Song Zheng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Song Zheng has a total of 74 co-authors. Here are the top co-authors by number of shared publications:
- Yaguang Kong
College of Automation, Hangzhou Dianzi University, Hangzhou, China
6 shared publications - Peng Jiang
School of Information Science and Technology, Hangzhou Normal University, Hangzhou, China
5 shared publications - Xiaoqing Zheng
School of Automation, Hangzhou Dianzi University, Hangzhou 310018, China
3 shared publications Jianhui Chen
Fujian Medical University Union Hospital
2 shared publications- Jinkun Men
Institute of Safety Science and Engineering, South China University of Technology, Guangzhou, China
2 shared publications Junjie Bai
Shengli Clinical College of Fujian Medical University
2 shared publications
Latest Publications
- Open Access
- 1 January 2026
- Published by American Association for the Advancement of Science (AAAS) in Research
- Vol. 9
IFI44 Promotes Clear Cell Renal Cell Carcinoma Progression via PRDX1 and Predicts Poor Prognosis
Clear cell renal cell carcinoma (ccRCC) is a lethal urologic malignancy with limited biomarkers for prognosis and therapeutic stratification. Interferon-induced protein 44 (IFI44) has been implicated in immune regulation, but its role in ccRCC is unclear. To address this gap, we comprehensively investigated the clinical significance, biological roles, and molecular mechanisms of IFI44 in ccRCC pathogenesis. Using integrative transcriptomic analysis of the Gene Expression Omnibus and the Cancer Genome Atlas-Kidney Clear Cell Carcinoma cohorts, we first identified IFI44 as a key candidate gene. Bioinformatic enrichment analyses and immune infiltration profiling were conducted to investigate potential mechanisms. In parallel, we established ccRCC cell lines with stable IFI44 knockdown and evaluated phenotypic changes using Cell Counting Kit-8, Transwell assays, wound-healing assays, and flow cytometry, thereby examining cell proliferation, apoptosis, migration, and invasion. We observed that IFI44 was markedly elevated in ccRCC tissues, and its increased level was closely associated with advanced tumor stage and poorer patient survival. Enrichment analyses indicated that IFI44 participates in pathways related to viral response, RNA splicing, and mRNA processing. Moreover, elevated IFI44 expression may be associated with an immunosuppressive tumor microenvironment, as suggested by increased infiltration of effector T cells and M1 macrophages, along with decreased infiltration of activated dendritic cells. In mechanistic studies, IFI44 knockdown markedly suppressed cell proliferation, triggered apoptosis, and reduced both migratory and invasive capacities, whereas PRDX1 overexpression rescued these phenotypes and PRDX1 was shown to interact with IFI44. In summary, the data show that IFI44 acts as an oncogene in ccRCC, promoting tumor progression through its interaction with PRDX1 while also shaping an immunosuppressive microenvironment, and suggest that IFI44 is a promising biomarker of prognosis and candidate therapeutic target for ccRCC.
