Ying Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ying Wang has a total of 128 co-authors. Here are the top co-authors by number of shared publications:
Jie Yu
Department of Nuclear Medicine, The Fifth Affiliated Hospital, Sun Yat-sen University, Zhuhai 519000, China
5 shared publications- Hongjun Jin
Guangdong Provincial Engineering Research Center of Molecular Imaging, the Fifth Affiliated Hospital, Sun Yat-sen University, Zhuhai 519000, China
4 shared publications Min Yang
Department of Nuclear Medicine, The Fifth Affiliated Hospital, Sun Yat-sen University, Zhuhai 519000, China
4 shared publicationsQingling Chen
Department of Nuclear Medicine, The Fifth Affiliated Hospital of Sun Yat-Sen University, Sun Yat-Sen University, Zhuhai, Guangdong, People’s Republic of China
4 shared publicationsFanwei Zhang
Sun Yat Sen Univ, Affiliated Hosp 5, Dept Nucl Med, 52 Meihua E Rd, Zhuhai 519099, Peoples R China
3 shared publications- Guojie Wang
Department of Radiology, Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, China
3 shared publications
Latest Publications
- 1 April 2026
- Published by Elsevier in Experimental Neurology
- Vol. 401, 115754
Altered glucose metabolic pattern in basal ganglia: 18F-FDG PET/CT semi-quantitative analysis of metabolic signatures in neonatal bilirubin encephalopathy
- 1 April 2026
- Published by Elsevier in Experimental Neurology
- Vol. 401, 115754
Altered glucose metabolic pattern in basal ganglia: 18F-FDG PET/CT semi-quantitative analysis of metabolic signatures in neonatal bilirubin encephalopathy
- 22 January 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 110, 103956
Anatomy-guided prompting with cross-modal self-alignment for whole-body PET-CT breast cancer segmentation
- 22 January 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 110, 103956
Anatomy-guided prompting with cross-modal self-alignment for whole-body PET-CT breast cancer segmentation
- 6 January 2026
- Published by Elsevier in Practical Radiation Oncology
- Vol. 16 (1) , 1-2
Rare Findings of Skin and Muscle Involvement in Systemic Sclerosis With Gastric Cancer
- 6 January 2026
- Published by Elsevier in Practical Radiation Oncology
- Vol. 16 (1) , 1-2
Rare Findings of Skin and Muscle Involvement in Systemic Sclerosis With Gastric Cancer
- Open Access
- 1 December 2025
- Published by Taylor & Francis in Neuropsychiatric Disease and Treatment
- Vol. ume 21, 2783-2793
Leveraging TabTransformer Deep Learning on Conventional MRI Radiomics for Accessible and Interpretable Diagnosis of Autism Spectrum Disorder
Purpose: This study aims to assess the diagnostic efficacy of a multi-region radiomics analysis utilizing conventional MRI sequences (T1-weighted imaging [T1WI] and T2-weighted imaging [T2WI]) for autism spectrum disorder (ASD), and to investigate the correlations between radiomics features and the severity of clinical symptoms, thereby exploring potential imaging biomarkers. Methods: This retrospective study included 207 pediatric participants (91 ASD, 116 typically developing controls). Radiomics features were extracted from manually segmented bilateral hippocampus, thalamus, caudate nucleus, and lenticular nucleus on T1WI and T2WI images. Three distinct classifiers (T1WI-only, T2WI-only, T1WI+T2WI combined) were developed using logistic regression (LR), support vector machine (SVM), and a TabTransformer deep learning (DL) model. Diagnostic performance was evaluated via five-fold cross-validation. Results: The TabTransformer DL model utilizing combined T1WI+T2WI features demonstrated superior performance, achieving an area under the curve of 0.900, accuracy of 0.834, sensitivity of 0.843, and specificity of 0.823. Specific radiomic features, predominantly from the left lentiform nucleus and bilateral caudate nucleus, were significantly correlated with clinical severity scores (ABC, CARS). Conclusion: Radiomics models leveraging routine MRI sequences demonstrate robust diagnostic utility for ASD. The identified subcortical features, correlating with core symptoms, may serve as viable imaging biomarkers. Future work requires external validation, exploration of automated segmentation, and investigation in larger, multi-center cohorts..
