Assessment of Multidimensional Health Care Parameters for Developing a Virtual Human Generative Model: A Cross-Sectional Study in Japan

    June 2023 in “ JMIR Research Protocols
    Masanobu Hibi, Shun Katada, Aya Kawakami, Kotatsu Bito, Mayumi Otsuka, Kei Sugitani, Adeline Muliandi, Nami Yamanaka, Takahiro Hasumura, Yasutoshi Ando, Takashi Fushimi, Teruhisa Fujimatsu, Tomoki Akatsu, Sawako Kawano, Ren Kimura, Shigeki Tsuchiya, Yuki Yamamoto, Mai Haneoka, Ken Kushida, Tomoki Hideshima, Eri Shimizu, Jumpei Suzuki, Aya Kirino, Hiroji Tsujimura, Shun Nakamura, Takashi Sakamoto, Yuki Tazoe, Masayuki Yabuki, Shinobu Nagase, Reiko Fukuda, Yukari Yamashiro, Nobutoshi Ojima, Motoki Sudo, Naoki Oya, Yoshihiko Minegishi, Kouichi Misawa, Nontawat Charoenphakdee, Zhengyan Gao, Kohei Hayashi, Kenta Oono, Yohei Sugawara, Shoichiro Yamaguchi, Takahiro Ono, Hiroshi Maruyama
    Image of study
    TLDR The study aims to create a model to predict health attributes using diverse health data from Japanese adults.
    A cross-sectional study was conducted on 1000 adults in Japan, aiming to develop a Virtual Human Generative Model (VHGM) using comprehensive health data. The data included biochemical and metabolic profiles, bacterial profiles, messenger RNA, proteome, and metabolite analyses, lifestyle surveys, physical, motor, cognitive, and vascular function analyses, alopecia analysis, and body odor components. The VHGM will be constructed using a heterogeneous incomplete variational autoencoder (HIVAE) and evaluated based on its ability to predict missing attributes and represent the empirical distribution. The study's findings may reveal unknown relationships among variables, potentially leading to new health care hypotheses and interventions. However, limitations include potential selection bias and the inability to generalize findings to the entire population. The primary sponsor of this study is Kao Corporation (Tokyo, Japan).
    Discuss this study in the Community →

    Cited in this study

    5 / 5 results