- 1 January 2026
- Published by American Association for the Advancement of Science (AAAS) in Research
- Vol. 9
Open AccessIFI44 Promotes Clear Cell Renal Cell Carcinoma Progression via PRDX1 and Predicts Poor Prognosis
Clear cell renal cell carcinoma (ccRCC) is a lethal urologic malignancy with limited biomarkers for prognosis and therapeutic stratification. Interferon-induced protein 44 (IFI44) has been implicated in immune regulation, but its role in ccRCC is unclear. To address this gap, we comprehensively investigated the clinical significance, biological roles, and molecular mechanisms of IFI44 in ccRCC pathogenesis. Using integrative transcriptomic analysis of the Gene Expression Omnibus and the Cancer Genome Atlas-Kidney Clear Cell Carcinoma cohorts, we first identified IFI44 as a key candidate gene. Bioinformatic enrichment analyses and immune infiltration profiling were conducted to investigate potential mechanisms. In parallel, we established ccRCC cell lines with stable IFI44 knockdown and evaluated phenotypic changes using Cell Counting Kit-8, Transwell assays, wound-healing assays, and flow cytometry, thereby examining cell proliferation, apoptosis, migration, and invasion. We observed that IFI44 was markedly elevated in ccRCC tissues, and its increased level was closely associated with advanced tumor stage and poorer patient survival. Enrichment analyses indicated that IFI44 participates in pathways related to viral response, RNA splicing, and mRNA processing. Moreover, elevated IFI44 expression may be associated with an immunosuppressive tumor microenvironment, as suggested by increased infiltration of effector T cells and M1 macrophages, along with decreased infiltration of activated dendritic cells. In mechanistic studies, IFI44 knockdown markedly suppressed cell proliferation, triggered apoptosis, and reduced both migratory and invasive capacities, whereas PRDX1 overexpression rescued these phenotypes and PRDX1 was shown to interact with IFI44. In summary, the data show that IFI44 acts as an oncogene in ccRCC, promoting tumor progression through its interaction with PRDX1 while also shaping an immunosuppressive microenvironment, and suggest that IFI44 is a promising biomarker of prognosis and candidate therapeutic target for ccRCC.
- 6 November 2024
- Published by Elsevier in Chemical Engineering Research and Design
- Vol. 212, 168-184
A temporal convolution network-based just-in-time learning method for industrial quality variable prediction
- 6 November 2024
- Published by Elsevier in Chemical Engineering Research and Design
- Vol. 212, 168-184
A temporal convolution network-based just-in-time learning method for industrial quality variable prediction
- Open Access
- 26 July 2024
- Published by MDPI AG in Applied Sciences
- Vol. 14  (15) , 6562
Collaborative Optimization Algorithm Based on Chameleon Swarm Algorithm Designed for the Problem of Beer Production Scheduling
Aiming at finding a better way to solve the problem of beer production scheduling, a new collaborative optimization based on the Manhattan Distance and Chameleon Swarm Algorithm is proposed. Firstly, a dynamic relaxation factor is introduced to the constraints at the system level, which combines the changing trend of the inconsistency information and the optimal solution of the discipline level. Additionally, the Manhattan Distance is used to replace the square of the Euclidean Distance at the system level. Thirdly, the Chameleon Swarm Algorithm is used to improve the update rule during the process of iteration. As these improvements are applied to the collaborative optimization, the steps of this new algorithm are given. Through the test case of a multivariate function, it can be found that the algorithm has been improved compared to the original algorithm. Then, a model for beer production scheduling is proposed, and the results of the optimization show that the improved collaborative optimization has better optima effectiveness and fewer iterations and is not sensitive to initial points, which proves that the improved collaborative optimization has a better ability to solve the problem of beer production scheduling than normal collaborative optimization and collaborative optimization with fixed relaxation factors.
- 26 July 2024
- Published by MDPI AG in Applied Sciences
- Vol. 14  (15) , 6562
Open AccessCollaborative Optimization Algorithm Based on Chameleon Swarm Algorithm Designed for the Problem of Beer Production Scheduling
Aiming at finding a better way to solve the problem of beer production scheduling, a new collaborative optimization based on the Manhattan Distance and Chameleon Swarm Algorithm is proposed. Firstly, a dynamic relaxation factor is introduced to the constraints at the system level, which combines the changing trend of the inconsistency information and the optimal solution of the discipline level. Additionally, the Manhattan Distance is used to replace the square of the Euclidean Distance at the system level. Thirdly, the Chameleon Swarm Algorithm is used to improve the update rule during the process of iteration. As these improvements are applied to the collaborative optimization, the steps of this new algorithm are given. Through the test case of a multivariate function, it can be found that the algorithm has been improved compared to the original algorithm. Then, a model for beer production scheduling is proposed, and the results of the optimization show that the improved collaborative optimization has better optima effectiveness and fewer iterations and is not sensitive to initial points, which proves that the improved collaborative optimization has a better ability to solve the problem of beer production scheduling than normal collaborative optimization and collaborative optimization with fixed relaxation factors.
