Who cooperate with the social research?

When conducting research—especially studies involving people or society—it is essential to obtain the cooperation of participants. But who chooses to cooperate, and how can we build lasting cooperative relationships? A quantitative text analysis of open-ended responses from a panel survey conducted over more than twenty years revealed that: (1) financial compensation serves as an initial motivation for participation; (2) receiving feedback on the surveys in which participants took part; (3) developing a sense of involvement in the research; and (4) the normalization of survey participation through these factors contribute to sustained cooperation. Conversely, the analysis also found that individuals tend to be less willing to participate during periods of non-regular employment or when on leave due to illness. This suggests that people who perceive themselves as deviating from a socially “standard” life course are more likely to refrain from participating in surveys. Recognizing that selection bias arises at the very stage of survey participation itself, maintaining long-term cooperative relationships with respondents is crucial for ensuring the collection of high-quality survey data.

Visualizing the Voices of Workers on Mental Health Leave through Participatory Action Research

Although a variety of measures have been implemented to address mental health challenges such as depression, the voices of individuals who have actually taken a leave of absence due to mental health difficulties have rarely been incorporated into these measures. Consequently, their lived experiences have often remained invisible. This invisibility has been shaped by strong societal stigma toward mental health conditions, power imbalances between employers and employees, and a lack of suitable research methodologies. Using a participatory action research approach, this study sought to make visible the lived experiences of people who have experienced work leave due to mental health issues. The process not only generated insights for creating recovery-friendly workplaces that extend beyond medical or welfare perspectives, but also fostered empowerment among the participants themselves. In this presentation, part of our ongoing project will be introduced, and the potential for translating these findings into practical implementation will be discussed.

Understanding Highly Creative Scientists: A Clinical Psychological Approach

Traditionally, the psychological study of creativity has positioned divergent thinking (the ability to generate many flexible, novel ideas) as its primary component. Consequently, the cognitive and personality traits of researchers who produce practical, real-world outcomes have remained largely unexplored. However, the understanding of creativity is recently shifting to emphasize the importance of convergent thinking (evaluating multiple ideas to narrow down an optimal solution) and the necessity of assessing "real products"—outcomes judged as high-quality by experts in the field. Therefore, this study defines creativity as the “ability to produce novel and useful products” (Mumford, 2003). We applied specialized clinical psychological methods (psychological assessments) to high-achieving STEM researchers and graduate students to describe the traits supporting their creativity. This presentation will introduce the cognitive and personality profiles of the resulting researcher groups, the relationship with their specialized fields, and a discussion of the developmental changes observed in the researcher's path to excellence.

Emergence of the Primacy Effect in Structured State-Space Models

Structured state-space models (SSMs) have been developed to offer more persistent memory retention than traditional recurrent neural networks, while maintaining real-time inference capabilities and addressing the time-complexity limitations of Transformers. Despite this intended persistence, the memory mechanism of canonical SSMs is theoretically designed to decay monotonically over time, meaning that more recent inputs are expected to be retained more accurately than earlier ones. Contrary to this theoretical expectation, however, the present study reveals a counterintuitive finding: when trained and evaluated on a synthetic, statistically balanced memorization task, SSMs predominantly preserve the initially presented data in memory. This pattern of memory bias, known as the primacy effect in psychology, presents a non-trivial challenge to the current theoretical understanding of SSMs and opens new avenues for future research.

Utility-Guided Agreement Generation from Negotiation Dialogues Using Large Language Models

This study aims to develop a mechanism that automatically generates agreement texts and determines consensus in negotiation scenarios using a large language model (LLM) based on utility. While conventional LLMs excel at natural language generation, they have difficulty explicitly reflecting the individual participants’ utilities and preferences. In this research, each AI agent’s utility is predefined and provided as input to the LLM, which then generates an agreement text. Furthermore, an agreement evaluation module is designed to analyze the generated text and assess it based on the participants’ utilities. By enabling the interpretation of decision-making processes expressed in natural language, the proposed mechanism enhances transparency and contributes to realizing explainable AI–based support for consensus building.Is this conversation helpful so far?

