Zhengcan Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhengcan Wu has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Jianping Gu
Department of Interventional and Vascular Radiology, Nanjing First Hospital, Nanjing Medical University, Nanjing, People’s Republic of China
2 shared publications Maofeng Gong
2 shared publicationsRui Jiang
2 shared publicationsXu He
2 shared publicationsBoxiang Zhao
1 shared publication- Chengfei Cai
Taizhou University
1 shared publication
Latest Publications
- Open Access
- 19 May 2025
- Published by MDPI AG in Electronics
- Vol. 14 (10) , 2056
Classification of Whole-Slide Pathology Images Based on State Space Models and Graph Neural Networks
Whole-slide images (WSIs) pose significant analytical challenges due to their large data scale and complexity. Multiple instance learning (MIL) has emerged as an effective solution for WSI classification, but existing frameworks often lack flexibility in feature integration and underutilize sequential information. To address these limitations, this work proposes a novel MIL framework: Dynamic Graph and State Space Model-Based MIL (DG-SSM-MIL). DG-SSM-MIL combines graph neural networks and selective state space models, leveraging the former’s ability to extract local and spatial features and the latter’s advantage in comprehensively understanding long-sequence instances. This enhances the model’s performance in diverse instance classification, improves its capability to handle long-sequence data, and increases the precision and scalability of feature fusion. We propose the Dynamic Graph and State Space Model (DynGraph-SSM) module, which aggregates local and spatial information of image patches through directed graphs and learns global feature representations using the Mamba model. Additionally, the directed graph structure alleviates the unidirectional scanning limitation of Mamba and enhances its ability to process pathological images with dispersed lesion distributions. DG-SSM-MIL demonstrates superior performance in classification tasks compared to other models. We validate the effectiveness of the proposed method on features extracted from two pretrained models across four public medical image datasets: BRACS, TCGA-NSCLC, TCGA-RCC, and CAMELYON16. Experimental results demonstrate that DG-SSM-MIL consistently outperforms existing MIL methods across four public datasets. For example, when using ResNet-50 features, our model achieves the highest AUCs of 0.936, 0.785, 0.879, and 0.957 on TCGA-NSCLC, BRACS, CAMELYON16, and TCGA-RCC, respectively. Similarly, with UNI features, DG-SSM-MIL reaches AUCs of 0.968, 0.846, 0.993, and 0.990, surpassing all baselines. These results confirm the effectiveness and generalizability of our approach in diverse WSI classification tasks.
- 19 May 2025
- Published by MDPI AG in Electronics
- Vol. 14 (10) , 2056
Open AccessClassification of Whole-Slide Pathology Images Based on State Space Models and Graph Neural Networks
Whole-slide images (WSIs) pose significant analytical challenges due to their large data scale and complexity. Multiple instance learning (MIL) has emerged as an effective solution for WSI classification, but existing frameworks often lack flexibility in feature integration and underutilize sequential information. To address these limitations, this work proposes a novel MIL framework: Dynamic Graph and State Space Model-Based MIL (DG-SSM-MIL). DG-SSM-MIL combines graph neural networks and selective state space models, leveraging the former’s ability to extract local and spatial features and the latter’s advantage in comprehensively understanding long-sequence instances. This enhances the model’s performance in diverse instance classification, improves its capability to handle long-sequence data, and increases the precision and scalability of feature fusion. We propose the Dynamic Graph and State Space Model (DynGraph-SSM) module, which aggregates local and spatial information of image patches through directed graphs and learns global feature representations using the Mamba model. Additionally, the directed graph structure alleviates the unidirectional scanning limitation of Mamba and enhances its ability to process pathological images with dispersed lesion distributions. DG-SSM-MIL demonstrates superior performance in classification tasks compared to other models. We validate the effectiveness of the proposed method on features extracted from two pretrained models across four public medical image datasets: BRACS, TCGA-NSCLC, TCGA-RCC, and CAMELYON16. Experimental results demonstrate that DG-SSM-MIL consistently outperforms existing MIL methods across four public datasets. For example, when using ResNet-50 features, our model achieves the highest AUCs of 0.936, 0.785, 0.879, and 0.957 on TCGA-NSCLC, BRACS, CAMELYON16, and TCGA-RCC, respectively. Similarly, with UNI features, DG-SSM-MIL reaches AUCs of 0.968, 0.846, 0.993, and 0.990, surpassing all baselines. These results confirm the effectiveness and generalizability of our approach in diverse WSI classification tasks.