- 8 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 24  (8) , 13530-13544
Information Filtering Unit-Based Long Short-Term Memory Network for Industrial Soft Sensor Modeling
The use of soft sensors is an alternative solution to physical sensors for real-time acquisition of key quality variables that cannot be measured online in real-time. LSTM-based methods have been improving the state-of-the-art in soft sensors. In order to mine the correlation between historical information and current input, and utilize the relatively important historical information, we propose an underlying recurrent neural network unit called information filtering unit (IFU) and an IFU-based LSTM soft sensor (IF-LSTM). In IFU, a feed forward neural network is firstly constructed to predict the hidden state of the previous time step based on the current input and subsequently a scaling factor is obtained, and the hidden state vector is then scaled to highlight those important components that contain non-redundant information with the current input. Finally, an IF-LSTM model is constructed by combining IFU and LSTM units for predicting hard-to-obtain key quality variables from easily-measured process variables, and its performance is validated on two benchmark industrial datasets (SRU and DC datasets). Experimental results show that the performance of the IF-LSTM outperforms popular RNN variants such as LSTM, GRU, Bi-LSTM and Bi-GRU in all metrics (RMSE, MAE, MAPE and R 2 ), and achieves an R 2 of 0.9929 on the DC dataset, indicating that the predicted values are well-fitted to the true values. Furthermore, comparative experiments with state-of-the-art methods in the literature show that the IF-LSTM model achieves the best RMSE of 0.0069 on the DC dataset, and the second best RMSE of 0.0124 on the SRU dataset.
- 8 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 24  (8) , 13530-13544
Information Filtering Unit-Based Long Short-Term Memory Network for Industrial Soft Sensor Modeling
The use of soft sensors is an alternative solution to physical sensors for real-time acquisition of key quality variables that cannot be measured online in real-time. LSTM-based methods have been improving the state-of-the-art in soft sensors. In order to mine the correlation between historical information and current input, and utilize the relatively important historical information, we propose an underlying recurrent neural network unit called information filtering unit (IFU) and an IFU-based LSTM soft sensor (IF-LSTM). In IFU, a feed forward neural network is firstly constructed to predict the hidden state of the previous time step based on the current input and subsequently a scaling factor is obtained, and the hidden state vector is then scaled to highlight those important components that contain non-redundant information with the current input. Finally, an IF-LSTM model is constructed by combining IFU and LSTM units for predicting hard-to-obtain key quality variables from easily-measured process variables, and its performance is validated on two benchmark industrial datasets (SRU and DC datasets). Experimental results show that the performance of the IF-LSTM outperforms popular RNN variants such as LSTM, GRU, Bi-LSTM and Bi-GRU in all metrics (RMSE, MAE, MAPE and R 2 ), and achieves an R 2 of 0.9929 on the DC dataset, indicating that the predicted values are well-fitted to the true values. Furthermore, comparative experiments with state-of-the-art methods in the literature show that the IF-LSTM model achieves the best RMSE of 0.0069 on the DC dataset, and the second best RMSE of 0.0124 on the SRU dataset.
