Trustworthy AI for Healthcare

Xiaoyang Wang (王晓阳) Shawn Preferred name: Shawn

I’m a PhD candidate at Drexel University, advised by Prof. Christopher C. Yang in the Health Informatics Research Lab.

Before joining Drexel, I earned my M.S. from the University of Pittsburgh, advised by Prof. Peter Brusilovsky, and my B.E. from Shanghai Normal University.

• Philadelphia, USA
• Email
• Google Scholar
• GitHub
• LinkedIn
• ORCID
• X

Research

News

EMNLP 2026

The Consistency Illusion: How Multi-Agent Debate Hides Reasoning Misalignment

Xiaoyang Wang, Christopher C. Yang.

Paper GitHub BibTeX

• This work introduces CARA (Cross-Agent Reasoning Alignment), a family of metrics that tests whether multi-agent LLMs which agree on an answer also share compatible reasoning, revealing a consistency illusion in which debate suppresses agent contradictions while their reasoning chains grow less similar.
• It proposes the Grounded Debate Protocol (GDP), a lightweight prompt-level intervention that substantially improves cross-agent reasoning alignment across two medical QA benchmarks and two model backbones without adding any LLM calls.

IEEE TIFS

FakeBench: Probing Explainable Fake Image Detection via Large Multimodal Models

Yixuan Li, Xuelin Liu, Xiaoyang Wang, Bu Sung Lee, Shiqi Wang, Anderson Rocha, and Weisi Lin.

Paper GitHub BibTeX

• This work introduces FakeBench, a multimodal benchmark designed to evaluate large multimodal models (LMMs) on explainable fake image detection rather than simple binary classification.
• The benchmark incorporates a fine-grained taxonomy of generative visual forgeries and human-in-the-loop textual descriptions to assess detection, reasoning, interpretation, and detailed forgery analysis.

Pattern Recognition

DeepSelective: Interpretable Prognosis Prediction via Feature Selection and Compression in EHR Data

Ruochi Zhang, Qian Yang, Xiaoyang Wang, Tian Wang, Qiong Zhou, Ziqi Deng, Kewei Li, Yueying Wang, Yusi Fan, Jiale Zhang, Lan Huang, Chang Liu, Fengfeng Zhou.

Paper Code BibTeX

• This work proposes DeepSelective, a novel end to end deep learning framework for predicting patient prognosis using EHR data, with a strong emphasis on enhancing model interpretability.
• DeepSelective combines data compression techniques with an innovative feature selection approach, integrating custom-designed modules that work together to improve both accuracy and interpretability.

Journal of Healthcare Informatics Research

MediHive: A Decentralized Agent Collective for Medical Reasoning

Xiaoyang Wang, Christopher C. Yang.

Paper BibTeX

• MediHive is a decentralized multi-agent framework for medical question answering, in which LLM-based agents self-assign specialized roles, resolve disagreements through conditional evidence-based debate, and iteratively fuse peer insights via a shared memory pool to reach consensus without any central coordinator.

@article{wang2026consistencyillusion,
  title   = {The Consistency Illusion: How Multi-Agent Debate Hides Reasoning Misalignment},
  author  = {Wang, Xiaoyang and Yang, Christopher C.},
  journal = {arXiv preprint arXiv:2606.08457},
  year    = {2026}
}

@article{li2025fakebench,
  title   = {FakeBench: Probing Explainable Fake Image Detection via Large Multimodal Models},
  author  = {Li, Yixuan and Liu, Xuelin and Wang, Xiaoyang and Lee, Bu Sung and Wang, Shiqi and Rocha, Anderson and Lin, Weisi},
  journal = {IEEE Transactions on Information Forensics and Security},
  year    = {2025},
  doi     = {10.1109/TIFS.2025.3597211}
}

@article{zhang2026deepselective,
  title   = {DeepSelective: Interpretable Prognosis Prediction via Feature Selection and Compression in EHR Data},
  author  = {Zhang, Ruochi and Yang, Qian and Wang, Xiaoyang and Wang, Tian and Zhou, Qiong and Deng, Ziqi and Li, Kewei and Wang, Yueying and Fan, Yusi and Zhang, Jiale and Huang, Lan and Liu, Chang and Zhou, Fengfeng},
  journal = {Pattern Recognition},
  year    = {2026},
  doi     = {10.1016/j.patcog.2025.112970}
}

