Flexible feature discovery and structural information control

In this paper, we propose a new information theoretic method called structural information control for flexible feature discovery. The new method has three distinctive characteristics, which traditional competitive learning fails to offer. First, the new method can directly control competitive unit activation patterns, whereas traditional competitive learning does not have any means to control them. Thus, with the new method, it is possible to extract salient features not discovered by traditional methods. Second, competitive units compete witheach other by maximizing their information content about input patterns. Consequently, this information maximization makes it possible to control flexibly competition processes. Third, in structural information control, it is possible to define many different kinds of information content, and we can choose a specific type of information according to a given objective. When applied to competitive learning, structural information can be used to control the number of dead or spare units, and to extract macro as well as micro features of input patterns in explicit ways. We first applied this method to simple pattern classification to demonstrate that information can be controlled and that different neuron firing patterns can be generated. Second, a dipole problem was used to show that structural information could provide representations similar to those by the conventional competitive learning methods. Finally, we applied the method to a language acquisition problem in which networks must flexibly discover some linguistic rules by changing structural information. Especially, we attempted to examine the effect of the information parameter to control the number of dead neurons, and thus to examine how macro and micro features in input patterns can explicitly be discovered by structural information.     

Applying dynamic thermography in the diagnosis of breast cancer

The major difficulty in the interpretation of breast thermography is the complexities of the vascular patterns (false negatives and false positives), and a secondary problem is the existence of cold tumors (false negatives). To overcome these difficulties, we investigated dynamic thermography after cold stress. When a breast is exposed to cold stress, the vascular pattern disappears, and after the stress is removed, the pattern gradually recovers. This phenomenon of thermal recovery can be visualized by sequential thermography or by digital subtraction thermography. In this study, we have attempted to increase the sensitivity of breast thermography by using thermal recovery and sequential and subtraction thermography, hoping that even small tumors with negative steady-state thermography may appear in dynamic thermography     

Parallel processing of network calculations in order to speed up transient stability analysis

Power systems have become very large, and in addition the forthcoming open electricity market will make power system operation more complex. Therefore, power system operation must become more efficient and flexible. Very fast power system simulation is a means of achieving more sophisticated power system on‐line monitoring and control, and parallel computing is a key technology for very fast power system simulation. This paper proposes an efficient and fast parallel network calculation algorithm that will contribute to the development of fast power system simulation programs. Numerical examples show that the proposed method using six processors is about 4.0 times as fast as the usual serial algorithm when applied to a large‐scale radial power system and about 3.2 times as fast when applied to a large‐scale loop power system. © 2001 Scripta Technica, Electr Eng Jpn, 135(3): 26–36, 2001     

Group behavior of agents with an emotional model

Recently, there has been much interest in the study of the formation of groups of agents that cause interactions between agents and invent new functions. We gave some agents an action rule based on the interactions of human feelings by using a circumplex model. It had been decided that the parameters of feelings in this model should have only two axes. In this report, eight basic action dimensions and pure feelings on four corresponding axes were given to agents as a model of feelings and actions based on the multiple factor analysis theory of R. Plutchik, and the behavioral characteristics of the group of agents were examined.     

Formation mechanism of monodispersed spherical core-shell ceria/polymer hybrid nanoparticles

Very unique core-shell ceria (cerium oxide)/polymer hybrid nanoparticles that have monodispersed spherical structures and are easily dispersed in water or alcohol without the need for a dispersant were reported recently. The formation mechanism of the unique nanoparticles, however, was not clear. In order to clarify the formation mechanism, these nanoparticles were prepared using a polyol method (reflux heating) under varied conditions of temperature, time, and concentration and molecular weight of added polymer (poly(vinylpyrrolidone)). The size of the resultant nanoparticles was strongly and complicatedly dependent on the set temperature used during reflux heating and the poly(vinylpyrrolidone) molecular weight. Furthermore, the size of the nanoparticles increased by a 2-step process as the reflux heating time increased. The IR spectral changes with increasing reflux time indicated the increase in the number of cross-linked polymers in the shell. From these results, the formation mechanism was discussed and proposed.     

Direct observation of internal structure in spray-dried yttria-doped zirconia granule

Spray-dried yttria-doped zirconia granules were made transparent by immersion in a liquid and the internal structure was characterized using an optical microscope. This unique technique was found to be applicable for this system by using an immersion liquid with appropriate refractive index, and it enabled observation of the internal structure to be made over the entire volume of granules, in clear contrast to conventional SEM observation. Distinct features, which were considered to be agglomerates, were found in the granules. This was supported by SEM observation.     

