Qi Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Qi Yang has a total of 159 co-authors. Here are the top co-authors by number of shared publications:
- Ling Guo
Sun Yat-sen University Cancer Center 2 Guangzhou
10 shared publications - Guoheng Huang
Guangdong University of Technology
8 shared publications - Chi-Man Pun
University of Macau
7 shared publications - Xiaochen Yuan
Macao Polytechnic University
6 shared publications - Hai-Qiang Mai
State Key Laboratory of Oncology in South China, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Department of Nasopharyngeal Carcinoma, Collaborative Innovation Center for Cancer Medicine Sun Yat‐sen University Cancer Center Guangzhou Guangdong People's Republic of China
5 shared publications - Huan-Xin Lin
State Key Laboratory of Oncology in South China, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
4 shared publications
Latest Publications
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Radiation and Plasma Medical Sciences
- Vol. 10 (7) , 1011-1023
AMGL: Adaptive Multimodal Graph Learning for Brain Disease Prediction
Graph-based approaches have been widely adopted in biomedical applications for modeling multimodal data, particularly in the accurate diagnosis and effective treatment of brain disorders. Most existing graph-based multimodal medical data processing methods typically extract features by fusing multimodal information through weighted operations, and then manually define graph structures based on specific modalities to learn patient representations via graph embedding. However, these methods often overlook the complex correlations and discrepancies across modalities, making it difficult to obtain highly relevant information. Moreover, the prior construction of an appropriate graph presents a considerable challenge, as manually defined structures are susceptible to spurious or noisy edges. These factors inevitably lead to incorrect predictions in real-world clinical scenarios. To address these limitations, we propose an end-to-end Adaptive Multimodal Graph Learning (AMGL) framework that comprises two key modules: Modal-Aware Integration Learning (MAIL) and Cluster-constrained Adaptive Graph Learning (CAGL). MAIL captures both inter-modal relevance and complementarity to construct enriched modality-aware representations, while CAGL performs adaptive graph learning based on data clustering and utilizes a Graph-Gated Neural Network (GGNN) for disease prediction. Experimental results on the TADPOLE and ABIDE datasets demonstrate that our method achieves superior classification accuracy and generalization capability, with an average performance gain of 2%–3% over state-of-the-art approaches. The code of AMGL is available at https://github.com/WuJUNEhhh/AMGL.
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Radiation and Plasma Medical Sciences
- Vol. 10 (7) , 1011-1023
AMGL: Adaptive Multimodal Graph Learning for Brain Disease Prediction
Graph-based approaches have been widely adopted in biomedical applications for modeling multimodal data, particularly in the accurate diagnosis and effective treatment of brain disorders. Most existing graph-based multimodal medical data processing methods typically extract features by fusing multimodal information through weighted operations, and then manually define graph structures based on specific modalities to learn patient representations via graph embedding. However, these methods often overlook the complex correlations and discrepancies across modalities, making it difficult to obtain highly relevant information. Moreover, the prior construction of an appropriate graph presents a considerable challenge, as manually defined structures are susceptible to spurious or noisy edges. These factors inevitably lead to incorrect predictions in real-world clinical scenarios. To address these limitations, we propose an end-to-end Adaptive Multimodal Graph Learning (AMGL) framework that comprises two key modules: Modal-Aware Integration Learning (MAIL) and Cluster-constrained Adaptive Graph Learning (CAGL). MAIL captures both inter-modal relevance and complementarity to construct enriched modality-aware representations, while CAGL performs adaptive graph learning based on data clustering and utilizes a Graph-Gated Neural Network (GGNN) for disease prediction. Experimental results on the TADPOLE and ABIDE datasets demonstrate that our method achieves superior classification accuracy and generalization capability, with an average performance gain of 2%–3% over state-of-the-art approaches. The code of AMGL is available at https://github.com/WuJUNEhhh/AMGL.
