Search

Large, distributed collections of miniaturized, wireless electronic devices1,2 may form the basis of future systems for environmental monitoring3, population surveillance4, disease management5 and other applications that demand coverage over expansive spatial scales. Aerial schemes to distribute the components for such networks are required, and—inspired by wind-dispersed seeds6—we examined passive structures designed for controlled, unpowered flight across natural environments or city settings. Techniques in mechanically guided assembly of three-dimensional (3D) mesostructures7–9 provide access to miniature, 3D fliers optimized for such purposes, in processes that align with the most sophisticated production techniques for electronic, optoelectronic, microfluidic and microelectromechanical technologies. Here we demonstrate a range of 3D macro-, meso- and microscale fliers produced in this manner, including those that incorporate active electronic and colorimetric payloads. Analytical, computational and experimental studies of the aerodynamics of high-performance structures of this type establish a set of fundamental considerations in bio-inspired design, with a focus on 3D fliers that exhibit controlled rotational kinematics and low terminal velocities. An approach that represents these complex 3D structures as discrete numbers of blades captures the essential physics in simple, analytical scaling forms, validated by computational and experimental results. Battery-free, wireless devices and colorimetric sensors for environmental measurements provide simple examples of a wide spectrum of applications of these unusual concepts. With a design inspired by wind-dispersed seeds, a series of three-dimensional passive fliers at the macro-, meso- and microscale are realized that can bear active electronic payloads.

We experimentally investigated the effects of both the compressibility and nozzle width on the local heat transfer distribution of microscale unconfmed slot jets impinging on a uniformly heated flat plate. We made heat transfer measurements under the following experimental conditions; Reynolds numbers of Re = 4000~10000, Mach numbers of Ma = 0.13~0.68, nozzle-to-plate distances of H/B = 3~25, lateral distances of x/B = 0~25, and nozzle widths of B = 300~700 μm having a nozzle aspect ratio of y/B = 30. A thermal infrared imaging technique was used to measure the impingement plate temperature. The experimental results show that for all tested Re and H/B values at a nozzle width of B = 300 μm, the Nusselt number maximum occurred nearly at the stagnation point and then monotonically decreased along the downstream. However, at B = 500 and 700 μm, the maximum Nusselt number point shifted toward x/B ≈ 1.5~2.0. And the Nusselt number increased, as x/B increased, from the stagnation point to the shifted maximum point and monotonically decreased afterward. This shifted maximum point may be attributable to vortex rings promoting sudden flow acceleration and entrainment of surrounding air moving along the jet axis. For the same Reynolds number, the Nusselt number in the stagnation region increased as the nozzle width increased due to a momentum increase of the jet flow caused by the formation of vortices. And, the Nusselt numbers for the smallest nozzle width of B = 300 μm (or highest Mach number at a given Reynolds number) at all H/B and Reynolds numbers tested significantly deviated from those for B = 500 and 700 μm in the downstream region corresponding to x/B > 5, suggesting that the compressibility, when it is high, can affect the heat transfer in the downstream region.

For an optimal design of the rubber domes that are commonly used for keyboard switching, there are design objectives that must be achieved, such as a minimization of the volume to within the required constraints of resistance force. The authors utilized the first-order Taylor series expansion to reduce the computational burden associated with the evaluation of the probabilities, so that the values for individual mean and variance of constraint could be estimated easily by the response surface models. In addition, a Z-score that can provide the mean values and a certain extent of a scattering of product quality was introduced to overcome the difficulties in finding robust optimal solutions from the variety of responses of rubber domes. The present study proposed a procedure for the robust optimization of rubber domes by using a Z-score as a 6σ robust optimization index along with a penalty function. In addition, a new robust optimization process, i.e., the IGEO (Identify, generate, explore and optimize), was proposed on the basis of the penalty function. The obtained optimal solutions agreed well with the simulated results from finite element analysis to within a maximum error range of 0.2543%.

