Projects

Efficient Whole Slide Pathology VQA via Token Compression
In this project, we propose the very first MLLM architecture, named Token Compression Pathology LLaVA (TCP-LLaVA) to perform WSI VQA via token compression.
Weimin Lyu, Qingqiao Hu, Kehan Qi, Zhan Shi, Wentao Huang, Saumya Gupta, Chao Chen
PDF
Learning Brain Tumor Representation in 3D High-Resolution MR Images via Interpretable State Space Models
In this project, we investigate the State Space Models, especially Mamba. We use Masked Image Modeling pre-training strategy, and fintune the model for 3D radiomics analysis. We evaluate the model in terms of downstream classification task and its latent space interpretability.
Qingqiao Hu, Daoan Zhang, Jiebo Luo, Zhenyu Gong, Benedikt Wiestler, Jianguo Zhang, Hongwei Bran Li
PDF Code
Bayeisan One-Encoder-Multiple-Decoder (BOEMD) Network
In this project, we1 proposed a simple yet effective Bayesian Neural Network (BNN) framework, especifically designed for inter-rater uncertainty quantification probelm in medical imaging. We call this framework, Bayesian One-Encoder-Multiple-Decoder (BOEMD) network.
Qingqiao Hu, Hao Wang, Jing Luo, Yunhao Luo, Zhiheng Zhang, Jan S. Kirschke, Benedikt Wiestler, Bjoern Menze, Jianguo Zhang, Hongwei Bran Li
PDF Code