We analyze respiration time series of newborn infants with apnea. At first, stationarity, determinism and predictability of the time series are investigated. The local stationarity decreases in regions where the amplitude suddenly varies as typically observed in apnea, and it is difficult to predict the time series in those regions. Thus, we focus on the characteristics of the decrease in stationarity when apnea occurs, and we propose a new method to construct time series of the fluctuation of the local stationarity from observed data. This time series is called time series of stationarity rates in this paper. In advance of applying this method to the respiration time series, we construct the time series of stationarity rates of numerical data generated by deterministic systems while slowly varying a parameter caused by another deterministic driving force, and we apply the surrogate-data method to demonstrate that the constructed time series can be characterized as deterministic. We apply these techniques to the time series of stationarity rates of the respiration data. Our results suggest that the constructed time series of stationarity rates is possibly deterministic in short term.
In order to evaluate the property of a ultrasonic linear scattering, we measured the ultrasonic attenuation characteristics using the fundamental component of the transmitted ultrasonic signal through protein and hen liver. Energy is thought that is absorbed by hydrolysis in the constituents of the tissue and by the relaxation process at low frequency and by relaxation process of motions of side chain alone at high frequency. We describe a study on a protein and hen liver in the relaxation process of tissue characterization by analyzing the ultrasonic absorption spectrum and the attenuation when the temperature is changed. Results showed that the attenuation of a protein and hen liver were increased linearly each absorption between 30 and 70 MHz band.
Pressure ulcer is a symptom frequently found in the patients incapable of moving for themselves in paraplegia or defects in central nervous system. It is caused mainly from chronic pressurization of skin tissue, but many other factors are also reported to promote the pressure ulcer. Shear force is among them. Biomechanical researches suggest mechanochemical reaction of cell that can disturb to repair the wound tissue, however, few studies were made about the relationship between the shear force and the pressure ulcer, because of the difficulty in measuring the shear force without affecting the skin condition itself. To measure the shear force without such interference, we have developed a planar shear force gauge, and measured the skin shear force for its preliminary evaluation. The device was found to be useful for the measurement of the local shear force and the skin deformation. However, further investigation on the material for the torsion frame and the fabrication accuracy is necessary.
Interactions between a lipid-protein membrane and taste substances were investigated using SPR (surface plasmon resonance). The membrane was the LB (Langmuir-Blodgett) membrane composed of DHP (dihex-adecyl phosphate) and the sugar-binding protein ConA. The LB membrane incorporating ConA showed larger responses to almost all the taste substances, such as quinine, caffeine and sucrose than the LB membrane without ConA. Additionally we introduced a dry method, where the resonance angle change is measured before and after drying water from the sample, to detect pungent substances. We compared results of other taste substances with those of pungent substances using this new method. Suggestion was also made on the interactions of astringent and pungent substances with the charged LB membrane.
The purpose of this study is to apply the growth mechanism of plant to sensing system. In this paper, the one-output for two-input sensing device imitating the mechanism of photosynthesis is introduced. As one example of the device applications for a transport system, an automobile which recognizes the driving environment is presented. As a result, the possibility of applying the photosynthesis mechanism for engineering is shown by proposing the quantitative setting of automobile speed which generally depends on the intuition of a driver.
Quantitative assessment of impairment in hemiplegic patients was attempted by accelerometry in the head of fibula. A portable microprocessor-based data acquisition system was used to measure the acceleration at the head of fibula during gait in ambulatory subjects and to evaluate temporal and spatial parameters of the gait. After examination, the recorded data are downloaded to a personal computer to analyze the gait cycle, power spectrum and phase-plane locus of two directions. Experiments were performed on 30 hemiplegic patients of different Brunnstrom stage (Br. stage) grading. The orthogonal accelerations at the head of fibula were measured when walking at a normal speed. The gait cycle for a subject of a low Br. stage was significantly larger than that for a high Br. stage. The power amplitude in frequency distribution for a low Br. stage subject was significantly smaller than that for a high Br. stage. The area of phase-plane locus is larger in higher Br stage. This monitor may permit a quantitative assessment of the recovery of hemiplegic patients, both in a hospital environment and later in their homes.
