DHAI Lab


The Wang Research Group (Data-Driven & Human-Centered AI in Healthcare and Medical Research [DHAI] Lab) works on developing and applying Machine Learning, Data Mining and Statistical methods on healthcare and (bio)medical data. Inspired by the human factors approach, DHAI lab also designs and develops Human-Centered Artificial Intelligence algorithms and tools for users to integrate, visualize, analyze, and interpret health data in order to improve the interoperability and accessibility of AI-assisted healthcare decision support.

Research Thrusts

Exploring the frontiers of Human-Centered AI

Interactive & Explainable Survival Analysis

Interactive & Explainable Survival Analysis

Developing advanced time-to-event prediction algorithms with human-in-the-loop explanations. Our platform enables trustworthy survival analysis for equipment failure prediction in IoT systems, customer churn analytics in business intelligence, cybersecurity risk forecasting, and patient outcome modeling in healthcare settings.

Digital Twin Powered Decision Making AI

Digital Twin Powered Decision Making AI

Leveraging reinforcement learning with online adaptation to create personalized decision-making policies powered by digital twin technology. Applications include smart building energy optimization, inventory and supply chain management, personalized learning paths in education, and adaptive treatment planning in healthcare.

Heterogeneous Graph Neural Networks

Heterogeneous Graph Neural Networks for Evolving Multimodal Data

Building sophisticated graph neural network architectures that integrate relational attributes with topological structures over time to analyze complex, evolving multimodal data. Key applications include brain disease detection, understanding metabolic mechanisms, molecular interaction analysis, and accelerating drug discovery processes.

LLM- and RAG-Enabled Multi-Agent AI Systems

LLM & RAG-Enabled Multi-Agent AI Systems

Creating intelligent multi-agent systems powered by large language models and retrieval-augmented generation for enhanced interaction and knowledge processing. Applications span enterprise knowledge assistants, scientific research copilots, business workflow automation, and clinician-facing question-answering systems in healthcare.

Ethical & Privacy Preserving AI Systems

Ethical & Privacy Preserving AI Systems

Designing and implementing privacy-first AI inference systems with zero-retention architecture for trustworthy deployment. Our work encompasses privacy-preserving analytics, secure model deployment, federated learning for IoT devices, and protected health data pipelines that ensure ethical AI practices while maintaining high performance.

Latest Papers

Xinyu Qin, Anil K. Sood, Ruiheng Yu, Sara Corvigno, Elaine Stur, Lu Wang (2026). A Treatment Response Optimized Clinical Decision Support AI System via Digital Twin Simulation. Manuscript submitted to the Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2026).

Xinyu Qin, Vicky Ye, Ruiheng Yu, Lu Wang (2026). Can Physician Expertise Improve Machine Learning Identification of Delirium?. Manuscript submitted to the Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2026).

Xinyu Qin, Martin Katzman, Alexandria Greifenberger, Elssa Toumeh, Sachinthya Lokuge, Tia Sternat, Ruiheng Yu, Lu Wang (2026). Estimating Treatment Effects for Depression in Longitudinal Therapy Switching Settings. Manuscript submitted to the Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2026).

Ruiheng Yu, Martin Katzman, Alexandria Greifenberger, Elssa Toumeh, Sachinthya Lokuge, Tia Sternat, Xinyu Qin, Lu Wang (2026). Interpretable Multimodal Depression Screening on Cam-CAN via fMRI Connectivity Graph Neural Networks and Clinical Measures. Manuscript submitted to the Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2026).

Ruiheng Yu, James H. Bray, Yunqi Liu, Xinyu Qin, Lu Wang (2026). Privacy-First Multi-Agent AI Powered Virtual Counselor for SBIRT: Integrating Large Avatar Models with Zero-Retention Architectures. Manuscript submitted to the Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2026).

See all