Influence of Bile Salts on the Permeability Through the Nasal Mucosa of Rabbits of Insulin in Comparison with Dextran Derivatives

The nasal drug absorption and the effect of absorption promoters have been studied in rabbits. Nasal mucosa excised from rabbits was mounted as a flat sheet in an in vitro chamber. The result indicates that the change in the porosity of the membrane by pretreatment with bile salts increased the permeability coefficient of sodium chloride in the nasal membrane. The permeabilities of dextran derivatives were enhanced by pretreatment with sodium glycocholate (GC). The permeability coefficient (P) of fluorescein isothiocyanate diethylaminoethyl dextran     

A 250-Mb/s/pin, 1-Gb double-data-rate SDRAM with a bidirectionaldelay and an interbank shared redundancy scheme

This paper describes three circuit technologies indispensable for high-bandwidth multibank DRAM's. (1) A clock generator based on a bidirectional delay (BDD) eliminates the output skew. The BDD measures the cycle time as the quantity charged or discharged of an analog quantity, and replicates it in the next cycle. This achieves a 0.18-mm ², two-cycle-lock clock generator operating from 25 to 167 MHz with a 30-ps resolution. (2) A quad-coupled receiver eliminates the internal skew caused by the difference between a rise input and a fall input by 40%. (3) An interbank shared redundancy scheme (ISR) with a variable unit redundancy (VUR) efficiently increases yield in multibank DRAM's. The ISR allows redundancy match circuits to be shared with two or more banks. The VUR allows the number of units replaced to be variable. These circuit technologies achieved a 250-Mb/s/pin, 8-bank, 1-Gb double-data-rate synchronous DRAM     

Recombinant Long-Acting Thioredoxin Ameliorates AKI to CKD Transition via Modulating Renal Oxidative Stress and Inflammation

An effective strategy is highly desirable for preventing acute kidney injury (AKI) to chronic kidney disease (CKD) transition. Thioredoxin-1 (Trx), a redox-active protein that has anti-oxidative and anti-inflammatory properties, would be a candidate for this but its short half-life limits its clinical application. In this study, we examined the renoprotective effect of long-acting Trx that is comprised of human albumin and Trx (HSA-Trx) against AKI to CKD transition. AKI to CKD mice were created by renal ischemia-reperfusion (IR). From day 1 to day 14 after renal IR, the recovery of renal function was accelerated by HSA-Trx administration. On day 14, HSA-Trx reduced renal fibrosis compared with PBS treatment. At the early phase of fibrogenesis (day 7), HSA-Trx treatment suppressed renal oxidative stress, pro-inflammatory cytokine production and macrophage infiltration, thus ameliorating tubular injury and fibrosis. In addition, HSA-Trx treatment inhibited G2/M cell cycle arrest and apoptosis in renal tubular cells. While renal Trx protein levels were decreased after renal IR, the levels were recovered by HSA-Trx treatment. Together, HSA-Trx has potential for use in the treatment of AKI to CKD transition via its effects of modulating oxidative stress and inflammation.     

Thermal conductivity of non-stoichiometric americium oxide: A molecular dynamics study

The aim of this paper is to investigate the influence of multi-valency of americium in its oxide for the lowering of the thermal conductivity and the uncertainty in measurement. In the present study, thermal conductivity of non-stoichiometric americium oxide was evaluated up to 2000 K by the non-equilibrium molecular dynamics calculations using the Born-Mayer-Huggins interatomic potential with the partially ionic model. The oxygen-to-americium ratio (O/Am) was varied from 1.6 to 1.9, which corresponded to the variation of the ratio of Am³⁺/Am⁴⁺. So, we prepared potential parameters for both Am³⁺ and Am⁴⁺. The calculated thermal conductivity of non-stoichiometric americium oxide decreased with an increase of temperature, and the degree of the temperature dependence became smaller with a decrease of the O/Am ratio. This was mainly caused by the phonon-scattering due to oxygen vacancies induced with Am³⁺ ions. Comparing two supercells in which (1) short-range ordered Am³⁺ clusters were contained and (2) Am³⁺ ions were randomly distributed, the thermal conductivity of the former seemed to be somewhat larger than that of the latter.