In the washing machines used for the pressure-cleaning of automotive parts, a pair of 2-way valves frequently reverses the flow direction of the water to spray high-pressurized clean water to remove all oil and dirt from machined engine and transmission blocks. Although this valve system is widely used due to its competitive price, its application can be restricted by the surging effect of a pressure ripple caused by inaccurate valve control during the rapid changes in water flow. As an alternative, a 3-way reversing valve can be used to replace the valve system, to provide rapid and accurate changes in water flow direction without the need of a precise control device. During the use of this alternative valve, however, a cavitation effect often occurs due to a complicated bottom plug shape. In this study, the cavitation index and POC (percent of cavitation) were introduced to numerically evaluate the cavitation effect. To minimize the cavitation effect, an optimal shape design was carried out via parametric study, in which a simple CAE-model was applied in order to avoid time-consuming multi-phase fluid-flow analysis and also to avoid the need for convergence, which is difficult to achieve. A full-factorial DOE and an artificial neural network model provided the optimal values required for a status of nocavitation occurrence: waist length, tail length and POC at 6.50 mm, 6.69 mm and 27%, respectively.

A direct-drive generator that uses a permanent magnet provides higher energy density and fewer constraint conditions. For a sufficient production of output power, however, an electronic machine system with a very large diameter must be used due to the need for operation at a low rotational speed. The structural weight of a direct-drive generator for a large-scale wind turbine can be as much as 80% of the total weight. Among the direct-drive generators, the AFPM (Axial Flux Permanent Magnet)-type machine has been the most attractive due to its higher torque effect per unit volume and higher power density. In this study, shape optimization was accomplished based on the desirability function for a direct-drive generator in 2.5MW wind turbine using the AFPM-type machine proposed in this study. Electro-magnetic and structural-coupling analyses were carried out to determine the optimal design variables that would meet the requirements of structural stiffness, such as limitations in air-gap clearance, as well as conforming to acceptable global mechanical behaviors. Compared with the initial generator, the structural weight and stress of the proposed model was reduced by 13.6 and 21% under conditions that satisfied the constraint requirements of the limitations of air-gap clearance.

Concept design requiring complicated feed-forward and feed-back processes makes it difficult for engineers to determine the global behaviors of design variables and objective functions in a design space. Although design of experiments and response surface models have been applied to overcome these problems, the design variables satisfying the objective functions can’t be found due to violations of given constraints. In this study, a new optimization process, i.e., the GEO (Generate, Explore and Optimize) process for the concept design of a tactical missile, based on a MOGA (Multi-Objective Genetic Algorithm) was proposed, which was first adapted to generate a Pareto Front in order to simultaneously satisfy the constraints and the objective functions. In the first step, the weights between the objective functions were determined by using an AHP (Analytic Hierarchy Process). Then, the design space exploration followed, and was implemented by surrogate models constructed from the Pareto Front with a neural network. In the last step, a TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) and a desirability function were applied together to determine the optimal solution. The TOPSIS merged the multi-objective design problem to a single entity, and the desirability function normalized each objective function.

A direct-drive generator for large-scale wind turbine provides higher energy density in comparison with a gear train type generator. An AFPM (Axial Flux Permanent Magnet)-type machine has been the most attractive for the direct-drive generator due to its higher torque effect per unit volume and higher power density. The electronic machine system with a very large diameter, however, has to be used due to operation at low rotational speed of main shaft. Supported structures for the electronic machine system occupy at least 80% of total weight of the direct-drive generator. The light weight design, therefore, is at issue in the industries manufacturing the AFPM-type generator for large-scaled wind turbines. In this study, a topology model of the direct-drive generator for 2.5MW wind turbine using the AFPM-type machine was proposed, which has hollow shapes of the rotor with annular disc, the stator and the main shaft mounted on coupled slew bearings. Analytic results obtained from an electro-magnetic and structural interaction analysis showed that the structural weight of the proposed model compared to conventional generator can be reduced by 30% in a condition satisfying the requirements of the structural behaviors.

The axle drive shaft has important roles such as transferring power and changing the steering angle between the axle and the wheel in a power train system. It is used in most heavy construction machinery, where a high degree of reliability is required in the power train system. However, for axle drive shafts with a long span axle, failures are common at the snap ring cut that is machined on the drive shaft when there is significant fatigue damage under repeated loading conditions. Stress relief grooves have been applied at the snap ring cut to reduce the stress concentration and improve the fatigue life of an axle drive shaft. Although several studies have described how the stress concentration can be reduced by the stress relief grooves, details of the geometries of the stress relief grooves have been subject to debate and even controversy. We investigated the effects of the stress relief grooves on the stress concentration, and estimated the fatigue life of the drive shaft by using finite element analysis, taking into account the geometric parameters such as size and location of the stress relief grooves. As a result, the stress relief grooves presented by non-dimensional geometric parameters for an r/h = 1.2 and a d/b = 2.0 enabled a 22.3% reduction of the stress concentration, and a maximum improvement in the fatigue life that was approximately 3.3 times that of drive shaft with no stress relief grooves applied. These can be an index for selecting optimal geometric shapes of the stress relief grooves.

