Zhong Tang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhong Tang has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
Yuxin Liu
School of Information and Management, Guangxi Medical University, Nanning, China
2 shared publicationsZhenkun Tang
School of Information and Management, Guangxi Medical University, Nanning, China
2 shared publicationsZhenyu Tang
Department of Computer Engineering, University of California, Irvine, CA, United States
2 shared publications- Chao Feng
Center for Genomic and Personalized Medicine, Guangxi Medical University, China
1 shared publication Huimin He
Department of Health Service Management, School of Information and Management, Guangxi Medical University, Nanning, China
1 shared publicationJunlai Qiu
Medical AI ReSearch (MARS) Group, Center for Genomic and Personalized Medicine, Guangxi Key Laboratory for Genomic and Personalized Medicine, Nanning, Guangxi Collaborative Innovation Center for Genomic and Personalized Medicine, University Engineering Research Center of Digital Medicine and Healthcare, Guangxi Medical University
1 shared publication
Latest Publications
- Open Access
- 29 May 2026
- Published by Frontiers Media SA in Frontiers in Bioinformatics
- Vol. 6, 1796762
A vasculogenic mimicry subtype unveiled by integrated multi-omics predicts prognosis and guides immunotherapy in MIBC
Introduction: The progression, recurrence, and metastasis of muscle-invasive bladder cancer (MIBC) are closely associated with vasculogenic mimicry (VM). Establishing a VM-based molecular subtyping system and clarifying its regulatory network are essential to overcome current therapeutic limitations in MIBC. Methods: This study collected 1,489 previously reported VM-related genes and identified differentially expressed genes between tumor cells and endothelial cells using single-cell transcriptomic analysis. The intersection of these gene sets yielded VM-related genes, of which 12 were selected to construct a prognostic model validated by multi-omics data, including bulk transcriptomic, genomic, mass cytometry, spatial transcriptomic, and metabolomic datasets. Single-cell and bulk RNA sequencing data were integrated to develop a VM-based molecular subtyping system. Results: MIBC patients were classified into three subtypes: VM, Mixed-VM, and Non-VM, each showing distinct prognostic outcomes consistently validated across multiple independent cohorts. The VM subtype demonstrated the most aggressive clinical behavior, characterized by high cellular plasticity and extensive metabolic reprogramming. Compared to the Non-VM subtype, VM-subtype patients exhibited an immunosuppressive tumor microenvironment and elevated tumor mutational burden (TMB), indicating a potentially favorable response to immunotherapy. Combined treatment with the MEK inhibitor TAK-733 and immune checkpoint inhibitors was proposed as a promising therapeutic strategy for this subgroup. Conclusion: This novel VM-based molecular subtyping system for MIBC shows strong potential for clinical application in prognosis prediction, immunotherapy response evaluation, and targeted drug discovery, providing a framework to guide personalized treatment strategies for MIBC patients.
- 29 May 2026
- Published by Frontiers Media SA in Frontiers in Bioinformatics
- Vol. 6, 1796762
Open AccessA vasculogenic mimicry subtype unveiled by integrated multi-omics predicts prognosis and guides immunotherapy in MIBC
Introduction: The progression, recurrence, and metastasis of muscle-invasive bladder cancer (MIBC) are closely associated with vasculogenic mimicry (VM). Establishing a VM-based molecular subtyping system and clarifying its regulatory network are essential to overcome current therapeutic limitations in MIBC. Methods: This study collected 1,489 previously reported VM-related genes and identified differentially expressed genes between tumor cells and endothelial cells using single-cell transcriptomic analysis. The intersection of these gene sets yielded VM-related genes, of which 12 were selected to construct a prognostic model validated by multi-omics data, including bulk transcriptomic, genomic, mass cytometry, spatial transcriptomic, and metabolomic datasets. Single-cell and bulk RNA sequencing data were integrated to develop a VM-based molecular subtyping system. Results: MIBC patients were classified into three subtypes: VM, Mixed-VM, and Non-VM, each showing distinct prognostic outcomes consistently validated across multiple independent cohorts. The VM subtype demonstrated the most aggressive clinical behavior, characterized by high cellular plasticity and extensive metabolic reprogramming. Compared to the Non-VM subtype, VM-subtype patients exhibited an immunosuppressive tumor microenvironment and elevated tumor mutational burden (TMB), indicating a potentially favorable response to immunotherapy. Combined treatment with the MEK inhibitor TAK-733 and immune checkpoint inhibitors was proposed as a promising therapeutic strategy for this subgroup. Conclusion: This novel VM-based molecular subtyping system for MIBC shows strong potential for clinical application in prognosis prediction, immunotherapy response evaluation, and targeted drug discovery, providing a framework to guide personalized treatment strategies for MIBC patients.