Practical Uses of AI in Business

At our company, we set up an internal “AI Committee” to promote the use of AI in our company. This presentation shows some of our activities, such as trying AI-assisted coding, enabling AI-based access to internal documents, and testing tools that create meeting minutes automatically. Through these efforts, we are learning how generative AI can help with everyday work. Although these trials are still in progress, we hope to expand AI use across the company based on what we have learned.

Delivery Planning Optimization with Educational Considerations

With the growing demand for online supermarkets in recent years, the need for efficient and sustainable delivery planning has become increasingly critical. This study presents an optimization-based approach to delivery planning tailored for online supermarket operations. By leveraging mathematical programming techniques, we developed a vehicle assignment and routing system that minimizes driver waiting time by accounting for picking completion times, while simultaneously improving key performance indicators such as delivery distance, time, and number of deliveries. Furthermore, the system incorporates driver skill levels to assign appropriate delivery courses and supports gradual training for unfamiliar areas, enabling drivers to expand their coverage as they gain experience. A field experiment conducted in Mie Prefecture in FY2024 demonstrated a 7% increase in delivery volume, reduced labor costs, and shorter delivery times. The proposed method contributes to the standardization of delivery planning by reducing reliance on individual expertise and capturing operational know-how in a systematic framework.

Security Evaluation of Evasion Attacks against Edge AI Hardware

AI technology is widely used in various areas. However, threats of evasion attacks that intentionally mislead AI inferences have been reported. Therefore, security evaluations and countermeasures against the threats are important to realize AI utilization with safety and security. This study introduces evasion attack techniques to evaluate the security of edge AI hardware.

Blockchain-Based Knowledge Sharing System

Scientific research is often conducted through collaboration among multiple researchers and laboratories. However, many challenges remain in sharing, inheriting, and reusing the knowledge and data generated during the research process. In particular, findings that are not published in papers are easily lost and difficult to share or reuse. To address these issues, we developed a web application called Shizui vault, which enables research teams to easily share scientific knowledge and results. This system allows users to record research outputs at the smallest unit, with each submission encrypted and recorded on the SYMBOL blockchain. This mechanism makes it possible to prove the existence and timestamp of each submission, thereby enhancing the transparency and reliability of the research process. Shizui vault serves as a new foundation for knowledge sharing, designed to preserve and transmit small discoveries and ideas in academic research to the future.

Object material identification by the human visual system in virtual reality

Understanding how the human visual system identifies material properties is a central question in vision science. Traditional psychophysical studies have relied on passive viewing paradigms, where observers judge material qualities of images presented on a monitor from fixed distances and angles. In contrast, in everyday settings such as gemstone appraisal, people actively move, tilt, and manipulate objects to extract diagnostic visual cues. In this study, we address this gap by allowing observers to freely explore virtual objects within an immersive virtual reality (VR) environment. This approach enabled us to quantify how active observation behaviors contribute to material discrimination. We report that free exploration reveals distinct behavioral and perceptual signatures of material recognition, highlighting the importance of self-initiated motion and viewpoint change in constructing the perceptual representation of object materials.

Impact of Astigmatic Axis on Text Readability

Although it has been suggested that differences in astigmatic axis affect vision and reading at the same cylindrical power, the impact of differences in blur patterns on characters has not been examined. We developed a quantitative evaluation method for character blur using information entropy. We compared the measurement results of reading speed with artificially created astigmatic eyes and the evaluation results of the reading tasks used, showing a significant negative correlation.

Environmental understanding and representation for autonomous visual navigation

For a robot to autonomously reach a goal, it needs some prior knowledge about its environment. Humans are no different, even when visiting a place for the first time, we rely on a map to find our way. In this talk, I will discuss what challenges exist with conventional maps used for robot navigation, and share some of our research exploring what kind of environmental maps work best for robots that mainly rely on cameras for navigation.

Topological waveguides enabling robust and functional signal transport

Topological insulators rising from topology in quantum materials have been attracting attention as novel technology for controlling quantum and classical wave phenomena and diffusion phenomena. Our study focuses on low-loss and high-efficient wave propagation in topological insulators, and aims to apply them to signal and energy transportation systems. However, there are remaining challenges for active functionalities of the topological waveguide systems by external stimuli because topological waveguides have an immune to external noise and internal defects. In this presentation, we introduce the actively tunable topological couples by driving frequency. Further, we demonstrated the frequency-modulated digital data communication with them.