Focal magnetic stimulation using a figure-of-eight coil is a useful method to investigate the intact nervous system. Analytical studies using models of magnetic stimulation of a nerve fiber indicate that the nerve fiber is stimulated by the gradient of the component of the induced electric field that is parallel to the fiber. Theoretically, the nerve fiber which is parallel to the induced electric field has the minimum threshold for the excitation. However, some experimental studies show that the nerve bundle perpendicular to the induced electric field, which is parallel to the wings of the figure-of-eight coil, has the minimum threshold for eliciting the muscle response. In this study, the reason for this inconsistency between the theoretical results and the experimental results is explained using an analytical model. It is verified that the threshold for the excitation of the nerve fiber which is parallel to the wings of figure-of-eight coil is lower than that of nerve fiber parallel to the coil wings when the nerve fiber is away from the center of the coil at some distance.
The study intends to clarify the effect of the static magnetic field on the cellular activities. This paper is concerned with the metabolic activity of the mitochondria from the viewpoint of its energy level. As a preliminary examination, the time required to render the cells to the low energy state to the G0/G1 phase of the cell cycle and to reduce its energy level, was examined by keeping the cells in the medium with low-density fetal bovine serum. Then, we examined the effect of the magnetic field on the mitochondria energy activity from the viewpoint of the cell metabolism estimated by the MTT assay. We also examined the cell respiration, expecting that the mitochondria energy activity level will be reflected on the respiration. The magnetic field was kept at 100 [mT], as well as 0 [mT] as the control. The results by MTT assay indicated that the mitochondria energy activity seems to be enhanced by the factor of 1.5. We found that the cell respiration is increased by the factor of 1.3 by the exposure to the magnetic field. It is seen by the t-test that the difference is statistically significant. It is thus concluded that the mitochondria energy activity is enhanced in the G0/G1 phase of the cell cycle by the exposure to the magnetic field. On the other hand, we also found that the mitochondria energy cycle period is little effected.
A telemetry system was developed to monitor a vital sign of the person who needs continuous monitoring. To make a long-term monitoring possible, a transmitter was made small and wearable as a finger-ring. As a vital sign, a photoplethysmographic signal is detected at the finger on which a ring-type telemeter is worn. The signal is sampled, and is transmitted to a receiver using a radio wave. At the receiver, the signal is demodulated into cardiac pulses. Then, the rate and the amplitude of the pulses are monitored continuously. If the pulse interval falls out of a predetermined range, or becomes undetectable, an alarm system is activated. The alarm system informs the emergency by an alarm sound or by automatic emergency call. A test-system was manufactured. It was as small as an ordinary ring, and easy to be worn all day long. The fundamental functions of this system were examined, and its performances were satisfactory.
A stereoscopic input system of 3D-coordinate is developed for the stereoscopic design of machinery parts according to their stereoscopic view by binocular parallax. 3D-coordinate is stereoscopically input by 3D-mouse cursor synthesized as an input device which freely moves in any direction of real and imaginary spaces. It is used for the realization and change of stereoscopic design of an image object displayed in a real time. An input-shape is given as a trajectory of an input device. The stereoscopic intermediate image is obtained on every input-operation making clear a trace of edge-point of an input device with rest domains characterized by different colors. The complete image of an object gives necessary information which is directly used for the realization of its actual parts with appropriate characteristics.
The nonlinear continuous system identification using radial basis function networks is studied in this paper. A learning law to adjust the network weight parameters and to suppress the effect of the modeling error is provided. The convergence of the learning law and the boundedness of adjusted parameters are proven based on the Lyapunov stability theorem. The results of simulations for a two-degree-of-freedom manipulator moving in a vertical plain show the effectiveness of the proposed identification scheme.
The evolution process of the system of galaxies is studied based on statistical time-series analyses of radiowave intensity variations of the extra galactic radio sources. The object galaxies are 21 quarsars and 7 BL Lacertae galaxies at cosmological distances. The fluctuations of the intensity of waves (11cm, 3.7cm) observed daily by the Green Bank Interferometer over a three year period are analysed to get three characteristics (power spectral index, fractal dimension, Hurst exponent) for each galaxy. Considerations based on the analysed results tell us that the galactic system evolves. The nonlineardynamical analysis of the time series is needed to find the time evolutional function to unique to each galaxy. It is hoped that a general law for the structure of evolutional process will emerge for all the galaxies.
