RESEARCH OUTPUTS
Selected work spanning computational pathology, PET synthesis, ultrasound imaging, and multimodal foundation models. Each project connects a clinical or scientific task with a concrete modeling strategy.
IEEE Transactions on Medical Imaging, 45(5), 2542–2560
A two-stage PET synthesis framework that learns a vector-quantized codebook prior from high-quality PET data and uses it to guide synthesis.
Image and Vision Computing, 167, 105918
A journal article on multi-order enhanced multiple instance learning for stronger whole-slide representation learning in computational pathology.
Advanced Ultrasound in Diagnosis and Therapy, 9(4), 467–482
A review of photoacoustic–ultrasound fusion imaging, AI-enhanced interpretation, and clinical translation for breast cancer diagnosis and therapy.
中华医学超声杂志（电子版）
A review of multimodal foundation models, their cross-applications with ultrasound imaging, and emerging opportunities for clinical medical AI.
arXiv preprint arXiv:2408.15032
State-space-duality-based multiple instance learning for efficient long-sequence modeling in computational pathology and whole-slide image analysis.