- 1 December 2025
- Published by Taylor & Francis in Neuropsychiatric Disease and Treatment
- Vol. ume 21, 2783-2793
Open AccessLeveraging TabTransformer Deep Learning on Conventional MRI Radiomics for Accessible and Interpretable Diagnosis of Autism Spectrum Disorder
Purpose: This study aims to assess the diagnostic efficacy of a multi-region radiomics analysis utilizing conventional MRI sequences (T1-weighted imaging [T1WI] and T2-weighted imaging [T2WI]) for autism spectrum disorder (ASD), and to investigate the correlations between radiomics features and the severity of clinical symptoms, thereby exploring potential imaging biomarkers. Methods: This retrospective study included 207 pediatric participants (91 ASD, 116 typically developing controls). Radiomics features were extracted from manually segmented bilateral hippocampus, thalamus, caudate nucleus, and lenticular nucleus on T1WI and T2WI images. Three distinct classifiers (T1WI-only, T2WI-only, T1WI+T2WI combined) were developed using logistic regression (LR), support vector machine (SVM), and a TabTransformer deep learning (DL) model. Diagnostic performance was evaluated via five-fold cross-validation. Results: The TabTransformer DL model utilizing combined T1WI+T2WI features demonstrated superior performance, achieving an area under the curve of 0.900, accuracy of 0.834, sensitivity of 0.843, and specificity of 0.823. Specific radiomic features, predominantly from the left lentiform nucleus and bilateral caudate nucleus, were significantly correlated with clinical severity scores (ABC, CARS). Conclusion: Radiomics models leveraging routine MRI sequences demonstrate robust diagnostic utility for ASD. The identified subcortical features, correlating with core symptoms, may serve as viable imaging biomarkers. Future work requires external validation, exploration of automated segmentation, and investigation in larger, multi-center cohorts..
- 25 October 2025
- Published by Springer Nature in European Journal of Nuclear Medicine and Molecular Imaging
- Vol. 53 (4) , 2387-2402
NHOC/NHOP as novel biomarkers for predicting lymph node metastasis in NSCLC using PET/CT radiomics and machine learning: a two-center retrospective study
Purpose NHOC and NHOP, defined as the normalized distances from peak uptake to tumour centroid and perimeter, are novel PET/CT metrics of tumour aggressiveness. This two-centre study assessed the baseline NHOC/NHOP for predicting lymph node metastasis (LNM) in non-small cell lung cancer (NSCLC), then developed and validated an interpretable machine learning model combining clinical data, NHOC/NHOP and PET radiomics for LNM and occult nodal metastasis (ONM) prediction. Methods 342 patients from two centres underwent 18F-FDG PET/CT scans, and data were divided into training (n = 188), internal (n = 63), and external (n = 91) sets. NHOC/NHOP and 284 radiomics features were initially extracted using LIFEx software. These features were normalized using Z-score and harmonized via ComBat. To avoid single algorithmic bias, eight machine-learning models were trained on the optimal radiomics features. The best-performing algorithm was employed to develop four predictive models including clinical, NHOC/NHOP, radiomics, and their combination. Shapley Additive Explanations (SHAP) values were used to interpret model contributions. Results Key independent predictors were PD-L1 value, lesion size and the novel biomarker NHOC, establishing the clinical model (PD-L1 and size) and the NHOC model. The multi-layer perceptron classifier (MLP) model achieved the highest Area Under the Curve (AUC) (0.82, 95% CI: 0.69–0.92). For LNM prediction, the combined model demonstrated superior performance across training (AUC 0.852), internal test (AUC 0.822), and external test (AUC 0.885) sets. It significantly outperformed clinical and NHOC models (p < 0.05). For ONM prediction, the combined model achieved the AUC (0.85) on the full datasets. SHAP analysis highlighted key features like GLCM_InverseVariance-PET and NGTDM_Strength-CT. A nomogram and online calculator were developed, with decision-curve analysis confirming superior net clinical benefit. Conclusion This study established an accurate, interpretable machine learning model for preoperative prediction of LNM and ONM in NSCLC. NHOC emerged as a novel independent predictor with respect to classical PET parameters.