It is clear that facial expressions are important on face-to-face communication. Humans can not only guess other person's psychological state from facial expressions but can also give own intentions to other persons by using them. If users can use facial expressions to a computer and the computer can correctly recognize them, an effective human interface can be constructed. Therefore, it is necessary to establish an effective facial expression recognition technique from facial image data. In this paper, we propose an application of the MTS (Mahalanobis-Taguchi-System) method for facial expression recognition. Especially, we treat the facial expression recognition from stationary facial image data. The MTS method is a statistical pattern recognition method using the Mahalanobis distance. This method can consider the correlation among each attribute in the recognition, and can optimize the number of attributes based on the results of recognition. We also propose an optimization method with the genetic algorithms. We construct the recognition system based on the MTS method, and then evaluate the effectiveness of the proposed methods through practical experiments. The experimental results revealed that our proposed methods could correctly recognize the facial expressions and could effectively optimize the number of attributes.
This paper presents a new system architecture of downlink data formatter for the earth observation satellites. A pilot system based on the system architecture has been developed. This new system is able to process downlink data stream from various kind of satellites and sensors, and therefore to solve some problems such as degradation of reliability in conventional downlink data formatting systems.In the evaluation experiments, the pilot showed sufficient performance to downlik data from MOS 1 MESSR, Landsat TM, and ERS AMI.
In general, it is hard to determine the network structure because it is related to the generalization ability. Moreover, it is also hard to analyze networks trained by back propagation learning. In order to solve these problems, structural learning with forgetting (SLF) has been proposed. In this paper, we improve SLF in terms of structuring ability, and propose parallel multi-layer networks. Using our method, (1) wastefully distributed representation of hidden units are suppressed without revival of unnecessary parameters, (2) forgetting is accelerated, (3) network structure is automatically determined, and (4) classification rules are extracted in a discrete valued inputs problem and a continuous valued one. This method is applied to the XOR problem and the thyroid function classification as a practical problem of continuous valued inputs. It is found that our method is twice faster than SLF and its success rate is about 100% in terms of obtaining the smallest number of hidden units.
The Sensor Glove II (SG II) is a glove type haptic device. The SG II supports 20 degrees of freedom which corresponds to the total D. O. F. of the human hand. It can measure the force and the angle of each joint and give force feedback in each joint. The motors move pulleys of the device through wires in tubes. It makes unexpected disturbances and makes device control difficulties. In this paper we control the SG II with a disturbance compensation based on a sliding mode observer. The sliding mode observer is an efficient scheme for compensating nonlinear disturbance. The results of the free motion experiment show its effectiveness.
This paper deals with the robustness and the fragility of the H∞ control of a nonlinear magnetic suspension system using the feedback linearization, from an experimental point of view. At first, two nonlinear models are derived with different assumptions, in order to evaluate effectiveness of modeling. Next, for the resulting models, the H∞ control with the feedback linearization are conducted respectively. Then, the time response of the gap against a step disturbance is shown as an experiment. Finally, the usefulness of the detailed nonlinear model, the effect of the feedback linearization, and the robustness and the fragility of the feedback linearization-based control are evaluated from comparative experiments.
Magnetocardiogram (MCG) measurement using high-Tc SQUID magnetometer is receiving attention as a diagnostic technique for observing the electrical activities of the heart because the instrument is easy to operate and the liquid nitrogen coolant is low cost compared with the liquid helium used in conventional SQUID magnetometers. The magnetic field of the heart is weak and is influenced easily by surrounding noise in the measurement process using high-Tc SQUID. So the estimation of the cardiac signal source becomes difficult. In this study, we carried out the MCG measurement on normal subjects and patients using a 16-channal high-Tc SQUID in a prefabricated magnetically shielded room. We used the spatial filter method (SFM) to reduce the errors in the measurement and analyses of the signal source. The results of the simulation and actual measurement of normal subjects and patients show that the method is useful for the estimation of the signal source and can also be used to analyze other biomagnetic signals. The method provides assistance in the diagnosis of cardiac diseases. The low cost and simple operation of the high-Te SQUID is advantageous in clinical medicine applications.
In the last two decades, linear-in-parameter nonlinear polynomial models for NARX (Nonlinear Auto-Regressive with eXogenous input) systems have received considerable attention. The keypoint of nonlinear polynomial model identification is how to select a set of significant terms employed to approximate the NARX system under study, from a large number of candidates. To this end, the orthogonal least squares method which is a local search procedure, and the genetic algorithm approach which has a high potential for global optimization have been proposed in the literature. However, it is considered that the methods reported so far in the literature still lack potential to identify the nonlinear polynomial models with relatively high-order. This limits the applicability of the nonlinear polynomial models to the real complex nonlinear systems. Motivated by this fact, in this paper, a new genetic algorithm approach to nonlinear polynomial model identification is proposed. Our contribution in this paper is to introduce a novel hierarchical encoding technique which is considered to be suitable to the structure of the nonlinear polynomial models. Simulation and application results are also included to verify the efficiency of the proposed identification algorithm.