- Open Access
- 30 March 2026
- Published by Frontiers Media SA in Frontiers in Medicine
- Vol. 13, 1773338
LNMSNet: a multi-task deep learning network for pulmonary nodules segmentation and malignancy classification
Introduction: Lung cancer remains the leading cause of global cancer incidence and mortality, with late-stage diagnosis contributing to a stark 5-year survival rate of only 8%. Systematic early detection via low-dose computerized tomography (LDCT) scan can dramatically improve outcomes, with survival exceeding 90% for stage I patients. Pulmonary nodules are the primary radiological precursor, but their accurate characterization is challenged by manual interpretation, inter-reader variability, and the difficulty of visually assessing small, ill-defined lesions on hundreds of CT slices. Methods: To this end, we propose LNMSNet, which extracts Multi-Scale features from Lung Nodules for the joint segmentation and malignancy classification. The model employs a U-shaped encoder-decoder with a ResNet-18 backbone. A key innovation of our LNMSNet is the MSConv module, which uses parallel multi-scale convolutions to capture both fine-grained texture and global contextual features, thereby enlarging the receptive field and improving boundary accuracy and size invariance. Results: We validated the proposed LNMSNet on a multi-center external dataset of 220 CT scans from two tertiary hospitals. The model showed superior performances in both tasks compared to other multi-task models and exhibited stable generalizability across institutions. Conclusion: The proposed LNMSNet effectively leverages multi-scale feature extraction and joint task optimization for accurate pulmonary nodule characterization.
- 30 March 2026
- Published by Frontiers Media SA in Frontiers in Medicine
- Vol. 13, 1773338
Open AccessLNMSNet: a multi-task deep learning network for pulmonary nodules segmentation and malignancy classification
Introduction: Lung cancer remains the leading cause of global cancer incidence and mortality, with late-stage diagnosis contributing to a stark 5-year survival rate of only 8%. Systematic early detection via low-dose computerized tomography (LDCT) scan can dramatically improve outcomes, with survival exceeding 90% for stage I patients. Pulmonary nodules are the primary radiological precursor, but their accurate characterization is challenged by manual interpretation, inter-reader variability, and the difficulty of visually assessing small, ill-defined lesions on hundreds of CT slices. Methods: To this end, we propose LNMSNet, which extracts Multi-Scale features from Lung Nodules for the joint segmentation and malignancy classification. The model employs a U-shaped encoder-decoder with a ResNet-18 backbone. A key innovation of our LNMSNet is the MSConv module, which uses parallel multi-scale convolutions to capture both fine-grained texture and global contextual features, thereby enlarging the receptive field and improving boundary accuracy and size invariance. Results: We validated the proposed LNMSNet on a multi-center external dataset of 220 CT scans from two tertiary hospitals. The model showed superior performances in both tasks compared to other multi-task models and exhibited stable generalizability across institutions. Conclusion: The proposed LNMSNet effectively leverages multi-scale feature extraction and joint task optimization for accurate pulmonary nodule characterization.