Manipulation and analysis of biomolecules using micro/nanostructures

Analysis of large DNA molecules is essential for both basic and applied research in biology, epidemiology, and medicine, particularly in studies involving bacteria and human cells. However, due to their enormous size and high flexibility, large DNA molecules are often difficult to analyze with high accuracy and speed using conventional techniques such as gel electrophoresis and column-based methods. In this study, we introduce methods developed by the authors that enable precise and rapid analysis of large DNA molecules by utilizing micro- and nanoscale structures. These approaches achieve molecular analyses by controlling the motion and deformation of DNA molecules through designed micro- and nanostructures. By optimizing the structural geometry and migration conditions, these methods allow fast and accurate analyses of large DNAs, including size-based separation, concentration, and single-molecule analysis.

Microfluidic flow sensor using on-chip pattern laser

Analysis of large DNA molecules is essential for both basic and applied research in biology, epidemiology, and medicine, particularly in studies involving bacteria and human cells. However, due to their enormous size and high flexibility, large DNA molecules are often difficult to analyze with high accuracy and speed using conventional techniques such as gel electrophoresis and column-based methods. In this study, we introduce methods developed by the authors that enable precise and rapid analysis of large DNA molecules by utilizing micro- and nanoscale structures. These approaches achieve molecular analyses by controlling the motion and deformation of DNA molecules through designed micro- and nanostructures. By optimizing the structural geometry and migration conditions, these methods allow fast and accurate analyses of large DNAs, including size-based separation, concentration, and single-molecule analysis.

Development of Next-Generation In-Body Monitoring Technologies Enabled by MEMS Wearable Devices

This study aims to develop next-generation in-body monitoring technology based on MEMS (Micro Electro Mechanical Systems) based wearable devices. It seeks to noninvasively acquire physiological information from the human body and enable the early detection of iron deficiency anemia. The proposed device can be attached to the skin like a band-aid and detects ferritin, a biomarker that indicates the body’s iron metabolism status, from body fluids such as sweat. This enables simple and continuous physiological monitoring without the need for conventional blood sampling. In this report, I report on the proposal, design, fabrication, and evaluation of a new MEMS device model developed to improve detection accuracy.

Analysis of Hair Type Characteristics Utilizing Bioinformatics Technologies

Human hair is typically classified into four major types—straight, wavy, curly, and coily—based on curl patterns, and further subdivided by physical traits such as thickness and texture. However, the molecular basis underlying these classifications, particularly in terms of protein composition, remains poorly understood. This study aims to elucidate the structural and molecular diversity of hair by integrating physical property measurements (e.g., diameter and tensile strength) with proteomic analysis. Hair samples were collected from individuals of diverse ethnic backgrounds, including Japanese, Korean, Turkish, European, and African descent. Using bioinformatics approaches, we analyze the relationship between hair morphology and protein expression profiles. This research is expected to provide new insights into the biological foundations of hair type variation and contribute to the development of personalized hair care solutions and ethnically tailored cosmetic or medical applications.

A Plant-Derived Secondary Metabolite Modulates Tumor Metabolism in Cancer Metastasis

Clinical proteomics of ovarian cancer ascites reveals proteomic subgroups and prognostic biomarkers

Ovarian cancer is known to disseminate directly within the peritoneal cavity through ascitic fluid and is one of the most lethal gynecologic malignancies. Malignant ascites contains not only cancer cells but also numerous non-tumor cells, including immune cells, which release proteases and their cytokine substrates. However, the proteolytic signaling landscape within ascites has not been systematically characterized. In this study, large-scale proteomic profiling of malignant ascites from 91 patients with high-grade serous ovarian carcinoma (HGSOC) was performed using a data-independent acquisition (DIA) approach. The proteomic signatures of ascitic fluid were evaluated to identify potential biomarkers and to explore the heterogeneity of proteolytic activity. Unsupervised analysis revealed three distinct molecular subgroups that exhibited significantly different levels of proteolytic activity. Furthermore, CoxBoost modeling identified biomarker candidates associated with patient prognosis. Enrichment analysis of these prognostic molecules highlighted coagulation and complement activation pathways as correlating with poor outcomes. These findings demonstrate the potential of ascitic proteome profiling as a valuable resource for understanding the tumor microenvironment and for identifying prognostic biomarkers in ovarian cancer.

