characteristic of signal of leg sEMG and examine possibility of direction recognition.

Due to declining birthrate and aging society, pet animals are introduced to live together with the elderly to improve their "quality of life", and the physiological and social effectiveness for the elderly has been confirmed. However，the introduction of pet animal is difficult for some people and places, due to hygiene and management problem. To address this issue, in this study, a therapy robot is developed.

In this report, for the people who resist the therapy robot without warmth, a pet robot with warmth as animal was developed, and the improvement of the robot was valid.

A seat-style walking training robot has been developed to aid early recovery of patients who have not regained the power to stand. To valid the seat-style walking training robot, the rehabilitation effects on the muscle will be studied. In this study, the muscle use degree related to the seat height and moving direction was analyzed based on the surface Electromyogram information of muscle. Specifically, the surface Electromyogram of muscle required for maintenance of posture was measured when subject travel to front, left, right, and back direction with different seat height. The use degree of the muscle is discovered by the analyzing of the surface Electromyogram information.

The nursing care support robots are highly necessary for the sever bedridden disabled. Since the intent of disabled.

Therefore, a Brain-Computer-Interface (BCI) method to recognize the intent of bedridden person is being developed using the cerebral activity measured by a fNIRS equipment. In this report, we measure the cerebral activity of the subject under a changing temperature environment and extract its feature. A temperature-adjust method was proposed based on the feature of cerebral activity.

In our study, an independent life support robot (ILSR) is under development for lower limb handicapped. In order to provide moving assistance, it’s necessary to identify user’s moving intention for ILSR. Therefore, a moving intention recognition method based on the activities of brain was proposed for providing moving assistance during the double hands tasks in the daily life of lower-limb handicapped.

In this report, functional-near infrared spectroscopy (fNIRS) was applied to measure the brain activities of primary motor cortex and frontal association area for subjects. The time-frequency feature extraction method and user’s moving intention recognition method were proposed. The effectiveness of proposed method was verified by experiments.

The aim of our research is to develop an obstacle avoidance method for a non-directional four-wheeled mobile robot using the ultrasonic sensor and laser range scanner sensor. In our previous study, the different of these two sensors had been discussed. In this report, an obstacle avoidance method is proposed using the feature of data measured by an ultrasonic sensor and laser range scanner sensor. The effectiveness of the proposed method was verified by obstacle avoidance experiment.

In this paper, we describe extension of the voice and picture authentication system. Since a picture has a feature called “picture superiority effect”, we can more easily remember the image of the picture than word such as password. An aim of our research and development is to build an authentication system using the picture superiority effect. We also used associative words by pictures to the voice authentication as passphrases. In the evaluation experiment that we carried out, we used the circular arrangement images. It was found that the circular arrangement is useful to improve a security level for authentication as a result.

In order to provide independent life support for the people with lower limbs disabilities, the movement assistance system, which can response to upper body motion, is under development in our laboratory. In the preceding study, moving intention recognition method of neural network based on upper body motion was developed. And, it was found that the NN which is trained by one subject upper body motion can be applied on others in certain extent. However, the recognition rate is low. In this report, the training method of NN by multi-subjects upper body motion is proposed, and the effectiveness is verified by experiments.

We propose a time-resolved histogram pattern using chaotic analysis as a new method for fish stress evaluation. It is possible to visualize stress over time, by using the contour obtained from this method. In order to improve accuracy, we estimated appropriate embedded parameters and evaluated chaoticity using the maximum lyapunov exponent. As a result, we constructed an appropriate chaotic attractor and confirmed chaoticity. This report illustrates the usefulness of this method.

This paper proposes basic studies on the introduction of communication robots for a system to support health consultation activities in Japanese schools (especially colleges or universities). Therefore, we would like to explain that investigation, these ideas are as follows; first of all, in order to clarify the subjective evaluation of health center’s users, visual analog scale (VAS) apps for mobile devices will be used. Next, in a health center, a communication robot will be set up so that health center’s users can use it. Finally, the interaction between the communication robot and its users will be recorded by rubric evaluation paper.

The purpose of this paper is to investigate the quantitative changes of autonomic nervous activity by using AC electromagnetic fields. Within this research it has been analyzed the indicator of autonomic nervous activity’s variation, LF/HF (ratio of Low Frequency (0.04-0.15Hz) Energy and High Frequency (0.15-0.40Hz) of fluctuation power), previously and post exposing AC electromagnetic fields. As a result, from analyzing all researched subjects, it has not shown any biological effects of AC electromagnetic fields. On the other hand, the obtained data based on the autonomic nervous system activity level at the time of subjective normality have been classified and analyzed. It actually means a finding of a certain rule in the variation of LF/HF. However this experiment haven’t had examined enough data regarding the researched subjects. For this reason, there is a need of gathering more data for the classification of autonomic nervous activity by applying AC electromagnetic fields.