The delivery of scalable, rich multimedia applications and services on the Internet requires sophisticated technologies for transcoding, distributing, and streaming content. Cloud computing provides an infrastructure for such technologies, but specific challenges still remain in the areas of task management, load balancing, and fault tolerance. To address these issues, we propose a cloud-based distributed multimedia streaming service (CloudDMSS), which is designed to run on all major cloud computing services. CloudDMSS is highly adapted to the structure and policies of Hadoop, thus it has additional capacities for transcoding, task distribution, load balancing, and content replication and distribution. To satisfy the design requirements of our service architecture, we propose four important algorithms: content replication, system recovery for Hadoop distributed multimedia streaming, management for cloud multimedia management, and streaming resource-based connection (SRC) for streaming job distribution. To evaluate the proposed system, we conducted several different performance tests on a local testbed: transcoding, streaming job distribution using SRC, streaming service deployment and robustness to data node and task failures. In addition, we performed three different tests in an actual cloud computing environment, Cloudit 2.0: transcoding, streaming job distribution using SRC, and streaming service deployment.

The structural change of 0.6Pb(Zr1-x Ti x )O3-0.4Pb(Zn1/6Ni1/6Nb2/3)O3 (PZT x –PZNN) from rhombohedral to tetragonal is significantly dependent on the ratio of Ti/Zr. PZT0.570–PZNN ceramics exhibit the highest piezoelectric properties (d 33 = 661 pC/N, k p = 0.68, and e 33 T /e 0 = 3928) as a result of the MPB formation. However, the optimum sintering temperature of this composition is approximately 1200 °C, which restricts the use of pure, cheap metal electrodes in the fabrication of multilayer ceramic actuators. In this study, a low temperature sintering of PZT0.570–PZNN ceramics with the addition of CuO is reported. It is found that the addition of CuO can significantly reduce the sintering temperature to 900 °C. The piezoelectric properties of the 2 mol% CuO-added PZT0.570–PZNN sintered at 900 °C are d 33 = 602 pC/N, k p = 0.65, e 33 T /e 0 = 3570, and T C = 251 °C.

The in-vivo micro-structure and corresponding material property of trabecular bone is important to simulate the mechanical behavior of macroscopic bone structure. In order to simulate the mechanical behavior of bone numerically, the apparent Young’s modulus of trabecular bone should be available. Generally a high-resolution finite-element model based on micro-CT-images could be was used to estimate this value. However, all the previous works regarding this issue have employed either eight-noded voxel elements or four-noded tetrahedral elements, which usually produces large amount of error in estimating an apparent material property. Therefore, rigorous studies on the accuracy of element type for predicting the material properties of cancellous bone have been made in this paper. Micro-CT-data were extracted from a femoral neck to construct three-dimensional finite-element models with three different element types and compression analyses were performed numerically (up to 1.3% strain) to estimate the apparent modulus. Compression tests using the specimens extracted from a cadaver were also performed to validate the simulated apparent material properties using different element types. As a result, ten-noded tetrahedral elements are recommended to obtain reliable material properties of cancellous bone instead of eight-noded voxel elements or four-noded tetrahedral elements.

The thyroid isthmus has different measurements in its width, height, and thickness, and its location with respect to the tracheal rings has been inconsistent, among the anatomical literature. This study was performed to clarify the location and morphometric characteristics of the thyroid isthmus. One hundred thyroid isthmuses of adult Korean cadavers (gender 58 males and 42 females, mean age 62.9 years, range 19-94 years) were used for this study. The distances from the inferior border of the cricoid cartilage to the superior and inferior margins of the isthmus were 4.9 ± 3.7 and 20.8 ± 5.8 mm, respectively. The width, height, and thickness of the thyroid isthmus were 11.1 ± 6.2, 15.9 ± 5.8 and 3.4 ± 1.7 mm, respectively. The thyroid isthmus was located on the 2nd to 4th, 1st to 3rd, and 1st to 4th tracheal rings, in 22, 18 and 18 % of the specimens, respectively. These results are expected to further the current knowledge of the location and morphometry of the thyroid isthmus and provide helpful information for surgical procedures in this region.