Overcoming Transitional Women’s Total Health Care Problems by EV Analyses

Clonal hematopoiesis; A hidden mechanism underlying cardiovascular disease

Chaordic Homeodynamics in Cardiomyocytes: Constant Period Chaos at the Sarcomere Level

Cardiac beating arises from periodic sarcomere shortening. In isolated cardiomyocytes warmed to 37–42 °C, we show that sarcomeres undergo Ca²⁺‑independent autonomous oscillations in which amplitude and phase wander irregularly, while cell‑level beating is preserved and each sarcomere’s period remains nearly constant (CV ≈ 1%) due to tight period locking. Observations were obtained from single‑sarcomere length waves during beating. High‑speed sarcomere‑length nanometry (500 fps, ~3 nm precision) together with recurrence quantification and Lyapunov analyses yielded a positive maximal Lyapunov exponent and significant differences versus surrogate data, confirming deterministic—rather than random—chaos. I propose the concept “Chaordic Homeodynamics” for this coexistence of order and fluctuation. Spontaneous phase slips between neighboring sarcomeres smoothed the rise of cell‑level tension, stabilizing output. These findings articulate a homeostatic principle that reconciles robustness with adaptability and may inform early arrhythmia markers and the design of bio‑inspired myocardial constructs, quantifying how living systems actively harness microscopic fluctuations.

Cellular dynamics and roles of brain macrophages during development

The central nervous system (CNS), consisting of the brain and spinal cord, is composed not only of neural lineage cells but also of vascular and immune cells. The precise and dynamic interactions among these diverse cell types are essential for maintaining the brain’s complex functions. This study focuses on microglia, the brain-resident macrophages of the immune system. Microglia are indispensable for maintaining brain homeostasis, performing a wide range of functions such as clearing unnecessary cells, shaping neural circuits, and regulating inflammatory responses. Recent advances in single-cell transcriptomics have revealed that microglia are not a homogeneous population but consist of multiple subtypes with distinct gene expression profiles. However, the developmental processes through which this diversity arises remain largely unknown. This presentation introduces recent approaches that combine in vivo imaging of embryonic mouse brains, high-resolution tissue analysis, fate mapping, and single-cell transcriptomics to elucidate microglial colonization process and the mechanisms underlying the acquisition of their diversity during brain development.

Establishment of a high-throughput long-read analysis platform for large-scale disease research

Long-read sequencing provides superior resolution for detecting structural variants, repeat expansions, and epigenetic modifications, offering new insights into the genomics of rare and intractable diseases. To enable large-scale long-read analysis of over 1,000 individuals, we established a high-throughput platform equipped with two PacBio Revio sequencers. The system performs long-read whole-genome and full-length transcriptome sequencing, with analytical pipelines that preserve polymerase kinetics in BAM files for methylation analysis. Full-length transcriptome profiling is conducted using MAS-Iso-Seq. All computational workflows are containerized with Singularity and executed on a high-performance computing cluster composed of multiple GPU servers with large memory capacity. The infrastructure incorporates scalable, secure storage and centralized data management. The platform achieved maximum yields of 150 Gb per genome and 160 Gb per MAS-Iso-Seq sample, enabling highly sensitive detection of genetic variants and transcript isoforms, even from challenging samples. Methylation profiling identified allele-specific gene silencing and somatic mosaic repeat expansions, while comprehensive transcriptome analysis revealed splicing abnormalities in neurological disorders. All data are securely archived within dedicated systems, ensuring robust data integrity and compliance as the infrastructure continues to evolve. By establishing this integrated long-read omics platform, we have enabled advanced genomic and transcriptomic analyses for large-scale studies of rare and intractable diseases.