In modern stressful society, it is desired to establish a psychosomatic stress relaxation method. So we focused on the 1/f fluctuation, which is said to give a comfortable feeling and healing effect on people. In the previous study, we show that the sympathetic nerve activity, LF/HF, is reduced significantly by listening to 1/f fluctuation music. However, we obtain that there is the correlation between the trend of LF/HF and the preference about "likes and dislikes” of music listening. This suggests that it cannot show the effect of the 1/f fluctuation without removing an element of the preference such as "likes and dislikes". So, in order to solve this problem, we propose to use the ultrasonic waves modulated by 1/f fluctuation. As a result, we show that the LF/HF decreases significantly when the subjects received the ultrasonic waves modulated by 1/f fluctuation, namely parasympathetic activity has shown to be dominant. It shows to be effective to alleviate the psychosomatic stress by using 1/f fluctuation.

We proposed the positive hypotheses of neural interferences based on physiological knowledge of neurons, ephapse and those quantum effects as engineering models. We thought the neural interferences of axons and synaptic interactions as ephapse are propagated by polaritons, which were a kind of quasi particles. By expanding method of the classical networks to quantum systems with wave equation and path integrals of polaritons, we could obtain both some tools and descriptions for quantum calculations of arbitrary neural circuits. The most important difference between the common (classical) neural network and quantum one are in with or without of an existence of interferences. And concretely we showed how those quantum methods, whose system contained much interference, were applied to the Bayes’ theory, entropy of information theory, and the two-step neural network as a kind of multi channels.

For wildly used walking assistance machine, it’s inevitable for users to make operation by themselves during the assistance process. However, walking assistance machine can actively move to follower user’s intention, it will be more easy and comfortable for lower-limb handicapped to walk freely. In this study, a direction intention identification method is proposed due to lower limb operations using a neural network.

Specially, the walking motion of subject was measured by motion sensor during the process of walking forward walking backward, walking right. According to the acceleration of right foot, direction intention of subjects was identified. By experiments, the validity of the proposed method was verified.

In our study, the control method of non-directional four-wheeled mobile robot without turning over is developed. In order to eliminate the influence of non-linear friction from different ground, a digital acceleration control method was applied in this research. Moreover, the kinematics and dynamics models were derived. The effectiveness of proposed method was verified by simulation.

Thermal conductivity and thermal diffusivity of hen egg-white lysozyme (HEWL) crystals were measured by the magnetic levitation technique of crystals and by the transient short hot wire method. The levitation of crystals was easily realized owing to the magneto-Archimedes effect. As the results, thermal conductivity and thermal diffusivity of HEWL crystals were found to be 0.410 W/(mK) and 3.77 × 10^-8 m²/s in the first measurement of 14 h later from the beginning of crystal growth, and 0.438 W/(mK) and 5.18 × 10^-8 m²/s in the second one of 20 h later, respectively. The experimental conditions, namely, HEWL concentration, pH, and temperature were 6.53 wt% and 3.30, and 17.2C, respectively. Magnetic field was fixed at 4.0 T. As a precipitant agent of crystallization, we used gadolinium chloride of 0.362 mol/kg. This agent is a paramagnetic subject, and is essentially important to cause the magneto-Archimedes effect.

In this paper, we analyze about the readability of online healthcare information in Japanese medical documents. Nowadays, in Japan, the Internet reaches nearly 90% penetrations. All people who interested in healthcare information gather the information about that through the Internet. If medical documents tend to be difficult to read, it will be likely to cause misunderstanding among patients. In the US states, life insurance policies are assured the text readability by using the readability indices (i.e. FRE). We proposed the readability indices for Japanese (i.e. jRE) named “Variable Replacement Method” which is based on the readability indices for English. Our proposed method is able to be used for assuring text readability in Japanese medical document on previous studies. This paper reports the results about text readability analysis using proposed method in Japanese healthcare information, for the purpose of verification whether proposed method is able to use other medical document.

A ghost loiters around in the Japanese Islands. It is a ghost called the vested interest. It is necessary for all nations in Japan to unite to exterminate this ghost. The real nature of the ghost in the Namino, Aso-shi area is shown in " holonic ". As for the human being becoming the leading role, there is a fatal defect while I am demanded with the times of "the heart" from the times of "the thing" .I caught right or wrong in " holonic management" and showed a method of the solution.

Due to increasing number of population of aging society in Japan, most people working in agriculture have become elders. Compared to other fruit trees, Buntan tree is especially shorter than others that bring physical stress on the lower back in working. In this report, The data from surface electromyogram.is shown for approaching new way of reducing the stress.

The length of stay (LOS), fee for service and DPC/PDPS in patients with type 2 diabetes mellitus (2^nd DM) were investigated concerning defined comorbidity complication of DPC/PDPS system by using DPC survey data. The LOS of 2^nd MD was found to be longer than that of the other disease and medical fee of 2^nd DM was also showed to be lower than that of the other disease. Only dementia among 6 kinds of comorbidity affected a LOS by decision tree analysis.

We have studied uric acid level as an indicator of lifestyle diseases, related to metabolic syndrome. The uric acid levels were determined by Self-Organizing Map. In this study, we compared the SOM results with those of the multivariate analysis method, which is used widely for multivariable data.