- 25 October 2025
- Published by Springer Nature in European Journal of Nuclear Medicine and Molecular Imaging
- Vol. 53 (4) , 2387-2402
NHOC/NHOP as novel biomarkers for predicting lymph node metastasis in NSCLC using PET/CT radiomics and machine learning: a two-center retrospective study
Purpose NHOC and NHOP, defined as the normalized distances from peak uptake to tumour centroid and perimeter, are novel PET/CT metrics of tumour aggressiveness. This two-centre study assessed the baseline NHOC/NHOP for predicting lymph node metastasis (LNM) in non-small cell lung cancer (NSCLC), then developed and validated an interpretable machine learning model combining clinical data, NHOC/NHOP and PET radiomics for LNM and occult nodal metastasis (ONM) prediction. Methods 342 patients from two centres underwent 18F-FDG PET/CT scans, and data were divided into training (n = 188), internal (n = 63), and external (n = 91) sets. NHOC/NHOP and 284 radiomics features were initially extracted using LIFEx software. These features were normalized using Z-score and harmonized via ComBat. To avoid single algorithmic bias, eight machine-learning models were trained on the optimal radiomics features. The best-performing algorithm was employed to develop four predictive models including clinical, NHOC/NHOP, radiomics, and their combination. Shapley Additive Explanations (SHAP) values were used to interpret model contributions. Results Key independent predictors were PD-L1 value, lesion size and the novel biomarker NHOC, establishing the clinical model (PD-L1 and size) and the NHOC model. The multi-layer perceptron classifier (MLP) model achieved the highest Area Under the Curve (AUC) (0.82, 95% CI: 0.69–0.92). For LNM prediction, the combined model demonstrated superior performance across training (AUC 0.852), internal test (AUC 0.822), and external test (AUC 0.885) sets. It significantly outperformed clinical and NHOC models (p < 0.05). For ONM prediction, the combined model achieved the AUC (0.85) on the full datasets. SHAP analysis highlighted key features like GLCM_InverseVariance-PET and NGTDM_Strength-CT. A nomogram and online calculator were developed, with decision-curve analysis confirming superior net clinical benefit. Conclusion This study established an accurate, interpretable machine learning model for preoperative prediction of LNM and ONM in NSCLC. NHOC emerged as a novel independent predictor with respect to classical PET parameters.
- 26 March 2025
- Published by Elsevier in Practical Radiation Oncology
- Vol. 15 (2) , 111-112
Primary Cutaneous Diffuse Large B-Cell Lymphoma, Leg Type
- 26 March 2025
- Published by Elsevier in Practical Radiation Oncology
- Vol. 15 (2) , 111-112
Primary Cutaneous Diffuse Large B-Cell Lymphoma, Leg Type
- 12 February 2025
- Published by Springer Nature
- p. 23-31
DuEU-Net: Dual Encoder UNet with Modality-Agnostic Training for PET-CT Multi-modal Organ and Lesion Segmentation
- 12 February 2025
- Published by Springer Nature
- p. 23-31
DuEU-Net: Dual Encoder UNet with Modality-Agnostic Training for PET-CT Multi-modal Organ and Lesion Segmentation
- Top Cited PapersOpen Access
- 21 January 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-20
Hypoxia-tropic delivery of nanozymes targeting transferrin receptor 1 for nasopharyngeal carcinoma radiotherapy sensitization
Nasopharyngeal carcinoma (NPC), a malignancy highly prevalent in East and Southeast Asia, is primarily treated with radiotherapy (RT). However, hypoxia-induced radioresistance presents a significant challenge. Nanozymes, nanomaterials with catalase-like activity, have emerged as a promising strategy for radiosensitization by converting elevated hydrogen peroxide in the tumor microenvironment into oxygen. Despite their potential, effectively targeting hypoxic lesions has been difficult. Here, we identify transferrin receptor 1 (TfR1) as an upregulated target in NPC, with its expression levels positively correlated with hypoxia. Human heavy-chain ferritin, a specific ligand of TfR1, selectively recognizes hypoxic NPC lesions in preclinical models. Based on these findings, we design a hypoxia-targeted nanozyme by loading platinum nanoparticles into ferritin. This nanozyme exhibits enhanced catalase-like activity and effectively alleviates tumor hypoxia in NPC xenografts. When combined with RT, a single injection of the nanozyme significantly inhibits tumor growth and prolongs mouse survival, outperforming sodium glycididazole, a clinically used radiosensitizer. In summary, our findings highlight TfR1 as an accessible cell surface target in hypoxic NPC lesions. The nanozyme targeting TfR1 holds promise for enhancing the therapeutic effectiveness of RT in NPC through an in situ oxygen-generation mechanism.