@article{wang2026medihive,
  title   = {MediHive: A Decentralized Agent Collective for Medical Reasoning},
  author  = {Wang, Xiaoyang and Yang, Christopher C.},
  journal = {Journal of Healthcare Informatics Research},
  year    = {2026},
  doi     = {10.1007/s41666-026-00239-7}
}

@inproceedings{wang2026strategy,
  title     = {Strategy-Guided Multi-Agent Debate for Medical Question Answering},
  author    = {Wang, Xiaoyang and Yang, Christopher C.},
  booktitle = {AMIA Annual Symposium Proceedings},
  year      = {2026},
  note      = {To appear}
}

@inproceedings{wang2025moehealth,
  title     = {MoE-Health: A Mixture of Experts Framework for Robust Multimodal Healthcare Prediction},
  author    = {Wang, Xiaoyang and Yang, Christopher C.},
  booktitle = {Proceedings of the 16th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (ACM BCB)},
  year      = {2025}
}

@inproceedings{lee2025automated,
  title     = {Automated Clinical Problem Detection from SOAP Notes using a Collaborative Multi-Agent LLM Architecture},
  author    = {Lee, Yeawon and Wang, Xiaoyang and Yang, Christopher C.},
  booktitle = {Proceedings of the 16th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (ACM BCB)},
  year      = {2025}
}

@inproceedings{wang2025multiattribute,
  title     = {Enhancing Multi-Attribute Fairness in Healthcare Predictive Modeling},
  author    = {Wang, Xiaoyang and Yang, Christopher C.},
  booktitle = {Proceedings of the 13th IEEE International Conference on Healthcare Informatics (ICHI)},
  year      = {2025}
}

@inproceedings{wang2025balancing,
  title     = {Balancing Fairness and Performance in Healthcare AI: A Gradient Reconciliation Approach},
  author    = {Wang, Xiaoyang and Yang, Christopher C.},
  booktitle = {Artificial Intelligence in Medicine (AIME)},
  year      = {2025},
  doi       = {10.1007/978-3-031-95838-0_47}
}

@inproceedings{chang2024explainable,
  title     = {Explainable AI for Fair Sepsis Mortality Predictive Model},
  author    = {Chang, Chia-Hsuan and Wang, Xiaoyang and Yang, Christopher C.},
  booktitle = {Artificial Intelligence in Medicine (AIME)},
  year      = {2024}
}

@inproceedings{lucas2024explainablefair,
  title     = {An ExplainableFair Framework for Prediction of Substance Use Disorder Treatment Completion},
  author    = {Lucas, Mary M. and Wang, Xiaoyang and Chang, Chia-Hsuan and Yang, Christopher C.},
  booktitle = {Proceedings of the 12th IEEE International Conference on Healthcare Informatics (ICHI)},
  year      = {2024}
}

@article{wang2024achieving,
  title   = {Achieving Equity via Transfer Learning with Fairness Optimization},
  author  = {Wang, Xiaoyang and Chang, Chia-Hsuan and Yang, Christopher C.},
  journal = {IEEE Access},
  year    = {2024},
  doi     = {10.1109/ACCESS.2024.3519465}
}

Background

Academic Service

• Journal Reviewer: JAMIA, J Healthc Inform Res, IEEE TIFS, IP&M, ACM HEALTH, MedIA, and Comput Electr Eng.
• Conference Reviewer: NeurIPS 2024, AAAI 2025, ICLR 2025, ICWSM 2025, IEEE ICHI 2025 & 2026, WWW 2026, IEEE ICME 2026, and Digital Twins for Health Society (DT4HS).

Teaching Experience

Visitor Map

Hello, world! 🌍

Thanks for visiting. I am open to collaborations on trustworthy AI for healthcare, multimodal learning, and multi-agent systems. Feel free to get in touch.

• Drexel University, Philadelphia, PA
• xw388@drexel.edu
• Locations via MapMyVisitors · Privacy