TETRIX is the expansion kit of LEGO MindStorms and includes powerful motor, battery and high-strength aluminum alloy components. TETRIX and LEGO MindStorms have potential to produce laboratory equipment easily. In this study, we carry out three-point bending test to investigate the strength of aluminum alloy channel in TETRIX. In addition, we produce the movable cushion robot that can carry a person and evaluate the safety of a robot made of TETRIX components. Bending strength of the aluminum alloy channel is 44 MPa, the cushion robot can carry up to 373 kg safely. TETRIX components have adequate strength for laboratory equipment.

In this paper, we examine a design method for low-sensitivity control with robust stability using multiplex feedback control. When the system has low-sensitivity characteristic, the steady state error for the reference input and influence of disturbance for the output become small. But the low-sensitivity control has bad robust stability characteristic, because of sum of the sensitivity function and the complementary function is equal to 1, and then the complementary sensitivity function tend to be 1. Yu et al. propose a design method of low-sensitivity control with robust stability for the plant having uncertain number of unstable poles using two-degree-of-freedom control and double feedback control. In this paper, we expand the result by Yu et al. and propose a design method of low-sensitivity control system with robust stability using multiplex feedback control.

Long time continuous moving images are acquired using a commercially available drive-recorder installed behind the rearview mirror from “Hadano-Nakai highway interchange” to Ueno park through “Metropolitan Expressway”. A simulation program which extracts the white line on the road using the r-θhistogram, was constructed for the preliminary experiment.

This study aims to classify the shape of maxillofacial skeleton and analyze swallowing function. We used TFSOM×SOM for the shape classification. We used CCA-SOM for the swallowing function analysis. The results show that this method can classify cephalometric radiograph according to some features of the shape, and We obtained swallowing function map.

The present study considers a fuzzy color system in which three membership functions are constructed on a tone triangle. This system can process a fuzzy input to a tone triangle system and output the center of gravity of three weights associated with respective grades. Three fuzzy sets (not black, white) are applied to the tone triangle relationship. By treating the attributes of chromaticness, whiteness, and blackness on the tone triangle, a target color can be easily obtained as the center of gravity of the output fuzzy set. In the present paper, the 0% triangle consists of lines of 0% chromaticness, 0% whiteness, and 0% blackness. The colors on the 0% triangle map into the right corner of the tone triangle (on or near C). The difference between the resulting inference output and the transforming output is shown by the nonlinear characteristic between the redness ro of the output (consequent) and the chromaticness co of the output (transformation). The vagueness is influenced by the possibility.

Water allocation is, basically, the process of distributing water to demanding users. The lack of fairness in sharing water of a river basin can generate conflict of interest among the riparian or stakeholders. This situation can be avoided by calling the participants to an amicable agreement for the typical decision problem. Thus, for the sake of establishing appropriate water allocation methodologies, in this study we suggest a model in cooperative game theory and fuzzy logic to support the decision-making process. The study analyzes the sharing of water between the three riparian countries of the Okavango river basin as a case study by using Shapley value to determine numerically the benefit of each country, which are considered as players in the game.

AutoEncoder, which acquires a specific feature space model from unsupervised data, has come to be one of the key technologies for designing a system based on neural networks. In this paper, we conduct an assessment of three types of constraint for AutoEncoder. As results of two experiments, we confirmed the sparse coding constraints is valuable for applying the acquired feature space to noisy photographed document analysis.

This paper is a method to analyze the functional network among brain areas using resting state fMRI. The method estimates hidden parameters using LMS algorithm and examines the relationship among brain areas using clustering. The results in three subjects showed that the proposed method clearly showed the relationship among brain areas.

There is a disease called osteoarthritis of the knee causing pain in that knee cartilage wear down, is that the load on the knee is large problem. In walking, the knee has been found that it takes the largest torque when heel strike. In the present study was carried out to reduce the torque applied to the knee joint by to assist the knee at the time of the ground by a motor that reproduces the spring effect. As a first step, the movement of the knee performed experiments reproduced by the motor was found to reduce the torque at the ground. Produced a brace fitted with a motor, it is scheduled to perform the assist experiment.

The fMRI is used for the brain activity measurement. In the fMRI experiment, a software program was sometimes used for visual stimuli. Those software programs must be tested to succeed the fMRI experiments. In this case, enough money and time need for testing a program in fMRI experiment. Altogether, outside tests of experimental program is desired. In order to achieve this, we created two interfaces using the Arduino. The created interfaces were a synchronization signal generator and a controller. The synchronization signal generator can be implemented in the software program. However, there is a possibility that an operation becomes slowly during execution of the program. Therefore, we created the synchronization signal generator as external equipment. The controller was sometimes used in fMRI experiment. We created a 4 buttons controller similar to a controller of our fMRI experimental environment. This paper describe proposals and verification of a special interface.

There exist a viewpoint that problem of simultaneous stabilizing two plants using a common controller is equivalent to stabilizing a single plant using a strongly stabilizing controller. However, the strongly stabilizing controllers cannot have a pole at the origin. It is said that the simultaneous stabilizing controller has bad robust servo characteristic. The purpose of this paper is to clarity the relation between the simultaneous stabilization and strong stabilization, and present the class of simultaneous stabilizable plants which can simultaneous stabilizing by simultaneous stabilizing controllers with robust servo characteristic.