- Open Access
- 1 May 2025
- Published by AME Publishing Company in Quantitative Imaging in Medicine and Surgery
- Vol. 15 (5) , 4193-4203
Association of computed tomography-imaged common iliac vein compression with inferior vena cava thrombosis: a propensity score-matched analysis
Background: Investigating the association between common iliac vein (CIV) compression and inferior vena cava (IVC) thrombosis could have important clinical implications, such as a potential reduction in the use of IVC filters in low-risk patients and guidance for decision-making in thrombosis management. This study aimed to clarify the impact of the CIV compression degree on the incidence of IVC thrombosis in patients with lower extremity deep vein thrombosis (DVT). Methods: This single-center, retrospective study was performed between January 2015 and July 2023. A total of 391 eligible patients (mean age 60.42±16.67 years; 53.96% male) were included. The quantitative evaluation included the minimum diameter of the CIV, and the compression percentage was calculated on computed tomography venography. Univariable analysis and subsequent multivariable analysis, as well as propensity score matching (PSM), were used to analyze odds ratios (ORs) with 95% confidence intervals (CIs). The association between IVC thrombosis and CIV compression on a continuous scale was evaluated via restricted cubic splines (RCSs). Results: Iliofemoral DVT and a larger minimum diameter were associated with an increased risk of IVC thrombosis, whereas older age and a higher compression percentage were linked to a decreased risk. After adjustments for age, male sex, hypertension, and iliofemoral DVT via PSM, the statistically significant associations of CIV minimum diameter (adjusted OR =1.10, 95% CI: 1.02–1.18; PConclusions: Compared with no or mild compression, moderate and severe CIV compression was associated with a reduced risk of IVC thrombosis. A higher CIV compression degree was consistently associated with a decreased IVC thrombosis risk when the minimum diameter was 49.81%.
- 1 May 2025
- Published by AME Publishing Company in Quantitative Imaging in Medicine and Surgery
- Vol. 15 (5) , 4193-4203
Open AccessAssociation of computed tomography-imaged common iliac vein compression with inferior vena cava thrombosis: a propensity score-matched analysis
Background: Investigating the association between common iliac vein (CIV) compression and inferior vena cava (IVC) thrombosis could have important clinical implications, such as a potential reduction in the use of IVC filters in low-risk patients and guidance for decision-making in thrombosis management. This study aimed to clarify the impact of the CIV compression degree on the incidence of IVC thrombosis in patients with lower extremity deep vein thrombosis (DVT). Methods: This single-center, retrospective study was performed between January 2015 and July 2023. A total of 391 eligible patients (mean age 60.42±16.67 years; 53.96% male) were included. The quantitative evaluation included the minimum diameter of the CIV, and the compression percentage was calculated on computed tomography venography. Univariable analysis and subsequent multivariable analysis, as well as propensity score matching (PSM), were used to analyze odds ratios (ORs) with 95% confidence intervals (CIs). The association between IVC thrombosis and CIV compression on a continuous scale was evaluated via restricted cubic splines (RCSs). Results: Iliofemoral DVT and a larger minimum diameter were associated with an increased risk of IVC thrombosis, whereas older age and a higher compression percentage were linked to a decreased risk. After adjustments for age, male sex, hypertension, and iliofemoral DVT via PSM, the statistically significant associations of CIV minimum diameter (adjusted OR =1.10, 95% CI: 1.02–1.18; PConclusions: Compared with no or mild compression, moderate and severe CIV compression was associated with a reduced risk of IVC thrombosis. A higher CIV compression degree was consistently associated with a decreased IVC thrombosis risk when the minimum diameter was 49.81%.
