Project Name: HANAMI – Heart sound Analysis and its Non-invasive healthcare Applications via Machine Intelligence.
Supporter: Zhejiang Lab, P. R. China (Acceptance Rate: < 15%).
Run Time: 17.08.2019 – 17.08.2020.
Role: Author of Proposal, Principal Investigator, Grant Holder
Partners: The University of Tokyo, Shenzhen University General Hospital, Imperial College London, Carnegie Mellon University, University of Augsburg.
Abstract: This project aims to investigate the feasibility of using machine listening for automatic analysing the heart sound recorded from the smart stethoscope. Firstly, we will collect, establish, and release a publicly accessible heart sound database, which will overcome the challenge of lacking comparable and suistainable open source database in relevant study. Secondly, we will make a comprehensive investigation of the fundamental knowledge (e.g., features, machine learning models, and relationship between the pathological heart sounds and their acoustical properties). Thirdly, both of the traditional machine learning approaches and the state-of-the-art deep learning methods will be studied and compared based on the released database. We hope the studies can attract more attentions from the scientific community of machine learning, signal processing, cardiology, and biomedical engineering.
Project Name: Deep Learning for Intensive Longitudinal Biomedical Signals and its Health Related AI Applications.
Supporter: Japan Society for the Promotion of Science (JSPS), Japan (Acceptance Rate: 10.6%).
Run Time: 30.09.2019 – 29.09.2021.
Role: Co-Author of Proposal, Co-Principal Investigator, Co-Grant Holder.
Partners: The University of Tokyo, Imperial College London, Osaka University, Carnegie Mellon University, University of Augsburg, Shizuoka University.
Abstract: This research aims to leverage the power of A.I. for analysing and monitoring the daily behaviour of the patients suffering from psychiatric diseases via the biomedical intensive longitudinal data. We will investigate the state-of-the-art techniques of machine learning, deep learning, and signal processing for their capacity on screening the patients from the healthy control. In addition, we will explore the feasibility to use the paradigm of A.I. to implement an automatic monitoring and evaluation system for subject’s health status by IoT sensor data. The achievements of this research can facilitate the development of smart wearables for building a human-centered A.I. world with a plenty of applications on healthcare and wellbeing.
Project Name: Development of Just-in-Time Adaptive Intervention for Behavioural Modification based on Continuous Psycho-behavioural Monitoring under Daily Life.
Supporter: Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan (Acceptance Rate: 24.8%).
Run Time: 01.04.2017 – 31.03.2020.
Role: Main Participant.
Partners: The University of Tokyo, Fujita Health University, Nagoya City University.
Abstract: The purpose of this study is to develop a risk detection type intervention guidance method that conducts behavioural change intervention in Just-in-Time, and to verify its clinical applicability.