Menu

Copyright © Satoru Hiwa All rights reserved | This template is made with by Colorlib

Researches

国内企業の方へ

当方の研究内容にご興味をお持ちいただいた方で、関連技術でお困りの場合、技術相談が可能です。数値最適化（学位取得）、脳活動を始めとした生体情報計測と最適化・機械学習に基づいた解析（現職での主研究テーマ）をはじめ、企業勤務時代に自動車や電池・電源・キャパシタ等の制御アルゴリズム開発、数理モデル化や数値シミュレーション、最適化アルゴリズムに基づく最適設計を実務経験しておりますので、製品開発や新規事業開発を意識した相談が可能です。まずはお気軽に当方までメールにてご連絡ください。内容に応じて、共同研究契約や技術相談の形態、費用を相談させていただきます。

例）

  • 製造工程や製品設計に存在するパラメータを計測データに基づいて最適な値に設定したい。
  • システムや製品に対して、生体情報に基づいた評価を行いたい。
  • 産学連携のテーマを模索している。

    • Cognitive and Computational Neuroscience

      Neuroscience of Creativity

      Supported by JSPS KAKENHI Grant Number JP24K03028, Grant-in-Aid for Scientific Research (B)

      Creativity is one of the most important non-cognitive skills that will be required of the children who will be living in the future. Creative thinking processes can be divided into divergent and convergent thinking, but it remains unclear how these two thinking processes should be trained to promote creativity and what characteristics highly creative individuals possess. This study aims to clarify the dependence of brain structure and function on creativity. The influence of the functional/structural brain network on creativity in divergent and convergent thinking, as well as the dependence between brain function and structure, will be revealed through the analysis of machine learning models of the relationship between brain and creativity traits. Besides, we develop the generative model-based data augmentation method to enhance the sample size from limited datasets for building above-mentioned machine-learning models. These results can be applied to the development of educational and intervention methods to promote creativity.

      See also:

    • T. Yamamoto, T. Sugiura, T. Hiroyasu and S. Hiwa*, "Task-guided Generative Adversarial Networks for Synthesizing and Augmenting Structural Connectivity Matrices for Connectivity-Based Prediction," bioRxiv, 2024.02.13.580039; doi: 10.1101/2024.02.13.580039 (Feb 2024, in review).

      • Topological data analysis on neuroimaging data

      Most of cognitive neuroscience studies have mainly taken the hypothesis-driven approach, however, neuroimaging data is extremely high-dimensional, that is why the data-driven approach would be effective. Topological data analysis (TDA), which visualize and analyze features of dataset focusing on the 'shape' of the dataset by representing it in forms of networks or graphs, is recently applied to various fields of data analysis. We apply TDA framework to cognitive neuroscience study instead of using hypothesis testing to enable data-driven knowledge discovery. The network topology of the neuroimaging dataset and its relations with the behavioral and psychological metrics will give new insights on human cognition and behavior. We also aim to develop new TDA method to emphasize them.

      See also:

    • S. Isojima, K. Tanioka, T. Hiroyasu, S. Hiwa*, "Preliminary Investigation of the Association Between Driving Pleasure and Brain Activity with Mapper-based Topological Data Analysis," International Journal of Intelligent Transportation Systems Research, 2023, doi: 10.1007/s13177-023-00371-3 (Sep 2023).

      • [Finished] Quantifying Mindfulness Based on Functional Brain Network

      Supported by JSPS KAKENHI Grant Number JP19K12145, Grant-in-Aid for Scientific Research (C)

      Mindfulness meditation, which is defined as nonjudgmental observation of one's current experiences such as emotion, thoughts and sensations inside and around them, is a key practice to promote our human well-being. However, it is difficult for novices to evaluate how correctly their meditation is performed. To overcome this issue, we aim to characterize the brain state during meditation based on fMRI data. Recent studies have revealed that meditation could affect brain plasticity. Furthermore, the brain activation and deactivation patterns have been found to differ between different meditation styles. However, the functional network structures of the various meditation styles have not yet been established because of the heterogeneity of conditions across studies or the diversity of meditation practices. Here we propose a novel data-driven approach to find the specific functional network organization associated with meditation using an evolutionary optimization algorithm. Moreover, the network structure derived by our method is used to quantify how well the practitioners can meditate.

    • Human-in-the-loop System

      Measuring and Visualizing Driving Pleasure

      This study will refer to the positive feeling of "fun" derived from driving a car as a driving pleasure. The central question of this study is, "Under what circumstances does driving pleasure occur while driving?" Numerous surveys have been conducted on automobile usage and attitudes toward automated driving, and there is concern about the loss of driving pleasure due to the development of automated driving technology. However, the factors that induce driving pleasure need to be clarified. We believe that if we can explain the conditions that cause driving pleasure and visualize its occurrence, it will lead to the development of advanced driver assistance and automated driving technologies that make the driver's psychological state more positive, thereby expanding the potential of automobiles as a means to enhance human well-being. In this research, we aim to clarify the conditions under which driving pleasure occurs and to develop technology to visualize its occurrence and disappearance based on biometric and vehicle information.

      See also:

    • S. Isojima, K. Tanioka, T. Hiroyasu, S. Hiwa*, "Preliminary Investigation of the Association Between Driving Pleasure and Brain Activity with Mapper-based Topological Data Analysis," International Journal of Intelligent Transportation Systems Research, 2023, doi: 10.1007/s13177-023-00371-3 (Sep 2023).

      • Development of Mindful Driving System

      Supported by FY2019 Strategic Information and Communications R&D Promotion Programme (SCOPE)

      Distracted driving is a major cause of traffic accidents and injuries. We aim to promote 'mindful driving' by detecting and quantifying the degree of driver's distraction while driving, using a multimodal measurement of human behavior (e.g., operations of steering wheel, acceleration and deceleration pedals) and biological information including brain activity, heartbeat and so on. The prediction model for driver's distraction are modeled through machine learning on these multimodal data and then implemented on the vehicle.

      See also:

    • T. Ogihara, K. Tanioka, T. Hiroyasu, S. Hiwa*, "Predicting the Degree of Distracted Driving Based on fNIRS Functional Connectivity: A Pilot Study," Frontiers in Neuroergonomics, Volume 3, Article 864938, doi: 10.3389/fnrgo.2022.864938 (Jul 2022).
    • 「ACADEMIC EXPERTS/同志社大学大学院 生命医科学研究科医工学・医情報学専攻 ヒューマンインフォマティクス研究室 日和 悟 准教授」 News Letter from ITS Japan (2022.12.21)
    • 戦略的情報通信研究開発推進事業（SCOPE）「ヒトと自動車のマルチモーダル計測に基づくマインドフル・ドライビングシステムの開発」 研究開発成果と使いたい人をマッチング/近畿総合通信局 (2022.2)

    • Fashion Research

      Analysis of Fashion Design Based on Informatics and Neuroscience (under planning)

      I am interested in fashion and clothing, but I am neither a designer nor an artist. Here, I use informatics and neuroscience to extract from images or photographs the abstract concepts, contexts, and messages that designers put into their clothes. I welcome designers and people in the clothing industry interested in collaborating on this theme.