The mechanism of neuromodulation with serotonin was electrophysiologically revealed during excitatory neurotransmission in the hippocampus by use of the stimulation-recording method. Stimulating electrodes were positioned at the granule layer, and recording electrodes were set at the CA 3 layer in the hippocampal slice preparation of rats. Presynaptic neurons stimulated with single or double pulse train induced excitatory postsynaptic potentials (EPSP). EPSP was recorded and evaluated in the CA 3 injected with serotonin (0, 10, 20, 50 μ M) by means of the wave analyzing program. Firstly, results of the single pulse stimulation show that serotonin pharmacologically reduces EPSP in the hippocampal CA 3 neurons. Secondly, evaluated EPSP responses to the double pulse stimulation are applied to an electrophysiological model. Consequently, it is suggested that serotonin inhibits the release of calcium dependent synaptic vesicles as a neuromodulating mechanism.
This paper is concerned with a new design method of two-degree-of-freedom optimal deadbeat controller including internalmodel. In this servosystem, the compensators can perform the desired value tracking characteristics and disturbance rejection characteristics. It is shown that such a servo system which includes an internalmodel into the local feedback loop can be designed. Two kinds of optimal deadbeat control signals which can minimize the different quadratic performance index, one affect the control signal and deviation, the other affect the control signal and output, can be easily obtained from matrix computations using sampled data of the step response of the system. Then the deadbeat controller can be designed by using the sequence of control signals. A numerical example is given to illustrate the effectiveness of the proposed method.
Capturing the timely information on customer needs is a critical success factor for mass-customization. Many methods for estimating customer's product preference have been developed. However, the methods that use POS data fail to collect the information on non-purchasers while those that use elaborate questionnaires such as conjoint analysis cost much and take time in implementing them. In this paper, we propose a method for analyzing customer's preference on products with multiple attributes like personal computers in an internet, online shopping environment. Particularly, the proposed method utilizes the data on each customer's browsing history among possible product configurations, which can be available to the seller at low cost and in real time. This method consists of three parts; (1) collecting the data on browsing history for each customer, (2) converting the browsing data into the ordinal preference data, and (3) modeling the relationship between product preference and product attributes for each customer. In an experiment, we show that the proposed method approximates well the outcome of conjoint analysis. Thus, it can be a tool for tracking the shift of customer's preference in a timely fashion.
Left ventricular thickness, diameter and their time variance play importantt role in an assessment of a cardiac function using echocardiography. In order to measure left ventricular wall thickness and diameter from echocardiogram, it is required to detect a border which is a boundary between cardiac muscle and blood. The method of border detection requires reproducibility, high accuracy and high-speed. Many of methods of border detection have been proposed before now. However, they require much calculation. In this paper, we propose the high-speed detection method using projection operators. The method is composed of two steps. Firstly, we extract candidate points for the border under the assumption that points on the border take almost extremes of gradients of an image. Then, we estimate the border curve using two kinds of projection operators under assumption that the border is a smooth and continuous curve. This method realizes about 10 times high-speed compared with a conventional method.
FPN (Flexible Private Network) is a new service, which enables a user to join multiple secure communication groups within an Intranet. FPN defines a secure communication group by a single, encryption key (session key). All communication entities in a FPN request a session key to the key management entity and share the same encryption key. Also, to keep the secrecy of FPNs well, session keys should be updated frequently. In this paper, we consider a realization method to apply FPN over the enterprise networks and define key distribution method for session key sharing. An application of this method for secure communication systems will show that inexpensive and flexible VPNs with various communication groups can be constructed easily. It is also shown that the proposed method works effectively in an enterprise network system.
Typhoons strike Kagoshima district every year and sometimes hit power systems very badly and cut off the electricity. In order to resume normal services as soon as possible, one needs to predict the damage by typhoons accurately. This paper proposes a prediction method of the damage of power systems by using a linear regression model, in which the date on typhoons and harms are efficiently normalized by genetic algorithm. From weather information about the forthcoming typhoon, it enables us to predict the damage of electric lines, poles, and so on. Effectiveness of this prediction method is assured by applying it to the actual data.