- 25 December 2023
- Published by Springer Nature
- p. 468-479
MixU-Net: Hybrid CNN-MLP Networks for Urinary Collecting System Segmentation
- 25 December 2023
- Published by Springer Nature
- p. 468-479
MixU-Net: Hybrid CNN-MLP Networks for Urinary Collecting System Segmentation
- 1 October 2023
- Published by Springer Nature
- p. 123-132
A Novel Video-CTU Registration Method with Structural Point Similarity for FURS Navigation
- 1 October 2023
- Published by Springer Nature
- p. 123-132
A Novel Video-CTU Registration Method with Structural Point Similarity for FURS Navigation
- 29 July 2023
- Published by Elsevier in Computerized Medical Imaging and Graphics
- Vol. 108, 102275
Learning how to detect: A deep reinforcement learning method for whole-slide melanoma histopathology images
- 29 July 2023
- Published by Elsevier in Computerized Medical Imaging and Graphics
- Vol. 108, 102275
Learning how to detect: A deep reinforcement learning method for whole-slide melanoma histopathology images
- Open Access
- 1 March 2021
- Published by AME Publishing Company in Translational Andrology and Urology
- Vol. 10  (3) , 1321-1331
Pazopanib in patients with metastatic renal cell carcinoma: a single-center, real-world, retrospective Chinese study
Background: The efficacy and safety of pazopanib in patients diagnosed with metastatic renal cell carcinoma (mRCC) have been demonstrated by a Chinese subgroup analysis of the COMPARZ (Pazopanib Versus Sunitinib in the Treatment of Locally Advanced and/or Metastatic Renal Cell Carcinoma) trial. However, the real-world data are still unknown. This single-center, retrospective study was designed to verify the real-world effects of pazopanib in Chinese patients with mRCC. Methods: Patients with mRCC and a clinical decision to initiate pazopanib as first-line therapy were eligible. The primary endpoint was progression-free survival (PFS), with overall survival (OS), objective response rate (ORR), and safety being evaluated as secondary endpoints. The effectiveness according to the International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) risk model, number of risk factors in the intermediate risk group, age, Eastern Cooperative Oncology Group (ECOG) performance status (PS), and the number and site of organ metastasis were also assessed. Results: A total of 32 patients were enrolled, including 23 (71.9%) males and 9 (28.1%) females. The median age was 57 years (range 29–75 years). With a median follow-up time of 23.8 months, a median PFS of 18.3 months, and an ORR of 37.5%. Median OS was not reached, and the 1-, 2-, and 3-year overall survival rates were 90.6%, 78.1, and 65.6%, respectively. According to IMDC risk model, 37.5% were placed in the favorable risk (FR) subgroup, 56.2% (the majority) were placed in the intermediate risk (IR) subgroup, and 6.3% were placed in the poor risk (PR) subgroup. Compared with the IR and PR groups, the FR group achieved the best ORR (58.3%) and median PFS (22.1 months). Having 1 risk factor, ECOG PS Conclusions: Pazopanib showed efficacy and safety in real-world Chinese mRCC patients.
- 1 March 2021
- Published by AME Publishing Company in Translational Andrology and Urology
- Vol. 10  (3) , 1321-1331
Open AccessPazopanib in patients with metastatic renal cell carcinoma: a single-center, real-world, retrospective Chinese study
Background: The efficacy and safety of pazopanib in patients diagnosed with metastatic renal cell carcinoma (mRCC) have been demonstrated by a Chinese subgroup analysis of the COMPARZ (Pazopanib Versus Sunitinib in the Treatment of Locally Advanced and/or Metastatic Renal Cell Carcinoma) trial. However, the real-world data are still unknown. This single-center, retrospective study was designed to verify the real-world effects of pazopanib in Chinese patients with mRCC. Methods: Patients with mRCC and a clinical decision to initiate pazopanib as first-line therapy were eligible. The primary endpoint was progression-free survival (PFS), with overall survival (OS), objective response rate (ORR), and safety being evaluated as secondary endpoints. The effectiveness according to the International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) risk model, number of risk factors in the intermediate risk group, age, Eastern Cooperative Oncology Group (ECOG) performance status (PS), and the number and site of organ metastasis were also assessed. Results: A total of 32 patients were enrolled, including 23 (71.9%) males and 9 (28.1%) females. The median age was 57 years (range 29–75 years). With a median follow-up time of 23.8 months, a median PFS of 18.3 months, and an ORR of 37.5%. Median OS was not reached, and the 1-, 2-, and 3-year overall survival rates were 90.6%, 78.1, and 65.6%, respectively. According to IMDC risk model, 37.5% were placed in the favorable risk (FR) subgroup, 56.2% (the majority) were placed in the intermediate risk (IR) subgroup, and 6.3% were placed in the poor risk (PR) subgroup. Compared with the IR and PR groups, the FR group achieved the best ORR (58.3%) and median PFS (22.1 months). Having 1 risk factor, ECOG PS Conclusions: Pazopanib showed efficacy and safety in real-world Chinese mRCC patients.