Liquefied hydrogen peroxide has attracted attention as a mono-propellant energy source for systems that require a simple, lightweight structure. Its application has been extended to aerospace and deep-sea submarine systems, because this system has no need of an oxidizer and no hazardous products are discharged. For further industrial applications, however, a feasibility study should be carefully conducted with respect to power-supply efficiencies such as specific power density. In the present study, a prototype of a vane motor driven by liquefied hydrogen peroxide with a high density (95%) was developed and performance characteristics such as the specific power density were estimated by measuring the pressure and rotation speed of the vane motor, and by numerical simulation using FSI (Fluid-Structure-Interaction) analysis. The obtained specific power density reached 0.02 kW/kgf, which approximates the levels of the most up-to-date fuel cells.

Orthodontics is the branch of dentistry that focuses on the study and treatment of malocclusions. Orthodontic treatments commonly involve the insertion of an archwire made from stainless steel or an alloy onto orthodontic brackets, thereby inducing movement of teeth into desired positions. However, archwire activation may also induce undesirable tooth rotation because of the applied moment. Appropriate magnitude of gable bend in archwires will prevent unwanted rotation. However, the relationship between rotation and gable bend cannot be easily determined. This study presents a numerical approach to predict tooth rotation with respect to gable bend upon archwire activation. The kriging interpolation method is introduced to predict tooth rotation, thus leading to suggestions in optimum bending angles by minimizing undesirable tooth rotations. Optimization is performed based on reliable kriging models.

To effectively calculate the stress intensity factors in welded components, a weld toe magnification factor is introduced here that will allow for the influence of geometrical effects. This factor was derived from the data obtained in a parametric study performed by FEM. Several sets of weld toe magnification factor solutions have been presented, but these are applicable only to non-load-carrying cruciform or T-butt joints, due possibly to the requirement of very complicated calculations. In the majority of cases for various welded joints such as cruciform, cover plate and longitudinal stiffener joints, the currently used weld toe magnification factors do not adequately describe the behavior of weld toe cracks. In the present study, the weld toe magnification factor solutions with exponents and fractions that have polynomial functions in terms of a/c and a/t — i.e., crack depths normalized by corresponding crack lengths and specimen thickness — were developed through a parametric study using 3-dimensional finite elements for the above mentioned three types of welded joints. These weld toe magnification factor solutions showed a tendency to increase at a/t < 0.2. Meanwhile, for 0.2 < a/t < 0.7, the effect of the attachment on the weld toe magnification factor decreased asymptotically. When the a/t ratio exceeded this range, the weld toe magnification factor became almost one of unity. The fatigue crack propagation life was evaluated by using the proposed weld toe magnification factor and by considering the propagation mechanisms of multiple-surface cracks, and it showed good agreement-to within a deviation factor of 2-between the experimental and calculated results for the fatigue crack propagation life for all welded joint.

Piezoelectric ceramic-polymer composites with 0–3 connectivity were fabricated from lead zirconate titanate (PZT) and poly(vinylidene-fluoride-chlorotrifluoroethylene [P(VDF-CTFE)] copolymer. PZT-P(VDF-CTFE) composites were prepared using the hot press method under various pressure levels of P(VDF-CTFE) at 170°C. The composites were investigated for density, as well as their dielectric and piezoelectric properties, while the particle size of the PZT powder and volume fraction were varied. The connectivity model for the composites was evaluated on the basis of the measured dielectric properties. The piezoelectric voltage constant (g 33 ) was found to achieve a maximum value with a volume fraction of 60 %. In addition, the PZT-P(VDF-CTFE) composite with a 60 % volume fraction was found to yield optimum d 33 × g 33 value of 7.27 pm2/N.