- 29 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Radiation and Plasma Medical Sciences
- Vol. PP (24697311) , 1
CFCM-Gen:Cross-modal Feature Co-optimization Mechanism for 3D Radiology Report Generation
3D medical imaging provides rich volumetric information and plays a critical role in disease diagnosis, however, due to the high-dimensional complexity of 3D radiology images, automatically generating radiology reports remains a challenging task. Specifically, this complexity makes cross-modal feature alignment between images and text challenging, weakening their semantic association mapping. To this end, we propose the Cross-modal Feature Co-optimization Mechanism for 3D Radiology Report Generation (CFCM-Gen), which aims to enhance cross-modal feature co-optimization to improve semantic alignment between images and text, thereby strengthening the reporting consistency and cross-modal semantic alignment of automatically generated reports. In addition, we construct the first 3D MRI-text aligned dataset specifically designed for nasopharyngeal carcinoma diagnosis, named NPC-RG, which provides high-quality data support for the 3D radiology report generation task. In the design of our method, firstly, we design Keywords-Prompt for Local-Global Feature Fusion Module (KPLG), which uses medical keywords as hints to guide the local features and global features of 3D radiology images to be fused. This enhances the semantic representation of key regions and provides more precise feature information for cross-modal alignment. Secondly, the Dynamic Bidirectional Modality Interaction Module (DBMI) is proposed to dynamically optimize bidirectional information flow. It refines the correspondence between image regions and textual words, mitigates semantic conflicts, and enhances the alignment accuracy of cross-modal features. Finally, we propose Block-specific Reinforcement Feedback Mechanism (BSRF) to optimize semantic modeling by text block partitioning and reinforcement learning reward signals. Experimental results show that CFCM-Gen performs well in terms of reporting consistency and semantic correspondence. The code is available at:https://github.com/Andyfever123/CFCM-Gen.
- 29 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Radiation and Plasma Medical Sciences
- Vol. PP (24697311) , 1
CFCM-Gen:Cross-modal Feature Co-optimization Mechanism for 3D Radiology Report Generation
3D medical imaging provides rich volumetric information and plays a critical role in disease diagnosis, however, due to the high-dimensional complexity of 3D radiology images, automatically generating radiology reports remains a challenging task. Specifically, this complexity makes cross-modal feature alignment between images and text challenging, weakening their semantic association mapping. To this end, we propose the Cross-modal Feature Co-optimization Mechanism for 3D Radiology Report Generation (CFCM-Gen), which aims to enhance cross-modal feature co-optimization to improve semantic alignment between images and text, thereby strengthening the reporting consistency and cross-modal semantic alignment of automatically generated reports. In addition, we construct the first 3D MRI-text aligned dataset specifically designed for nasopharyngeal carcinoma diagnosis, named NPC-RG, which provides high-quality data support for the 3D radiology report generation task. In the design of our method, firstly, we design Keywords-Prompt for Local-Global Feature Fusion Module (KPLG), which uses medical keywords as hints to guide the local features and global features of 3D radiology images to be fused. This enhances the semantic representation of key regions and provides more precise feature information for cross-modal alignment. Secondly, the Dynamic Bidirectional Modality Interaction Module (DBMI) is proposed to dynamically optimize bidirectional information flow. It refines the correspondence between image regions and textual words, mitigates semantic conflicts, and enhances the alignment accuracy of cross-modal features. Finally, we propose Block-specific Reinforcement Feedback Mechanism (BSRF) to optimize semantic modeling by text block partitioning and reinforcement learning reward signals. Experimental results show that CFCM-Gen performs well in terms of reporting consistency and semantic correspondence. The code is available at:https://github.com/Andyfever123/CFCM-Gen.
- Open Access
- 31 March 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Concurrent chemoradiotherapy plus nimotuzumab versus chemoradiotherapy alone for locoregionally advanced nasopharyngeal carcinoma with a suboptimal response to induction chemotherapy: a randomized phase 2 trial
Induction chemotherapy (IC) followed by concurrent chemoradiotherapy (CCRT) is the standard treatment for locoregionally advanced nasopharyngeal carcinoma (LA-NPC); however, patients with a suboptimal response to IC, defined as detectable Epstein–Barr virus DNA and/or stable or progressive disease after IC, remain at high risk of treatment failure. Here we report an open-label, randomised, phase 2 trial evaluating whether adding nimotuzumab, a humanised anti–epidermal growth factor receptor antibody, to CCRT improves outcomes in this high-risk population. A total of 246 patients with untreated, non-keratinising, stage II–IVA LA-NPC were randomly assigned (1:1) to receive CCRT with or without nimotuzumab. The primary endpoint was 2-year progression-free survival (PFS); secondary endpoints included overall survival, distant metastasis–free survival, locoregional relapse–free survival, short-term response rate, and safety. At a median follow-up of 47 months, the 2-year PFS was 81.0% (90% confidence interval [CI], 74.3–86.1) in the nimotuzumab plus CCRT group and 80.8% (90% CI, 74.2–85.7) in the CCRT-alone group (hazard ratio, 0.93 [90% CI, 0.63–1.37]; p = 0.70). Survival outcomes were similar between groups, while low-grade rash occurred more frequently with nimotuzumab. These findings indicate that adding nimotuzumab to CCRT does not improve survival in patients with LA-NPC with a suboptimal response to IC, underscoring the need for predictive biomarkers and alternative therapeutic strategies. Trial registration: NCT04223024.