- 20 February 2025
- Published by Springer Nature in CardioVascular and Interventional Radiology
- Vol. 48 (3) , 304-313
Relationship Between Left Moderate or Severe Common Iliac Vein Compression and Occurrence of Involved Iliac Vein Thrombosis in Patients with Infrainguinal Deep Vein Thrombosis
Purpose To investigate the association between the degree of common iliac vein (CIV) compression and the incidence of iliac vein involvement thrombosis (IVT) in patients with infrainguinal deep vein thrombosis (DVT). Methods This was a single center retrospective study. Between January 2015 and July 2023, infrainguinal DVT patients with or without IVT were included. Patient baseline demographics, clinical characteristics, and radiographic variables were collected and analyzed. Univariate analysis followed by multivariate analysis was used for IVT to determine odds ratio (OR) with a 95% confidence interval (CI). The association of the degree of CIV compression and IVT was evaluated by restricted cubic spines (RCS) on a continuous scale. Results 222 left-sided DVT patients [140 (63.1%) patients diagnosed with IVT, and 82 (36.9%) without] were included. Univariate analyses revealed a higher prevalence of female gender (55.7 vs. 41.5%, p = .040) and CIV compression among patients with left-sided IVT compared to those without IVT. After adjusting for female gender, a strong positive association between IVT and moderate (compression degree > 50% and ≤ 75%) (adjusted OR 3.75; 95% CI 1.571–8.974; p = .003) to severe (> 75%) (adjusted OR 4.05; 95% CI 1.615–10.134; p = .003) CIV compression degrees remained, indicating moderate or severe compression significantly increased IVT risk. RCS showed that a greater CIV minimum diameter or a smaller compression percentage was negatively associated with IVT incidence at a CIV minimum diameter was > 4.1 mm (p < .001) or compression percentage was < 61.0% (p < .001). Conclusion Moderate and severe CIV compression were associated with increased IVT risk compared to no or mild compression. A decreasing degree of CIV compression was consistently associated with a decreasing risk of IVT when the minimum diameter was > 4.1 mm or the compression was < 61.0%. Graphical Abstract
- 20 February 2025
- Published by Springer Nature in CardioVascular and Interventional Radiology
- Vol. 48 (3) , 304-313
Relationship Between Left Moderate or Severe Common Iliac Vein Compression and Occurrence of Involved Iliac Vein Thrombosis in Patients with Infrainguinal Deep Vein Thrombosis
Purpose To investigate the association between the degree of common iliac vein (CIV) compression and the incidence of iliac vein involvement thrombosis (IVT) in patients with infrainguinal deep vein thrombosis (DVT). Methods This was a single center retrospective study. Between January 2015 and July 2023, infrainguinal DVT patients with or without IVT were included. Patient baseline demographics, clinical characteristics, and radiographic variables were collected and analyzed. Univariate analysis followed by multivariate analysis was used for IVT to determine odds ratio (OR) with a 95% confidence interval (CI). The association of the degree of CIV compression and IVT was evaluated by restricted cubic spines (RCS) on a continuous scale. Results 222 left-sided DVT patients [140 (63.1%) patients diagnosed with IVT, and 82 (36.9%) without] were included. Univariate analyses revealed a higher prevalence of female gender (55.7 vs. 41.5%, p = .040) and CIV compression among patients with left-sided IVT compared to those without IVT. After adjusting for female gender, a strong positive association between IVT and moderate (compression degree > 50% and ≤ 75%) (adjusted OR 3.75; 95% CI 1.571–8.974; p = .003) to severe (> 75%) (adjusted OR 4.05; 95% CI 1.615–10.134; p = .003) CIV compression degrees remained, indicating moderate or severe compression significantly increased IVT risk. RCS showed that a greater CIV minimum diameter or a smaller compression percentage was negatively associated with IVT incidence at a CIV minimum diameter was > 4.1 mm (p < .001) or compression percentage was < 61.0% (p < .001). Conclusion Moderate and severe CIV compression were associated with increased IVT risk compared to no or mild compression. A decreasing degree of CIV compression was consistently associated with a decreasing risk of IVT when the minimum diameter was > 4.1 mm or the compression was < 61.0%. Graphical Abstract