- 21 January 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-20
Top Cited PapersOpen AccessHypoxia-tropic delivery of nanozymes targeting transferrin receptor 1 for nasopharyngeal carcinoma radiotherapy sensitization
Nasopharyngeal carcinoma (NPC), a malignancy highly prevalent in East and Southeast Asia, is primarily treated with radiotherapy (RT). However, hypoxia-induced radioresistance presents a significant challenge. Nanozymes, nanomaterials with catalase-like activity, have emerged as a promising strategy for radiosensitization by converting elevated hydrogen peroxide in the tumor microenvironment into oxygen. Despite their potential, effectively targeting hypoxic lesions has been difficult. Here, we identify transferrin receptor 1 (TfR1) as an upregulated target in NPC, with its expression levels positively correlated with hypoxia. Human heavy-chain ferritin, a specific ligand of TfR1, selectively recognizes hypoxic NPC lesions in preclinical models. Based on these findings, we design a hypoxia-targeted nanozyme by loading platinum nanoparticles into ferritin. This nanozyme exhibits enhanced catalase-like activity and effectively alleviates tumor hypoxia in NPC xenografts. When combined with RT, a single injection of the nanozyme significantly inhibits tumor growth and prolongs mouse survival, outperforming sodium glycididazole, a clinically used radiosensitizer. In summary, our findings highlight TfR1 as an accessible cell surface target in hypoxic NPC lesions. The nanozyme targeting TfR1 holds promise for enhancing the therapeutic effectiveness of RT in NPC through an in situ oxygen-generation mechanism.
- Open Access
- 1 December 2024
- Published by AME Publishing Company in Quantitative Imaging in Medicine and Surgery
- Vol. 14 (12) , 8140-8154
Deep learning-based low count whole-body positron emission tomography denoising incorporating computed tomography priors
Background: Deep-learning-based denoising improves image quality and quantification accuracy for low count (LC) positron emission tomography (PET). Conventional deep-learning-based denoising methods only require single LC PET image input. This study aims to propose a deep-learning-based LC PET denoising method incorporating computed tomography (CT) priors to further reduce the dose level. Methods: Fifty patients who underwent their routine whole-body 2-deoxy-2-[18F]fluoro-D-glucose (18F-FDG) PET/CT scans in March 2022 were retrospectively and non-consecutively recruited. For full count (FC) PET, patients were injected with 3.7 MBq/kg FDG and scanned for 5 bed positions with 2 min/bed. LC PET of 1/10 (LC-10) and 1/20 (LC-20) count levels of FC PET were obtained by randomly down-sampling the FC list mode data, which were then paired with FC PET for training U-Net (U-Net-1) and cGAN (cGAN-1). Networks incorporated CT images as input (U-Net-2 and cGAN-2) were also implemented. Quantitative analysis of physical and clinical indices was performed and statistically assessed with Wilcoxon sign-rank test with Bonferroni correction. Results: Mean square error and structural similarity index were the best for cGAN-2, followed by U-Net-2, cGAN-1 and U-Net-1. The errors of mean standardized uptake value (SUV) and maximum SUV were lowest for cGAN-2, followed by cGAN-1, U-Net-2 and U-Net-1. For cGAN-2, image quality and lesion detectability scores were 3.71±0.94 and 4.25±0.83 for LC-10, 3.57±0.79 and 3.61±1.21 for LC-20, while they were 3.49±0.92 and 4.42±0.08 for FC. Notably, some small lesions were “masked out” on cGAN/U-Net-1 but can be retrieved on cGAN/U-Net-2 denoised PET for LC-20. Conclusions: Deep-learning-based LC PET denoising incorporating CT priors is more effective than conventional deep-learning-based denoising with single LC PET input, especially at lower dose levels.