Prediction of Traumatic Brain Injury Based on Axonal Strain Using a Human Brain FE Model

Functional Profiling of the Gut Microbiome in Anxiety and Depression

This study investigates the link between the gut microbiome and mental health, focusing on anxiety and depression. Using the CES-D scale, we divided 17 participants into groups based on mood: a control group and a depression group. Through gut microbiome analysis, distinct bacterial species and metabolic functions were observed in the depression group. PICRUSt2 and LEfSe analyses identified specific bacterial species and functions associated with depression, suggesting a role for the gut microbiota in mental health disorders. The findings underscore the gut microbiome’s potential influence on mood disorders and suggest avenues for therapeutic interventions targeting microbial pathways to support mental well-being.

Bioproduction utilizing insects as living factories

In our modern society, most functional molecules surrounding us are produced through chemical synthesis in flasks, where numerous reactions and catalysts have been developed and utilized. However, the synthesis of molecules requiring strict regio- and stereo-selectivity, or high functional group specificity, often involves multistep procedures, leading to increased costs and waste. Thus, the development of new reactions and reaction systems remains an urgent challenge.Meanwhile, highly selective chemical transformations catalyzed by biological enzymes have independently evolved within the fields of biosynthesis and xenobiotic metabolism. These biological functions have long been treated as separate from organic synthetic chemistry, and their active utilization has been relatively limited. In this presentation, I will introduce our recent findings and discuss the potential of integrating biological catalysis into resource-circulating molecular production systems.

Insect control through interventions in group formation behavior

Group formation is a crucial survival strategy widely observed across animal species. Understanding the mechanisms underlying group formation is important not only for basic biology but also for potential practical applications such as pest control. Even among closely related species, some form groups while others do not, suggesting that group behavior has evolved during the process of speciation. Our previous studies have revealed that among fruit flies (Diptera: Drosophilidae), there are species that actively form groups and others that rarely do so. Using the model species Drosophila melanogaster, we are now identifying specific neural circuits in the brain that regulate group formation. In this presentation, I will discuss the neural mechanisms underlying group formation in insects and explore how such understanding may contribute to novel approaches for pest control.

Targeting the mosquito circadian clock to disrupt hearing function and behaviors

Aiming at elucidation of higher-order visual system in fish and development of eel larval feed

We try to elucidate the higher-order visual system in fish and aim at development of feed for eel larvae by means of neuroscientific approaches to contribute to the fisheries industries through more efficient catches and aquacultural production of fish. We investigated the visual circuits and their functions of fish and found that fish have a higher-order visual system similar to that of mammals. The visual circuits and functions of a fish species that is expected to have higher cognitive abilities is also being revealed through international collaborative researches with the CNRS in France. 　　 In addition, we revealed the development of taste organs as eel larvae grow and the brain regions activated by feed stimuli. Furthermore, we are performing feeding experiments using feed added with growth promoting substances on zebrafish larvae to aim at high growth of fish.

Does miscommunication between humans and animals pose a risk to their interaction?

During close encounters between humans and animals, each species relies on its own communicative signals. However, same signals can carry different meanings across species. This is true even when comparing humans and non-human primates, which are closely related in evolutionary terms. When humans anthropomorphize primate behavior—interpreting their expressions through a human emotional lens—they may unintentionally display signals that primates perceive as threatening or intrusive. Our research indicates that humans are strongly motivated to interact with primates. Yet, close proximity not only increases aggression from the animals but also raises the risk of disease transmission. To mitigate these interactions, we need to examine the signals exchanged between different species, and how these are shaped by human cultures. Collaborating with researchers in Japan and the United States, we ought to build a more culturally and biologically informed approach to the study of human-animal interactions.

Understanding mechanisms of plant acclimation to nutritional stress

Between soil impoverishment and global warming, the instability of agricultural yields is constantly increasing. It is now crucial to develop less polluting and more sustainable agricultural methods by improving our understanding of the natural adaptation of plants to their environment. Plants can remember some environmental stresses. Nitrogen nutrition is a major environmental factor that regulates the growth and productivity of crops, but it is unclear whether plants remember nutritional stress. My results strongly suggest that plants can remember nitrogen deficiency stress and react. Understand its mechanisms will permit to imprint the memory, during seedling production process, to create seedlings with both high yield and stress tolerance.

