Research

Research interests and programs

Medical and biomedical AI form the foundation of our research. We develop reliable learning systems for multimodal data, scientific discovery, clinical decision support, and visual and spatial reasoning.

Multimodal and vision–language AI

We study models that connect images, language, prompts, and structured information for grounded analysis, reporting, and decision support.

  • Vision–language models
  • Image–text learning
  • Visual grounding
  • Prompt-guided analysis

Medical and biomedical image analysis

We develop methods for diagnosis, segmentation, risk assessment, and structured reporting across 2D and 3D biomedical imaging.

  • Ultrasound
  • Brain MRI
  • Histopathology
  • Dental and craniofacial CBCT

Reliable and interpretable AI

We investigate model robustness, generalization, uncertainty, explanation, and evaluation methods that keep domain experts meaningfully involved.

  • Model reliability
  • Interpretability
  • Uncertainty
  • Human-in-the-loop evaluation

EHR and clinical data intelligence

We analyze electronic health records, clinical time series, laboratory measurements, and other structured data for early warning, risk prediction, and clinical decision support.

  • Electronic health records
  • Clinical time series
  • Risk prediction
  • Decision support

3D vision understanding and world models

We study how AI systems represent, understand, and reason about objects, scenes, motion, and interactions in three-dimensional environments.

  • 3D vision
  • Scene understanding
  • World models
  • Visual and spatial reasoning

Agentic AI, spatial intelligence, and AR/VR

We explore agents that reason and act in visual and spatial environments, including synthetic 3D worlds and immersive systems for healthcare, education, and workforce training.

  • Agentic AI
  • Spatial intelligence
  • AR/VR/MR
  • Immersive simulation

AI for scientific and biomedical discovery

We apply AI to evidence-rich scientific problems involving bioinformatics, computational pathology, multimodal biomedical data, and image–omics integration.

  • Bioinformatics
  • Computational pathology
  • Image–omics
  • Spatial transcriptomics

AI for drug discovery

We are interested in machine learning methods that support molecular representation, candidate prioritization, and data-driven biomedical discovery.

  • Molecular AI
  • Graph learning
  • Candidate prioritization
  • Biomedical discovery

AI for forensic analysis

We investigate computational methods for analyzing forensic evidence, including craniofacial, dental, imaging, and other multimodal information.

  • Forensic imaging
  • Craniofacial analysis
  • Dental evidence
  • Multimodal reasoning

Interested in collaborating?

We welcome research partnerships in healthcare, imaging, clinical data, scientific AI, spatial intelligence, drug discovery, and forensic analysis.

Start a conversation