This paper proposes a method to extract subjective visual curves in an image including discontinuous curves and noisy points based on the proximity and the continuity in the perceptive grouping factors. A pair of points is selected as the starting point and the goal point, and some points are connected sequentially from the starting point by evaluating the proximity and the continuity between neighboring points. The global form of the curve that is composed by the lines between the connected points is also evaluated by the proximity and the continuity, and the evaluated value is voted into the voting space. After the voting process about all pairs of points is completed, the addresses that have the large voted values are extracted as the components of curves. Experimental results show that the curves were extracted that had the similar forms to be perceived by the visual sense in noisy images.
A multiwavelength differential absorption lidar system for measuring vertical concentration profiles of atmospheric SO2, NO2, and O3 was developed. The system contains two Nd: YAG pumped dye lasers, each capable of alternate oscillation with two wavelengths, and can emit a sequence of four wavelengths in the ultraviolet and visible regions. The geometrical form factor was calculated for biaxial and coaxial configurations, and the former was adopted for its ability to eliminate scattering from close range during long range measurement. The system can be used for high-resolution, multiwavelength DIAL measurement of a single species or simultaneous two-wavelength DIAL measurement of two species. As a preliminary measurement, O3 concentration was measured in the lower troposphere. The return signals are in agreement with the calculated geometrical form factors, and simultaneous null and DIAL profile measurements show that the instrumental error due to beam misalignment was low.
This paper describes a load dispatch method which minimizes power cost, i.e. [fuel cost]/[electric output], for a power system with thermal plants and energy storage facilities. The proposed method employs fractional programming to convert a minimization problem with fractional objective function to a series of quadratic minimization problem, and semi-definite programming to solve converted problems. The method provides the optimum time-dependent power output/input and storage level of energy storage facilities as well as time-dependent power output of thermal plants. The method has been applied to a power system with 5 thermal plants, 2 energy storage facilities of various performances and 5 load demands. The optimum load scheme of 4 time mesh points is obtained for the thermal plants and energy storage facilities. The fractional programming successfully converges the optimal scheme through a few iterations. The semi-definite programming deals with a variable matrix of 164 dimensions, and 185 inequality constraints.
The search ability of the function discovery system named S-System has been improved in this paper. In the system. there exist many individuals called bugs. Two strategies for improving the search ability are introduced: (a) when a bug moves in a direction that lowers the fitness, it turns back, and (b) only a limited number of bugs can exist in a phylum. Strategy (a) shortens the number of generations required in finding the true function and strategy (b) improves the discovery rate. The effectiveness of these strategies is confirmed by implementing them in a problem of finding Kepler's law using the S-System.
We proposed a novel scheme for microscope auto-focusing in high precision. The objects in the different planes, which are taken in the same image, can be in focus separately, by using the different focusing areas. These areas are selected in the regions around the edge of the rough focusing image, and the shapes of their focusing areas are determined according to their different situations. In order to meet the requirement of different situations, a database was used for storing the different parameters.
The observation of the magnetic flux line lattice in superconductors using the Bitter method provides the information of the microscopic response of the vortex system to the forces acting due to vortex-vortex interaction as well as the interactions with pinning centers. The static and dynamic structure of the magnetic flux line lattice in superconductors has been the focus of much research. In this letter we propose a system using image processing which is automatically able to analyze the distribution of the magnetic flux line lattice in superconductors obtained through the Bitter method.
This paper proposes a magnification method for MR Brain Image due to fuzzy clustering. A pixel of boundary area for organization in MR Image is mixed pixel (mixel) that consists of two organizations. Organization mixture proportion is estimated to distinguish between mixel and pure pixel using fuzzy clustering. Next, using information of eight vicinity pixels magnifies MR Image. It is observed that the proposed method is able to accurately magnify MR Image in comparison with the conventional magnification methods. The experimental results indicate the effectiveness of proposed method.
A dynamic tunneling algorithm (DTA) based on dynamical system is an effective method for global optimization problem. The algorithm has been only applied to unconstrained optimization problems. In this paper, we propose an algorithm, which can apply the DTA to constrained problems by resolving region of boundary of constraints. A typical numerical example illustrates the validity of the proposed algorithm.
In this paper, the authors propose a method of identifying individuals using an angled facial image. We think that we can obtain the more effective data for identifying faces by setting the CCD camera at an angle to the faces. When we set the CCD camera like this, the effective information is able to be obtained from the large region of one side of the faces, the information of nose height which is one of three dimensional facial information and the cheek undulation which is part of the outline of their profiles.