Pioneering High-redshift Cosmology

Reconstruction of the extreme solar and geomagnetic storm in Feb 1872

Intense solar eruptions occasionally accelerate charged particle to high energies. Such solar energetic particles (SEPs) can reach the Earth, depending on the relative geometry between the Sun and the Earth. Sometimes, SEPs have sufficiently high energy and flux to be detected by ground-level detectors known as ground-level enhancements (GLEs). Among such GLEs, GLE #5 on 23 Feb 1956 is associated with the hardest spectra and greatest flux and have been used for an assessment benchmark for the space weather hazards. However, we know little about their source data. Few knew where their source records were gone. We have recently located most of their source records and significantly revised reconstructions for the maximal fluence, spectrum hardness, and temporal evolution. Our result indicates a softer spectrum and increases the maximal fluence by 10% in contrast with the previous studies. The new datasets significantly revise the source datasets, improve the time-series accuracy, and form basis for further discussions on their spectra.

Reconstructing the Linguistic Silk Road

The ancient Silk Road was not only a route for the exchange of goods and religions, but also a vital network for the transmission of languages and scripts. This presentation introduces an attempt to reconstruct the “Linguistic Silk Road” through a study of twenty Buddhist dictionaries (known as yinyi) compiled across East Asia between the 7th and 13th centuries. These works are more than mere glossaries—they represent the intellectual crystallization of intercultural understanding and translation, serving as bridges between Eastern and Western systems of knowledge. By constructing and analyzing the Database of Ancient Buddhist Dictionaries, this study explores how people in antiquity comprehended foreign languages and incorporated them into their own linguistic systems, offering new perspectives that connect the humanities and digital scholarship.

Analysis of the Contribution of Urban Law to Climate Change

Climate change is no longer a problem confined to specific regions, but has become a common challenge for all humanity. Beyond rising temperatures, there is a consistent trend of increased frequency and intensity in extreme weather events, such as fluctuations in annual rainfall and sudden torrential downpours. In particular, measures to “adapt” to increasingly severe flooding have become an urgent issue directly linked to the survival of cities. In recent years, the (Japanese) government has championed the concept of “watershed flood control”. This approach encourages collaboration not only among river authorities but also with watershed residents, local governments, and businesses. It promotes a multi-layered strategy encompassing flood prevention, damage mitigation, and rapid recovery. To realise this, amendments to the Urban River Flood Damage Countermeasures Act introduced measures such as promoting the development of rainwater retention functional preservation zones and rainwater retention facilities. However, within Japan's urban planning legal framework, such climate change “adaptation” measures are not sufficiently integrated into urban planning. While the Climate Change Adaptation Act encourages local authorities to formulate regional plans, its linkage with the Urban Planning Act and Building Standards Act is weak. Consequently, climate change countermeasures and decarbonisation efforts are rarely reflected in decisions regarding actual land use, development activities, and their permitting. Furthermore, the installation of rainwater storage and infiltration facilities, greening, and similar measures are not mandatory, except under ordinances in some local authorities. This study employs comparative legal research, a traditional legal scholarship methodology, to examine Germany's urban legal system and derive implications for Japanese law. Germany's Baugesetzbuch (equivalent to Japan's Urban Planning Act) explicitly mandates consideration of climate protection and adaptation to climate change within land use planning. Furthermore, its legal framework enables specific provisions in urban planning, such as designating green belts, restricting impervious surfaces, and locating rainwater storage facilities and renewable energy installations. Thus, climate change adaptation within urban space is institutionally embedded in Germany's urban legal system. This research aims to elucidate the characteristics of Germany's urban legal framework and derive implications, such as the necessity of mandating the integration of climate change adaptation with urban planning. Through a comparative legal approach, it seeks to reconfigure environmental, disaster prevention, and urban planning considerations in a cross-cutting manner, thereby creating urban legislation resilient to climate change.

The Relationship of Entrepreneurship Education (EE) and Entrepreneurial Intention in University