- 1 December 2024
- Published by AME Publishing Company in Quantitative Imaging in Medicine and Surgery
- Vol. 14 (12) , 8140-8154
Open AccessDeep learning-based low count whole-body positron emission tomography denoising incorporating computed tomography priors
Background: Deep-learning-based denoising improves image quality and quantification accuracy for low count (LC) positron emission tomography (PET). Conventional deep-learning-based denoising methods only require single LC PET image input. This study aims to propose a deep-learning-based LC PET denoising method incorporating computed tomography (CT) priors to further reduce the dose level. Methods: Fifty patients who underwent their routine whole-body 2-deoxy-2-[18F]fluoro-D-glucose (18F-FDG) PET/CT scans in March 2022 were retrospectively and non-consecutively recruited. For full count (FC) PET, patients were injected with 3.7 MBq/kg FDG and scanned for 5 bed positions with 2 min/bed. LC PET of 1/10 (LC-10) and 1/20 (LC-20) count levels of FC PET were obtained by randomly down-sampling the FC list mode data, which were then paired with FC PET for training U-Net (U-Net-1) and cGAN (cGAN-1). Networks incorporated CT images as input (U-Net-2 and cGAN-2) were also implemented. Quantitative analysis of physical and clinical indices was performed and statistically assessed with Wilcoxon sign-rank test with Bonferroni correction. Results: Mean square error and structural similarity index were the best for cGAN-2, followed by U-Net-2, cGAN-1 and U-Net-1. The errors of mean standardized uptake value (SUV) and maximum SUV were lowest for cGAN-2, followed by cGAN-1, U-Net-2 and U-Net-1. For cGAN-2, image quality and lesion detectability scores were 3.71±0.94 and 4.25±0.83 for LC-10, 3.57±0.79 and 3.61±1.21 for LC-20, while they were 3.49±0.92 and 4.42±0.08 for FC. Notably, some small lesions were “masked out” on cGAN/U-Net-1 but can be retrieved on cGAN/U-Net-2 denoised PET for LC-20. Conclusions: Deep-learning-based LC PET denoising incorporating CT priors is more effective than conventional deep-learning-based denoising with single LC PET input, especially at lower dose levels.
- Open Access
- 5 April 2024
- Published by American Chemical Society (ACS) in ACS Omega
- Vol. 9 (15) , 17423-17431
Synthesis and Evaluation of [64Cu]Cu-NOTA-HFn for PET Imaging of Transferrin Receptor 1 Expression in Nasopharyngeal Carcinoma
As recurrent and metastatic nasopharyngeal carcinoma (NPC) is the most common cause of death among patients with NPC, there is an urgent clinical need for the development of precision diagnosis to guide personalized treatment. Recent emerging evidence substantiates the increased expression of transferrin receptor 1 (also known as cluster of differentiation 71, CD71) within tumor tissues and the inherent targeting capability of natural heavy-chain ferritin (HFn) toward CD71. This study aimed to synthesize and assess a radiotracer ([64Cu]Cu-NOTA-HFn) designed to target CD71 for positron emission tomography (PET) imaging in an NPC tumor-bearing mouse model. The entire radiolabeling process of [64Cu]Cu-NOTA-HFn was completed within 15 min with high yield (>98.5%) and high molar activity (72.96 ± 21.33 GBq/μmol). The in vitro solubility and stability experiments indicated that [64Cu]Cu-NOTA-HFn had a high water solubility (log P = -2.42 ± 0.52, n = 6) and good stability in phosphate-buffered saline (PBS) for up to 48 h. The cell saturation binding assay indicated that [64Cu]Cu-NOTA-HFn had a nanomolar affinity (Kd = 10.9 ± 6.1 nM) for CD71-overexpressing C666-1 cells. To test the target engagement in vivo, prolonged-time PET imaging was performed at 1, 6, 12, 24, and 36 h postinjection (p.i.) of [64Cu]Cu-NOTA-HFn to C666-1 NPC tumor-bearing mice. The C666-1 tumors could be visualized by [64Cu]Cu-NOTA-HFn and blocked by nonradiolabeled HFn. PET imaging quantitative analysis demonstrated that the uptake of [64Cu]Cu-NOTA-HFn in C666-1 tumors peaked at 6 h p.i. and the best radioactive tumor-to-muscle ratio was 10.53 ± 3.11 (n = 3). Ex vivo biodistribution assay at 6 h p.i. showed that the tumor uptakes were 1.43 ± 0.23%ID/g in the nonblock group and 0.92 ± 0.2%ID/g in the block group (n = 3, p < 0.05). Immunohistochemistry and immunofluorescence staining confirmed positive expression of CD71 and the uptake of HFn in C666-1 tumor tissues. In conclusion, our experiments demonstrated that [64Cu]Cu-NOTA-HFn possesses a very high target engagement for CD71-positive NPC tumors and provided a fundamental basis for further clinical translation.
- 5 April 2024
- Published by American Chemical Society (ACS) in ACS Omega
- Vol. 9 (15) , 17423-17431
Open AccessSynthesis and Evaluation of [64Cu]Cu-NOTA-HFn for PET Imaging of Transferrin Receptor 1 Expression in Nasopharyngeal Carcinoma
As recurrent and metastatic nasopharyngeal carcinoma (NPC) is the most common cause of death among patients with NPC, there is an urgent clinical need for the development of precision diagnosis to guide personalized treatment. Recent emerging evidence substantiates the increased expression of transferrin receptor 1 (also known as cluster of differentiation 71, CD71) within tumor tissues and the inherent targeting capability of natural heavy-chain ferritin (HFn) toward CD71. This study aimed to synthesize and assess a radiotracer ([64Cu]Cu-NOTA-HFn) designed to target CD71 for positron emission tomography (PET) imaging in an NPC tumor-bearing mouse model. The entire radiolabeling process of [64Cu]Cu-NOTA-HFn was completed within 15 min with high yield (>98.5%) and high molar activity (72.96 ± 21.33 GBq/μmol). The in vitro solubility and stability experiments indicated that [64Cu]Cu-NOTA-HFn had a high water solubility (log P = -2.42 ± 0.52, n = 6) and good stability in phosphate-buffered saline (PBS) for up to 48 h. The cell saturation binding assay indicated that [64Cu]Cu-NOTA-HFn had a nanomolar affinity (Kd = 10.9 ± 6.1 nM) for CD71-overexpressing C666-1 cells. To test the target engagement in vivo, prolonged-time PET imaging was performed at 1, 6, 12, 24, and 36 h postinjection (p.i.) of [64Cu]Cu-NOTA-HFn to C666-1 NPC tumor-bearing mice. The C666-1 tumors could be visualized by [64Cu]Cu-NOTA-HFn and blocked by nonradiolabeled HFn. PET imaging quantitative analysis demonstrated that the uptake of [64Cu]Cu-NOTA-HFn in C666-1 tumors peaked at 6 h p.i. and the best radioactive tumor-to-muscle ratio was 10.53 ± 3.11 (n = 3). Ex vivo biodistribution assay at 6 h p.i. showed that the tumor uptakes were 1.43 ± 0.23%ID/g in the nonblock group and 0.92 ± 0.2%ID/g in the block group (n = 3, p < 0.05). Immunohistochemistry and immunofluorescence staining confirmed positive expression of CD71 and the uptake of HFn in C666-1 tumor tissues. In conclusion, our experiments demonstrated that [64Cu]Cu-NOTA-HFn possesses a very high target engagement for CD71-positive NPC tumors and provided a fundamental basis for further clinical translation.