Seat tightness at the fully shut position should be a consideration in the development of a butterfly valve for use in a liquefied natural gas (LNG) vessel. A flexible solid metal seal offers sufficient tightness of the butterfly valve and meets the specifications for cryogenic temperature. In the present study, characteristics for a cryogenic butterfly valve, such as the flow coefficient and the pressure loss coefficient, were estimated by numerical fluid analysis carried out to simulate 3-D flow and to study performance as it was affected by the opening angles of the valve disc. A design criterion to ensure the seat tightness of the butterfly valve at the fully shut position was proposed, in which the contact pressure between the metal seal and the valve disc would be compared with the fluid pressure. Numerical structural analysis showed that the contact pressure can be calculated by simulation of the frictional contact behavior on the surface of the metal seal and the valve disc. As a result, an adequate flexibility of the metal seal and the valve disc was required in order to accomplish a contact pressure that would be high enough to satisfy the seat tightness requirement. Under cryogenic temperature, thermal shrinkage caused the metal seal to adhere closely to the valve disc periphery at both sides and raised the contact pressure to a relatively high value, though there was no contact across a small area at the center position, which is susceptible to leakage. An additional displacement of the metal seal and the valve disc appeared at an operating fluid pressure of 6.9 bar and produced sufficient contact pressure at the no-contact area. This was verified by experimental leakage tests performed at room and cryogenic temperatures.

Robust design for the primary suspension of a railway vehicle was performed according to the optimization of 10 dynamic responses representing driving safety and ride comfort, in which response surface models (RSMs) from the design of experiments (DOEs) were applied. To evaluate the probabilistic feasibility of robustness, an intensive computational process is mandatory. In the present study, the authors utilized the first-order Taylor series expansion to reduce the computational burden associated with a probabilistic feasibility evaluation, thus easily obtaining both an individual mean and variance of constraints. To overcome the difficulty of optimizing the mean and probabilistic variances for the 10 dynamic responses, a process capability index (Cpk) was introduced, which shows the mean value and scattering of the product quality to a certain extent and normalizes the objective functions irrespective of varying dimensions. Consequently, the robust design to optimize the 10 dynamic responses minimized the Cpk subjected to the constraint of Cpk ≥ 2, which satisfied 6σ. The proposed method improved the Cpk that violated the constraints obtained by the RSMs from DOEs and minimized the variance of the Cpk.

We investigated the Sb-doping effects on ZnO thin film using RF (radio frequency) magnetron sputtering and RTA (rapid thermal annealing). The structural and electrical properties of the thin films were measured by X-ray diffraction, SEM (scanning electron microscope), and Hall effect measurement. Thin films were deposited at a high temperature of 800°C in order to improve the crystal quality and were annealed for a short time of only 3 min. The structural properties of undoped and Sb-doped films were considerably improved by increasing oxygen content in the Ar-O2 gas mixture. Sb-doping also significantly decreased the electron concentration, making the films p-type. However, the crystallinity and surface roughness of the films degraded and the mobility decreased while increasing Sb-doping content, likely as a result of the formation of smaller grain size. From this study, we observed the transition to the p-type behavior at 1.5 at.% of Sb. The thin film deposited with this doping level showed a hole concentration of 4.412 × 1017 cm−3 and thus is considered applicable to p-type ZnO thin film.

This article describes the precise anatomic relation and close proximity of the common peroneal nerve (PN) to the tibial nerve branch entering the gastrocnemius muscle (NLG) in the popliteal region. For the study, 22 legs from 13 Korean fresh cadavers (5 men and 8 women ranging in age from 50 to 80 years) were dissected. An x-axis was set as a transverse line crossing the lateral and medial epicondyle of the femur. A longitudinal y-axis accorded with a perpendicular midline between the medial malleoulus of the tibia and the lateral malleolus of the fibula. The PN diverged from the sciatic nerve above the interepicondylar line (-0.4 cm, +10.3 cm) and ran inferolaterally at 20.2 degrees +/- 2.4 degrees . The PN crossed the midpoint of the x-axis (+3.4 cm, 0 cm). The NLG diverged from the tibial nerve above the interepicondylar (transverse) line (0 cm, 2.3 cm) and ran inferiorly and laterally at 17.7 degrees +/- 4.3 degrees . The distance from the diverging point of the NLG from the tibial nerve to the nearest point of the PN was 2.5 +/- 0.5 cm. The distance from the point at which the PN crossed the x-axis to the nearest point of the NLG was 2.7 +/- 0.3 cm. In procedures that involve handling of the lateral gastrocnemius muscle itself or the nerves to the lateral gastrocnemius, surgeons should be aware of the close proximity of the PN to the NLG in the popliteal region.

