Ryoka Systems (Japan)

Rapid on-site cytopathology evaluation (ROSE) has been considered an effective method to increase the diagnostic ability of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA); however, ROSE is unavailable in most institutes worldwide due to the shortage of cytopathologists. To overcome this situation, we created an artificial intelligence (AI)-based system (the ROSE-AI system), which was trained with the augmented data to evaluate the slide images acquired by EUS-FNA. This study aimed to clarify the effects of such data-augmentation on establishing an effective ROSE-AI system by comparing the efficacy of various data-augmentation techniques. The ROSE-AI system was trained with increased data obtained by the various data-augmentation techniques, including geometric transformation, color space transformation, and kernel filtering. By performing five-fold cross-validation, we compared the efficacy of each data-augmentation technique on the increasing diagnostic abilities of the ROSE-AI system. We collected 4059 divided EUS-FNA slide images from 36 patients with pancreatic cancer and nine patients with non-pancreatic cancer. The diagnostic ability of the ROSE-AI system without data augmentation had a sensitivity, specificity, and accuracy of 87.5%, 79.7%, and 83.7%, respectively. While, some data-augmentation techniques decreased diagnostic ability, the ROSE-AI system trained only with the augmented data using the geometric transformation technique had the highest diagnostic accuracy (88.2%). We successfully developed a prototype ROSE-AI system with high diagnostic ability. Each data-augmentation technique may have various compatibilities with AI-mediated diagnostics, and the geometric transformation was the most effective for the ROSE-AI system.

This poster presents a collaborative activity among R-CCS and JAXA towards the development of a common in-situ visualization infrastructure on the Arm A64FX supercomputers at both sites, which includes the Fugaku and Fujitsu FX700 on the R-CCS, and Fujitsu FX1000, named as TOKI-SORA, on the JAXA side. This large data visualization-related activity is just a part of a wide-ranging collaboration among the operational groups of the Japanese HPC sites running those Arm A64FX supercomputers. Although each HPC site has its own missions and targets, the visualization and analysis of extreme-scale simulation results still remain a common challenging problem, and thus, we have been evaluating the use of commercial and open-source software aiming to provide a useful in-situ visualization environment for the HPC users. In this poster, we will present the current development stage of this environment on the Fugaku.

The experimental data for the 30P30N three-element high-lift airfoil obtained in 2017 at the JAXA-LWT2 wind tunnel with a Kevlar-wall anechoic test section were revisited, and far-field noise data were compared between phased arrays, single microphones, and a very large eddy simulation (VLES). This investigation aims to explore further utilization of the data for validations of noise prediction based on high-fidelity CFD, such as hybrid RANS-LES and wall-modeled LES methods. The results show a good consistency in the far-field noise spectra between the single microphone and the integrated acoustic maps with the phased array in three directivity angles. With higher directivity resolution, the data from seven single microphones are confirmed to be of good quality in the noise characteristics of the high-lift airfoil. The numerical prediction of far-field noise by the VLES also shows excellent comparisons with the measurements in terms of directivity characteristics, demonstrating the usefulness of the experimental data for CFD validations and the effectiveness of numerical results in interpreting the experimental results.

This paper presents near-field pressure signature prediction for the 1.62%-scale model of National Aeronautics and Space Administration (NASA) X-59 using Japan Aerospace Exploration Agency (JAXA)'s unstructured grid solver FaSTAR. Simulations were conducted for two cases: model-only and static pressure rail. The chosen pressure rail was the shark fin type rail (Rail7), which was newly designed and developed at the Japan Aerospace Exploration Agency (JAXA) and used in the corresponding tests at the JAXA 1 m × 1 m supersonic wind tunnel. The pressure signatures obtained in simulations were compared with those measured in the wind tunnel test. In particular, the Rail7 simulation showed better prediction accuracy in terms of the position and level of the shocks. A series of analyses demonstrates the advantage of the shark fin type rails in validation between CFD and wind tunnel testing. In addition, parametric studies were conducted to investigate the characteristics of the shark-fin-type rail in further detail. These studies were focused on reducing the deflection and attenuation effects of the model shocks occurring owing to interference with the shocks generated by the rail, as well as reducing the pressure increase observed at the aft end of the waveform. The results revealed that a thinner rail width, smaller edge angle (i.e., sharper rails), and larger swept angle were potential options for reducing the interference effects of rail shocks. Furthermore, it was demonstrated that the pressure increase at the aft-end could be reduced by moving the model sufficiently away from the rail. Based on the parametric study results, the potential improved shape of the shark-fin-type rail was validated. The proposed improvement prescribes to reduce the width and edge angle by half and increase the swept angle to 60°. The results demonstrated that using the improved rail geometry, the floating-rail results were significantly closer to the free-air results. This indicated that the effects of rail shock have been sufficiently reduced, which is a significant advantage in terms of design verification of low-boom waveforms. We are investigating more optimized rail geometries based on these studies and are working on new rail designs.