- Open Access
- 19 June 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 114088-114099
A Generic Semi-Supervised Deep Learning-Based Approach for Automated Surface Inspection
Automated surface inspection (ASI) is critical to quality control in industrial manufacturing processes. Recent advances in deep learning have produced new ASI methods that automatically learn high-level features from training samples while being robust to changes and capable of detecting different types of surfaces and defects. However, they usually rely heavily on manpower to collect and label training samples. In this paper, a generic semi-supervised deep learning-based approach for ASI that requires a small quantity of labeled training data is proposed. While the approach follows the MixMatch rules to conduct sophisticated data augmentation, we introduce a new loss function calculation method and propose a new convolutional neural network based on a residual structure to achieve accurate defect detection. An experiment on two public datasets (DAGM and NEU) and one industrial dataset (CCL) is carried out. For public datasets, the experimental results are compared against several best benchmarks in the literature. For the industrial dataset, the results are compared against deep learning methods based on benchmark neural networks. The proposed method achieves the best performance in all comparisons. In addition, a comparative experiment of model performance given a different number of labeled samples is conducted, demonstrating that the proposed method can achieve good performance with few labeled training samples.
- 19 June 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 114088-114099
Open AccessA Generic Semi-Supervised Deep Learning-Based Approach for Automated Surface Inspection
Automated surface inspection (ASI) is critical to quality control in industrial manufacturing processes. Recent advances in deep learning have produced new ASI methods that automatically learn high-level features from training samples while being robust to changes and capable of detecting different types of surfaces and defects. However, they usually rely heavily on manpower to collect and label training samples. In this paper, a generic semi-supervised deep learning-based approach for ASI that requires a small quantity of labeled training data is proposed. While the approach follows the MixMatch rules to conduct sophisticated data augmentation, we introduce a new loss function calculation method and propose a new convolutional neural network based on a residual structure to achieve accurate defect detection. An experiment on two public datasets (DAGM and NEU) and one industrial dataset (CCL) is carried out. For public datasets, the experimental results are compared against several best benchmarks in the literature. For the industrial dataset, the results are compared against deep learning methods based on benchmark neural networks. The proposed method achieves the best performance in all comparisons. In addition, a comparative experiment of model performance given a different number of labeled samples is conducted, demonstrating that the proposed method can achieve good performance with few labeled training samples.
- Open Access
Application of Evolutionary Game Theory in Safety Management of Chemical Production
The chemical industry is essential in the social economy, and the issue of production safety has aroused widespread concern. Chemical safety incidents occupy the headlines from time to time, and chemical production safety management is particularly important. This paper presents an application model based on evolutionary game theory in the assessment and analysis of chemical production safety management. The model uses evolutionary game theory to construct a strategic interactive payoff matrix between the management department of a chemical plant and the chemical plant using a replicated dynamic equation to analyze their strategic interaction and to reveal the evolution of behavioral strategy selection. The evolution results were verified and simulated. The application of this model provides an effective safety management basis and recommendations for the management of chemical plants, providing a foundation for the safe production and healthy development of chemical plants.
Open AccessApplication of Evolutionary Game Theory in Safety Management of Chemical Production
The chemical industry is essential in the social economy, and the issue of production safety has aroused widespread concern. Chemical safety incidents occupy the headlines from time to time, and chemical production safety management is particularly important. This paper presents an application model based on evolutionary game theory in the assessment and analysis of chemical production safety management. The model uses evolutionary game theory to construct a strategic interactive payoff matrix between the management department of a chemical plant and the chemical plant using a replicated dynamic equation to analyze their strategic interaction and to reveal the evolution of behavioral strategy selection. The evolution results were verified and simulated. The application of this model provides an effective safety management basis and recommendations for the management of chemical plants, providing a foundation for the safe production and healthy development of chemical plants.