Guanghua Yang

Guanghua Yang


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 June 2025 in Scientific Reports
Verified
Journal Article
Open access
PDF file
HXHuimin Xiao
Huimin Xiao
GYGuanghua Yang
Guanghua Yang
ZLZhuocheng Li
Zhuocheng Li
CYChanghua Yi
Changhua Yi
To assess the suitability of Transformer-based architectures for medical image segmentation and investigate the potential advantages of Graph Neural Networks (GNNs) in this domain. We analyze the limitations of the Transformer, which models medical images as sequences of image patches, limiting its flexibility in capturing complex and irregular tumor structures. To address it, we propose U-GNN, a pure GNN-based U-shaped architecture designed for medical image segmentation. U-GNN retains the U-Net-inspired inductive bias while leveraging GNNs’ topological modeling capabilities. The architecture consists of Vision GNN blocks stacked into a U-shaped structure. Additionally, we introduce the concept of multi-order similarity and propose a zero-computation-cost approach to incorporate higher-order similarity in graph construction. Each Vision GNN block segments the image into patch nodes, constructs multi-order similarity graphs, and aggregates node features via multi-order node information aggregation. Experimental evaluations on multi-organ and cardiac segmentation datasets demonstrate that U-GNN significantly outperforms existing CNN- and Transformer-based models. U-GNN achieves a 6% improvement in Dice Similarity Coefficient (DSC) and an 18% reduction in Hausdorff Distance (HD) compared to state-of-the-art methods. The source code will be released upon paper acceptance.

Verified
Journal Article
Open access
PDF file
WJWeiyang Jin
Weiyang Jin
MWMengying Wang
Mengying Wang
JWJingwei Wang
Jingwei Wang
JFJinyi Fan
Jinyi Fan
JFJie Fang
Jie Fang
GYGuanghua Yang
Guanghua Yang
MICB-targeting CAR-NK (chimeric antigen receptor-modified natural killer cells) therapy may serve as off-the-shelf immunotherapy. We designed soluble Anti-MICB-scFv blocks tumor immune evasion targeting the MICB antigen, thereby enhancing CAR-NK cytotoxicity while reactivating endogenous immune attacks against malignancies. The Anti-MICB-CAR includes two Anti-MICB-scFv connected by an F2A linker, the CD8 hinge and transmembrane domain, the 4-1BB co-stimulatory domain, the CD3ζ activation domain, and IL-15. The expression efficiency of Anti-MICB-CAR in NK cells was investigated by flow cytometry; ELISA demonstrated that Anti-MICB-CAR-NK secreted free Anti-MICB-scFv and detected IL-15 secretion. Flow cytometry and CCK8 were utilized to study Anti-MICB-CAR-NK on tumor cell viability. The PANC-1 xenograft model was established in order to elucidate the anti-tumor effects of Anti-MICB-CAR-NK in vivo. In vitro investigations have demonstrated that the treatment of tumor cells with Anti-MICB-CAR-NK supernatant + NK cells or Anti-MICB-CAR-NK cells not only significantly increased the cytotoxic activity of tumor cells, but also secreted and produced higher levels of IL-15, IFN-γ, TNF-α, perforin, and granzyme B compared with NK cells. Anti-MICB-CAR-NK cells exhibit strong cytotoxic activity against tumor cells with high MICB expression. In vivo, Anti-MICB-CAR-NK cells exhibited a substantial inhibitory effect on tumor growth. The IHC results reveal that Anti-MICB-CAR-NK cells show a more pronounced ability to infiltrate the tumor. We demonstrated the successful expression of Anti-MICB-CAR in NK cells, which enhances the anti-tumor activity of NK cells both in vitro and in vivo. This stress ligand-targeting approach provides a promising strategy for solid tumors.