This study numerically investigates in detail the detailed noise generation around slat tracks, which are known to create complex flows and generate high levels of noise at slats. To gain a better understanding of the complex flow structures and noise generation mechanisms, a very large eddy simulation based on a lattice Boltzmann method is employed to analyze the unsteady flow around a slat with a 30-degree sweep angle and a simplified track attached to it. The cross-flow around the slat track creates two types of shear layers that wrap around the track and interact with both the main wing leading edge and the lower surface of the slat, increasing noise level significantly, especially in a high-frequency range. With several modifications applied to the model geometry, this study reveals the primary noise sources and the physical mechanism of noise generation and provides noise reduction concepts.

In this study, CFD simulations were conducted to investigate the noise reduction effects of solid- and porous-plates, as well as their combination and the side-edge effect of the porous-plates, when installed on the side braces of a main landing gear. Initially, a noise reduction study was conducted using a simplified infinite-long side brace-only model. It was shown that the distant porous-plate with both side-edges bent forward was highly effective in reducing the noise by reducing pressure fluctuation observed in the wake generated from the flat side-edges. Additionally, the use of porous side-edges bent forward attached directly to the side brace was also shown to be effective in reducing noise, which has the advantage of being easier to install on the side brace. Subsequently, the effectiveness of these two promising noise reduction methods was evaluated through their application to a three-dimensional geometry of the side brace. It was confirmed that both plate devices with the porous side-edges bent forward effectively reduce the noise level.

Abstract Natural laminar flow design has been applied to the next generation of subsonic aircraft, which are expected entry into service in the 2020s, with the aim of reducing greenhouse gas emissions, which is one of the Sustainable Development Goals. The natural laminar flow wing design was applied to highly swept wings, such as the main wing and the vertical tail plane of TRA2022, which is defined as a 120-pax technical reference aircraft in subsonic aircraft research of the Japan Aerospace Exploration Agency. An automatic inverse design system with inverse problem design tools and flow solvers outside of the system was developed. A natural laminar flow design was carried out more efficiently with the help of various innovations. A natural laminar flow design was successfully achieved for not only the main wing but also the vertical tail plane. In both cases, the laminar flow region was extended by more than 20% in the area. This result shows that the natural laminar flow wing design technology of the Japan Aerospace Exploration Agency is effective for highly swept wings. The objective of the natural laminar flow wing design of the Japan Aerospace Exploration Agency has been achieved.KeywordsInverse designDrag reductionBoundary layer transitionTransition prediction

Novel therapeutic approaches combining immune-checkpoint inhibitors are needed to improve clinical outcomes for patients with cancer. Lymphocyte-activation gene 3 (LAG-3) is an immune-checkpoint molecule that inhibits T-cell activity and antitumor immune responses, acting through an independent mechanism from that of programmed death-1 (PD-1) and cytotoxic T-lymphocyte-associated antigen-4 (CTLA-4). Here, we describe the development and preclinical characterization of relatlimab, a human antibody that binds to human LAG-3 with high affinity and specificity to block the interaction of LAG-3 with the ligands MHC II and fibrinogen-like protein-1, and to reverse LAG-3-mediated inhibition of T-cell function in vitro. Consistent with previous reports, in mouse models, the combined blockade of LAG-3 and PD-1 with surrogate antibodies resulted in enhanced antitumor activity greater than the individual blockade of either receptor. In toxicity studies in cynomolgus monkeys, relatlimab was generally well tolerated when combined with nivolumab. These results are consistent with findings from the RELATIVITY-047 phase II/III trial showing that relatlimab combined with nivolumab is a well-tolerated regimen that demonstrates superior progression-free survival compared with nivolumab monotherapy in patients with unresectable or metastatic melanoma.

View Video Presentation: https://doi.org/10.2514/6.2021-1513.vid This study examines the aerodynamic characteristics of the preliminary design shapes of the Yamato’s “PUPA™ Heavy” concept, a dual-mode detachable cargo pod with a 400kg-class payload, designed to be carried either by being attached onto cargo eVTOL (electrical vertical take-off and landing) aircrafts or loaded onto ground transports together with load units that the company is currently using. Because of its use on the ground, the external dimensions of the pod in ground transportation mode should be defined exactly by the standard of the conventional load units. Under this restriction, there exists a tradeoff between aerodynamics and load volume. In this study, CFD-based assessment of the aerodynamic issues of the preliminary design shapes which keep a balance between aerodynamics and load volume, in addition with the design of the hinged tail section which makes those factor balance feasible.

There is considerable need to generate large amounts of training data that include various operating conditions for fault detection and diagnosis with machine learning in a reusable rocket engine. A system-level simulation model has been developed in which reduced-order models are employed to simulate the global behavior of a reusable rocket engine. Although some components of the engine are not modeled at the system level due to their complexity, they are included among the items inspected during fault detection and diagnosis. This study has developed a regression model for simulating the behavior of such components based on the results of static-firing tests on a reusable rocket engine developed in Japan. The regression model used measurements that can be modeled by the system-level simulation and treated the ones that cannot be modeled as response variables. To identify the operating conditions, the explanatory variables are divided using the Gaussian mixture model in advance, and the Ridge regression models are then trained from the clustered explanatory variables on each cluster. This method reasonably predicts the response variables, even if the static-firing testing includes varying operating conditions such as the combustion phase with varied throttling and the chill-down phase.