- 31 March 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessConcurrent chemoradiotherapy plus nimotuzumab versus chemoradiotherapy alone for locoregionally advanced nasopharyngeal carcinoma with a suboptimal response to induction chemotherapy: a randomized phase 2 trial
Induction chemotherapy (IC) followed by concurrent chemoradiotherapy (CCRT) is the standard treatment for locoregionally advanced nasopharyngeal carcinoma (LA-NPC); however, patients with a suboptimal response to IC, defined as detectable Epstein–Barr virus DNA and/or stable or progressive disease after IC, remain at high risk of treatment failure. Here we report an open-label, randomised, phase 2 trial evaluating whether adding nimotuzumab, a humanised anti–epidermal growth factor receptor antibody, to CCRT improves outcomes in this high-risk population. A total of 246 patients with untreated, non-keratinising, stage II–IVA LA-NPC were randomly assigned (1:1) to receive CCRT with or without nimotuzumab. The primary endpoint was 2-year progression-free survival (PFS); secondary endpoints included overall survival, distant metastasis–free survival, locoregional relapse–free survival, short-term response rate, and safety. At a median follow-up of 47 months, the 2-year PFS was 81.0% (90% confidence interval [CI], 74.3–86.1) in the nimotuzumab plus CCRT group and 80.8% (90% CI, 74.2–85.7) in the CCRT-alone group (hazard ratio, 0.93 [90% CI, 0.63–1.37]; p = 0.70). Survival outcomes were similar between groups, while low-grade rash occurred more frequently with nimotuzumab. These findings indicate that adding nimotuzumab to CCRT does not improve survival in patients with LA-NPC with a suboptimal response to IC, underscoring the need for predictive biomarkers and alternative therapeutic strategies. Trial registration: NCT04223024.
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. PP (21682194) , 1-16
HDPL: Hypergraph-based Dynamic Prompting Learning for Incomplete Multimodal Medical Learning
Multimodal learning has garnered significant attention in the medical field due to its ability to provide a more comprehensive perspective utilizing various types of data, that aids in making more accurate decisions. However, the complexity of medical data, coupled with missing modalities, severely hinders predictive accuracy. Existing methods for multimodal learning with missing modalities still face considerable challenges. For instance, approaches that construct multimodal shared feature spaces often result in high computational costs, while methods that infer missing modalities based on complete ones may overly rely on the complete modalities, potentially skewing results. Pre-trained transformer methods address these issues but still have limitations, such as it can only process one missing modality at testing-stage. This is partly because structured data, unlike sequential data, lacks inherent minimum semantic units or natural order. Additionally, the positional encodings generated by this type of methods may introduce information interference when applied to structured data, leading to poor alignment with sequential data during modality fusion in transformer models. To tackle these challenges, we introduce HDPL: Hypergraph-based Dynamic Prompt Learning for Incomplete Multimodal Medical Learning, comprising three modules. The High-Order Hypergraph Embedding module can identify the minimal semantic units within structured data and utilizes hypergraph structures to extract high-dimensional features from clinical data. The Multimodal Medical Data Integrator module closes the distance of the embedding vectors corresponding in the shared space of modality-features, facilitating the integration of modalities in transformer. The Dynamic Network Structure Optimization module is a dynamic learning network by dynamically change the width and depth of network, improving the overall performance of the model, and it alleviates the shortcomings caused by incomplete modality to some extent. Through comprehensive experimentation, we demonstrate the efficiency and robustness of our model in dealing missing modalities and reducing training-burdens. Our code and dataset are available at https://github.com/colorful823/HDPL.
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. PP (21682194) , 1-16
HDPL: Hypergraph-based Dynamic Prompting Learning for Incomplete Multimodal Medical Learning
Multimodal learning has garnered significant attention in the medical field due to its ability to provide a more comprehensive perspective utilizing various types of data, that aids in making more accurate decisions. However, the complexity of medical data, coupled with missing modalities, severely hinders predictive accuracy. Existing methods for multimodal learning with missing modalities still face considerable challenges. For instance, approaches that construct multimodal shared feature spaces often result in high computational costs, while methods that infer missing modalities based on complete ones may overly rely on the complete modalities, potentially skewing results. Pre-trained transformer methods address these issues but still have limitations, such as it can only process one missing modality at testing-stage. This is partly because structured data, unlike sequential data, lacks inherent minimum semantic units or natural order. Additionally, the positional encodings generated by this type of methods may introduce information interference when applied to structured data, leading to poor alignment with sequential data during modality fusion in transformer models. To tackle these challenges, we introduce HDPL: Hypergraph-based Dynamic Prompt Learning for Incomplete Multimodal Medical Learning, comprising three modules. The High-Order Hypergraph Embedding module can identify the minimal semantic units within structured data and utilizes hypergraph structures to extract high-dimensional features from clinical data. The Multimodal Medical Data Integrator module closes the distance of the embedding vectors corresponding in the shared space of modality-features, facilitating the integration of modalities in transformer. The Dynamic Network Structure Optimization module is a dynamic learning network by dynamically change the width and depth of network, improving the overall performance of the model, and it alleviates the shortcomings caused by incomplete modality to some extent. Through comprehensive experimentation, we demonstrate the efficiency and robustness of our model in dealing missing modalities and reducing training-burdens. Our code and dataset are available at https://github.com/colorful823/HDPL.