The aim of this study was to elucidate the anatomical location of nerve entry points of Flexor digitorum profundus (FDP) and its implications for non-surgical neurolysis. A total of 21 amputated forearms of 11 Korean fresh cadavers were dissected. Two transverse x-axes joined the medial and lateral epicondyles and the radial and ulnar styloid processes. The longitudinal y-axis joined the midpoints of the proximal and distal transverse x-axes. The locations of the points were marked relative to the forearm length (x) and forearm width (y). The number of nerve entry points from median nerve and ulnar nerve were average 3.91 +/- 0.62 (range 3-5, median 4) and 2.14 +/- 0.65 (range 1-3, median 2) respectively. Most (82.9%) nerve entry points of FDP from the median nerve were within two circles, with 15 mm diameter. The two circles were on medial 1/10 of forearm width from the y-axis, and on proximal 1/3 (1:2) and 2/5 (2:3) of forearm length on x-axis. Most (80.0%) nerve entry points of the ulnar nerve innervating FDP were within a 15 x 30 mm rectangle. Its center was located at +26.5% on x-axis and -36.0% on y-axis. The nerve entry points used to be selected in performing non-surgical neurolysis with either ethyl alcohol (50%) or phenol (5-12%).

Unicondylar knee arthroplasty (UKA) has gained popularity in the recent years for treating medial compartment osteoarthritis of the knee joint in the Asian population. There is little information about the anthropometry of the resected bony surfaces of the knee joint in these population groups for designing the appropriate size-matched UKA components. We studied the anthropometry of the resected medial tibial condyles in 50 male and 50 female Korean cadavers by using three-dimensional computer tomography. We measured the anteroposterior dimension, the mediolateral dimension at defined points and the condylar aspect ratio. These measurements were compared with similar dimensions of the tibial components from five conventionally used UKA designs. Statistical analysis was performed using Student's t-test, Paired t-test and Pearson's correlation coefficient. We found that three of the tibial component designs showed mediolateral overhang for the whole range of measured anteroposterior dimensions of the resected medial tibial condyles, whereas one of the designs (DePuy) showed mediolateral undersizing for the smaller AP dimensions and overhang for the larger AP dimensions of the resected medial tibial condyles. Another design (Smith and Nephew) showed mediolateral undersizing for the whole range of measured anteroposterior dimensions of the resected medial tibial condyle. We found a decrease in the condylar aspect ratio with increasing AP dimension for our cadaver population data. However, the majority of the conventional tibial prosthesis showed either a constant condylar aspect ratio or an increasing aspect ratio (DePuy) with the increasing AP dimension of the resected medial tibial condyle. Our study may provide guidelines for designing appropriate tibial UKA components for a majority of Asian sub-populations and encourage similar studies in other population groups.

This study examines the anatomic relationships and variability of the facial nerve trunk and its branches, with emphasis on the intraparotid connections between the divisions. Microdissections were performed on 30 Korean half-heads, and the facial nerve trunks and branches were exposed. The average depth of the stylomastoid foramen from the skin surface was 21.0 +/- 3.1 mm, and the distance between the stylomastoid foramen and the bifurcation of the temporofacial (upper) and cervicofacial (lower) divisions was 13.0 +/- 2.8 mm. In 26 of 30 dissections (86.7%), the facial nerve trunk bifurcated into two main divisions, and a trifurcation pattern was seen in the other four cases (13.3%). According to the origin of the buccal branches, we classified the branching patterns of the facial nerve into four categories. In type I (13.8% of cases), the buccal branches arose from the two main divisions of the trunk but not from other branches of the facial nerve. In type II (44.8% of cases), the buccal branches arising from the two main divisions were interconnected with the zygomatic branch. In type III (17.3% of cases), the marginal mandibular branch sent nerve twigs to the buccal branch, which originated from the upper and lower divisions. In type IV (17.3% of cases), the nerve twigs from the zygomatic and marginal mandibular branches merged to the buccal branch arising from the two main divisions. Communications between the facial and auriculotemporal nerve branches, which are known as "communicating auriculotemporal nerves," were observed in 28 of the 30 cases (93.3%). Familiarity with these common variations in the facial anatomy provides useful information for the surgeon in careful dissection, preservation of the facial nerve, and complete removal of the tumors in parotidectomies.