- Open Access
- 27 December 2022
- Published by Springer Nature in BMC Public Health
- Vol. 22 (1) , 1-9
How to establishing an indicators framework for evaluating the performances in primary TB control institutions under the new TB control model? Based on a Delphi study conducted in Guangxi, China
Background: In China, the new TB control model of trinity form had been implemented in all parts, and the comprehensively evaluation to the performances in primary TB control institutions were closely related to the working capacity and quality of TB service, but there was still no an unified evaluation indicators framework in practice and few relevant studies. The purpose of this study was to establish an indicators framework for comprehensively evaluating the performances in primary TB control institutions under the new TB control model of trinity form in Guangxi, China.Methods: The Delphi method was used to establish an indicators framework for comprehensively evaluating the performances in primary TB control institutions under the new TB control model of trinity form, and the analytic hierarchy process(AHP) was used to determine the weights of all levels of indicators, from September 2021 to December 2021 in Guangxi, China.Results: A total of 14 experts who had at least 10 years working experience and engaged in TB prevention and control and public health management from health committee, CDC, TB designated hospitals and university of Guangxi were consulted in two rounds. The average age of the experts were (43.3 ± 7.549) years old, and the effective recovery rate of the questionnaire was 100.0%. The average value of authority coefficient of experts (Cr) in the two rounds of consultation was above 0.800. The Kendall’s harmony coefficient (W) of experts’ opinions on the first-level indicators, the second-level indicators and the third-level indicators were 0.786, 0.201 and 0.169, respectively, which were statistically significant (P < 0.05). Finally, an indicators framework was established, which included 2 first-level indicators, 10 second-level indicators and 37 third-level indicators. The results of analytic hierarchy process (AHP) showed that the consistency test of all levels of indicators were CI < 0.10, which indicating that the weight of each indicator was acceptable.Conclusion: The indicators framework established in this study was in line with the reality, had reasonable weights, and could provide a scientific evaluation tool for comprehensively evaluating the performances in primary TB control institutions under the new TB control model of trinity form in Guangxi, China.
- 27 December 2022
- Published by Springer Nature in BMC Public Health
- Vol. 22 (1) , 1-9
Open AccessHow to establishing an indicators framework for evaluating the performances in primary TB control institutions under the new TB control model? Based on a Delphi study conducted in Guangxi, China
Background: In China, the new TB control model of trinity form had been implemented in all parts, and the comprehensively evaluation to the performances in primary TB control institutions were closely related to the working capacity and quality of TB service, but there was still no an unified evaluation indicators framework in practice and few relevant studies. The purpose of this study was to establish an indicators framework for comprehensively evaluating the performances in primary TB control institutions under the new TB control model of trinity form in Guangxi, China.Methods: The Delphi method was used to establish an indicators framework for comprehensively evaluating the performances in primary TB control institutions under the new TB control model of trinity form, and the analytic hierarchy process(AHP) was used to determine the weights of all levels of indicators, from September 2021 to December 2021 in Guangxi, China.Results: A total of 14 experts who had at least 10 years working experience and engaged in TB prevention and control and public health management from health committee, CDC, TB designated hospitals and university of Guangxi were consulted in two rounds. The average age of the experts were (43.3 ± 7.549) years old, and the effective recovery rate of the questionnaire was 100.0%. The average value of authority coefficient of experts (Cr) in the two rounds of consultation was above 0.800. The Kendall’s harmony coefficient (W) of experts’ opinions on the first-level indicators, the second-level indicators and the third-level indicators were 0.786, 0.201 and 0.169, respectively, which were statistically significant (P < 0.05). Finally, an indicators framework was established, which included 2 first-level indicators, 10 second-level indicators and 37 third-level indicators. The results of analytic hierarchy process (AHP) showed that the consistency test of all levels of indicators were CI < 0.10, which indicating that the weight of each indicator was acceptable.Conclusion: The indicators framework established in this study was in line with the reality, had reasonable weights, and could provide a scientific evaluation tool for comprehensively evaluating the performances in primary TB control institutions under the new TB control model of trinity form in Guangxi, China.