- 29 January 2026
- Published by Elsevier in Computerized Medical Imaging and Graphics
- Vol. 128, 102711
MRI-based deep learning model predicts recurrent nasopharyngeal carcinoma in post-radiation nasopharyngeal necrosis
- 29 January 2026
- Published by Elsevier in Computerized Medical Imaging and Graphics
- Vol. 128, 102711
MRI-based deep learning model predicts recurrent nasopharyngeal carcinoma in post-radiation nasopharyngeal necrosis
- 24 January 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 110, 103958
MADAT: Missing-aware dynamic adaptive transformer model for medical prognosis prediction with incomplete multimodal data
- 24 January 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 110, 103958
MADAT: Missing-aware dynamic adaptive transformer model for medical prognosis prediction with incomplete multimodal data
- Open Access
- 21 July 2025
- Published by Wiley in Advanced Science
- Vol. 12 (39) , e00950
The CXCL10‐CXCR3 Axis Induces Tumor‐Associated Neutrophils to Interfere with CTLs‐Mediated Antitumor Activity in EBV‐Associated Epithelial Cancers
Although Epstein–Barr virus (EBV)‐associated epithelial cancers are categorized as immunologically “hot” tumors, they have unsatisfactory responses to immunotherapy. Increasing evidence has shown that therapeutic failure is due to an immunosuppressive tumor microenvironment established by EBV. In this study, a negative correlation is found between the infiltration of neutrophils and that of cytotoxic T lymphocytes (CTLs) containing granzyme B in EBV‐associated epithelial cancers. The CXCL10‐CXCR3 axis in EBV‐associated epithelial cancer cells controls the extrusion of neutrophil extracellular traps (NETs), which interferes with the antitumor activity of EBV antigen‐specific T cells in vitro and in vivo. NETs are positively correlated with the number of dysfunctional CTLs in EBV‐associated epithelial cancers, and are confirmed to be an independent prognostic factor for patients with EBV‐associated epithelial cancers. In conclusion, these findings reveal a novel mechanism of immunosuppression of tumor‐associated neutrophils (TANs) in EBV‐associated epithelial cancers. Targeting NETs formation in TANs may be a potential strategy for improving the efficacy of immunotherapy against EBV‐associated epithelial cancers.
- 21 July 2025
- Published by Wiley in Advanced Science
- Vol. 12 (39) , e00950
Open AccessThe CXCL10‐CXCR3 Axis Induces Tumor‐Associated Neutrophils to Interfere with CTLs‐Mediated Antitumor Activity in EBV‐Associated Epithelial Cancers
Although Epstein–Barr virus (EBV)‐associated epithelial cancers are categorized as immunologically “hot” tumors, they have unsatisfactory responses to immunotherapy. Increasing evidence has shown that therapeutic failure is due to an immunosuppressive tumor microenvironment established by EBV. In this study, a negative correlation is found between the infiltration of neutrophils and that of cytotoxic T lymphocytes (CTLs) containing granzyme B in EBV‐associated epithelial cancers. The CXCL10‐CXCR3 axis in EBV‐associated epithelial cancer cells controls the extrusion of neutrophil extracellular traps (NETs), which interferes with the antitumor activity of EBV antigen‐specific T cells in vitro and in vivo. NETs are positively correlated with the number of dysfunctional CTLs in EBV‐associated epithelial cancers, and are confirmed to be an independent prognostic factor for patients with EBV‐associated epithelial cancers. In conclusion, these findings reveal a novel mechanism of immunosuppression of tumor‐associated neutrophils (TANs) in EBV‐associated epithelial cancers. Targeting NETs formation in TANs may be a potential strategy for improving the efficacy of immunotherapy against EBV‐associated epithelial cancers.
- 31 December 2024
- Published by Elsevier in Journal of Biomedical Informatics
- Vol. 161, 104764
Visual-linguistic Diagnostic Semantic Enhancement for medical report generation
- 31 December 2024
- Published by Elsevier in Journal of Biomedical Informatics
- Vol. 161, 104764
Visual-linguistic Diagnostic Semantic Enhancement for medical report generation