I am a Ph.D. student in the Department of Computer Science & Engineering, Lehigh University, where I am advised by Prof. Lichao Sun. I was a research intern at the AI&I Department, Mayo Clinic , where I am developing vision-language pre-training and federated learning algorithms for electronic health records (EHR).

My recent research focuses on Multi-modal and Federated Learning for Biomedicine, as well as efficient machine/deep learning system design, with the goal of Accelerating Discovery to Delivery for Clinical Excellence. Ultimately, my career goal is to develop trustworthy and reliable artificial general intelligence (AGI) systems for computational precision medicine and health.

I have published approximately 20 research papers with total .

🔥 News

Nature Medicine
sym

BiomedGPT: A Generalist Vision-Language Foundation Model for Diverse Biomedical Tasks

Kai Zhang, Rong Zhou, Eashan Adhikarla, Zhiling Yan, Yixin Liu, Jun Yu, Zhengliang Liu, Xun Chen, Brian D. Davison, Hui Ren, Jing Huang, Chen Chen, Yuyin Zhou, Sunyang Fu, Wei Liu, Tianming Liu, Xiang Li, Yong Chen, Lifang He, James Zou, Quanzheng Li, Hongfang Liu, Lichao Sun

  • TL;DR: We introduce a unified and generalist Biomedical Generative Pre-trained Transformer (BiomedGPT) model, which leverages self-supervision on large and diverse datasets to accept multi-modal inputs and perform a range of downstream tasks.
  • Source Code: BiomedGPT .
FLSys 2023
sym

Memory-adaptive Depth-wise Heterogenous Federated Learning

Kai Zhang, Yutong Dai, Hongyi Wang, Eric Xing, Xun Chen, Lichao Sun

  • TL;DR: We introduce a memory-adaptive depth-wise learning solution in FL, which adaptively decomposes the full model into blocks according to the memory budgets of each client and trains blocks sequentially to obtain a full inference model.
  • Source Code: FeDepth .
EMNLP 2022
sym

Efficient Federated Learning on Knowledge Graphs via Privacy-preserving Relation Embedding Aggregation

Kai Zhang, Yu Wang, Hongyi Wang, Lifu Huang, Carl Yang, Xun Chen, Lichao Sun

  • TL;DR: Our framework allows for sharing entity embeddings of knowledge graphs across multiple clients while protecting privacy to prevent any potential leakage.
  • Source Code: FedR .