Hongyu He

Hongyu He

PhD Student & Research Assistant

National University of Singapore

About Me

I am a PhD student in Biomedical Engineering at the National University of Singapore (NUS), where I began my doctoral studies in August 2024. My academic journey spans across multiple prestigious institutions and diverse engineering disciplines, providing me with a unique interdisciplinary perspective that bridges electrical engineering, biomedical engineering, and artificial intelligence.

My research passion lies at the intersection of artificial intelligence and biomedical sciences, where I am developing innovative computational approaches to accelerate scientific discovery. I believe that AI has the potential to revolutionize how we understand biological systems and develop new therapeutic solutions.

Before starting my PhD, I gained valuable research experience at Duke University, where I worked as a Research Associate focusing on medical imaging and machine learning applications. My diverse background includes work on privacy-preserving federated learning, deepfake detection, and advanced medical image analysis.

Research Focus

  • Use self-evolving AI with improved creativity to help make new discoveries in biomedical sciences
  • Improve creativity of generative machine learning models and enable them to propose novel contents, focusing on post-training of LLMs and fundamental mechanisms of diffusion

Education

PhD in Biomedical Engineering

National University of Singapore (NUS)
August 2024 - Present
Research interests: AI for biomedical science, Generative model collapse, Increase LLM creativity

Master of Science in Electrical & Computer Engineering

Duke University, Durham, North Carolina
August 2020 - May 2022
Research interests: Machine Learning and Data Analytics

Bachelor of Engineering in Electrical Engineering

Xi'an Jiaotong University, Xi'an, Shaanxi, China
August 2016 - June 2020
Tsien Hsue-shen Class for Top Students

Professional Experience

Associate in Research

Yong Loo Lin School of Medicine, National University of Singapore
February 2024 - Present
Conducting advanced research at the intersection of AI and biomedical sciences, focusing on automated discovery systems and generative AI applications in healthcare.

Associate in Research

Department of Radiology, Duke University Medical Center
July 2022 - October 2023
Human Bone Project: Implemented a multi-label classification model that predicts both bone type and MRI series type (ax/sag/cor); Annotated MRI bone images for the training of bone segmentation model; Implemented 3D bone segmentation model.

Body Composition Project: Predict disease risk using CT-based body composition analysis.

Publications

Automated Discovery of Age-Related Retinal Biomarkers Using Deep Learning Framework

Hongyu He, Dianbo Liu
ARVO 2025 (Poster Session)

SuperMask: Generating High-resolution object masks from multi-view, unaligned low-resolution MRIs

Hongyu He, Hanxue Gu, Roy Colglazier, Jordan Axelrod, Robert French, Maciej A Mazurowski
MIDL 2023

Faceguard: Proactive deepfake detection

Yuankun Yang, Chenyue Liang, Hongyu He, Xiaoyu Cao, and Neil Zhenqiang Gong
2021 (preprint available on arXiv)

Stochastic Channel-Based Federated Learning With Neural Network Pruning for Medical Data Privacy Preservation: Model Development and Experimental Validation

Rulin Shao, Hongyu He, Ziwei Chen, Hui Liu, Dianbo Liu
JMIR Form Res 2020;4(12):e17265
DOI: 10.2196/17265

Stochastic Channel-Based Federated Learning for Medical Data Privacy Preserving

Rulin Shao, Hongyu He, Hui Liu, Dianbo Liu
Machine Learning for Health at NeurIPS, 2019

OpenHI2 – Open source histopathological image platform

Pargorn Puttapirat, Haichuan Zhang, Jingyi Deng, Yuxin Dong, Jiangbo Shi, Hongyu He, Zeyu Gao, Chunbao Wang, Xiangrong Zhang, Chen Li
2019 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

Contact Information