Liver surgeons favor using the entity called the 'dorsal liver' (i.e. the caudate lobe and other paracavally located liver parenchyme of segments 7 and 8). According to minute dissection of 48 livers, we describe the territories of the left/right portal veins, hepatic ducts and hepatic arteries in the dorsal liver. In the caudate lobe, the right hepatic artery, rather than the left hepatic artery (23/48 vs 19/48 for right vs left, respectively), tended to supply the 'left' portal vein territory. Similarly, paradoxical drainage patterns, such as the right hepatic duct draining the left portal vein territory, were found in seven of 48 livers. In the territory of the hilar bifurcation, right hepatic artery dominance was also evident and various bile drainage patterns were found. These included double drainage by the bilateral hepatic ducts (3/48) and drainage into the confluence of bilateral ducts (6/48). In contrast, the arterial supply and biliary drainage of the paracavally located parenchyme of segments 7 and 8 usually depended on the proper segmental arteries and ducts and their variations were within the range of those found in other parts of the right lobe. Therefore, the dorsal liver concept may not be anatomical but, rather, simply aimed at usefulness in surgery. Nevertheless, clear subdivision of the caudate lobe according to biliary drainage and/or arterial supply seemed difficult because of the paradoxical relatioships among the portal vein, hepatic artery and bile duct. Consequently, the present results support extended surgery based on the dorsal liver concept for carcinomas involving the caudate lobe.

We report a rare case of a left middle thyroid artery in a 71-year-old male cadaver. The middle thyroid artery arose from the anterior surface of the left common carotid artery. This artery ran anteroinferiorly and then bifurcated as it entered the lower lateral part of the left thyroid lobe. Both the right and left superior thyroid arteries arose from the common carotid arteries below the carotid bifurcation. The presence of a middle thyroid artery may cause unexpected bleeding during thyroid or neck surgery, and this rare variation needs to be kept in mind during the surgery.

The birthdates of GABAergic amacrine cells in the rat retina were investigated by immunocytochemistry using anti-GABA and anti-bromodeoxyuridine (BrdU) antisera. The ratio of co-localization of GABA to BrdU increased gradually from embryonic-day 13 (E13) and showed a peak value on E18 in the central retina and on E20 in the periphery. After birth, until postnatal-day 3 (P3), a few co-localized cells were observed in the inner nuclear layer (INL). However, in the peripheral retina, co-localized cells were observed in the INL and ganglion cell layer until P5. Our results suggest that the birthdates of GABA-immunoreactive cells vary, depending on cell-type and that there is a temporal lag in the GABA-immunoreactive cell production in the peripheral retina relative to the central retina.

Substance P (SP) immunoreactivity in the guinea pig retina was studied by light and electron microscopy. The morphology and distribution of SP-immunoreactive neurons was defined by light microscopy. The SP-immunoreactive neurons formed one population of amacrine cells whose cell bodies were located in the proximal row of the inner nuclear layer. A single dendrite emerged from each soma and descended through the inner plexiform layer toward the ganglion cell layer. SP-immunoreactive processes ramified mainly in strata 4 and 5 of the inner plexiform layer. SP-immunoreactive amacrine cells were present at a higher density in the central region around the optic nerve head and at a lower density in the peripheral region of the retina. The synaptic connectivity of SP-immunoreactive amacrine cells was identified by electron microscopy. SP-labeled amacrine cell processes received synaptic inputs from other amacrine cell processes in all strata of the inner plexiform layer and from bipolar cell axon terminals in sublamina b of the same layer. The most frequent postsynaptic targets of SP-immunoreactive amacrine cells were the somata of ganglion cells and their dendrites in sublamina b of the inner plexiform layer. Amacrine cell processes were also postsynaptic to SP-immunoreactive neurons in this sublamina. No synaptic outputs onto the bipolar cells were observed.

Two cases of left lobe hypoplasia of the human liver were dissected minutely to show the intrahepatic vessels and ducts. In the left side of the round ligament, one case had a large cyst-like mass containing dilated vessels and ducts, while the other was accompanied by three rugby-ball-like parenchymal masses, each of which was covered by the peritoneum. The former specimen carried the usual Spiegel's lobe, while the latter was very small. Fascial and peritoneal fusion was evident at and around these two left lobes. In both cases the left hepatic vein showed the usual course and terminal end, segment IV was present to a greater or lesser degree. Thus, we hypothesized that, due to fascial and peritoneal fusion, membranous stricture occurred at the peripheral portions of the left portal trunk and left hepatic duct. The discussion focuses on a comparison of intrahepatic vascular configurations in four recently reported cases, including the present two livers.