YWYong Wang

Yong Wang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Yong Wang has a total of 90 co-authors. Here are the top co-authors by number of shared publications:

  • Xuantong Gong

    6 shared publications
  • Jie Han

    Department of Ultrasound, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

    5 shared publications
  • Bo Wang

    Department of Ultrasound, National Cancer Center, National Clinical Research Center for Cancer, Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

    4 shared publications
  • Liwei Wang

    National Key Laboratory of General Artificial Intelligence, School of Intelligence Science and Technology, Peking University, Beijing, People’s Republic of China

    4 shared publications
  • Qingli Zhu

    Department of Ultrasound, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China

    4 shared publications
  • Zihan Niu

    Peking Union Medical College Hospital, Beijing, China

    4 shared publications

Latest Publications

  • A foundation generative model for breast ultrasound image analysis

    by Haojun Yu, Youcheng Li, Nan Zhang, Zihan Niu, Xuantong Gong, Yanwen Luo, Haotian Ye, Siyu He, Quanlin Wu, Wangyan Qin, et al.

    Foundation models have emerged as powerful tools for addressing various tasks in clinical settings. However, their potential development for breast ultrasound analysis remains untapped. Here we present BUSGen, the first foundation generative model designed for breast ultrasound image analysis. Pretrained on over 3.5 million breast ultrasound images, BUSGen has acquired extensive knowledge of breast structures, pathological features and clinical variations. With few-shot adaptation, BUSGen can generate repositories of realistic and informative task-specific data, facilitating the development of models for a wide range of downstream tasks. Extensive experiments highlight BUSGen's exceptional adaptability, significantly exceeding real-data-trained foundation models in breast cancer screening, diagnosis and prognosis. In breast cancer early diagnosis, our approach outperformed all board-certified radiologists (n = 9), achieving an average sensitivity improvement of 16.5% (P < 0.0001). In addition, we characterized the scaling effect of using synthetic data. Finally, BUSGen enabled de-identified data sharing, making progress forward in secure medical data utilization.


  • Open Access

    A Chain-of-thought Reasoning Breast Ultrasound Dataset Covering All Histopathology Categories

    by Haojun Yu, Youcheng Li, Zihan Niu, Nan Zhang, Xuantong Gong, Huan Li, Zhiying Zou, Haifeng Qi, Zhenxiao Cao, Zijie Lan, et al.

    Breast ultrasound (BUS) is an essential tool for diagnosing breast lesions, with millions of examinations per year. However, publicly available high-quality BUS benchmarks for AI development are limited in data scale and annotation richness. In this work, we present BUS-CoT, a BUS dataset for chain-of-thought (CoT) reasoning analysis, which contains 11,439 ultrasound images from 11,850 lesions and 4,838 patients, covering all 99 WHO-defined histopathology categories. For model training and evaluation, we provide a curated high-quality subset of 5,163 lesion-focused images annotated by experienced radiologists. To facilitate research on incentivizing CoT reasoning, we construct the reasoning processes based on observation, feature, diagnosis and pathology labels, annotated and verified by experienced experts. Moreover, by covering lesions of all histopathology types, we aim to facilitate robust AI systems in rare cases, which can be error-prone in clinical practice. The data and code are publicly available at https://doi.org/10.6084/m9.figshare.30838715.


  • Open Access

    Kupffer Phase Radiomics Signature in Sonazoid Contrast‐Enhanced Ultrasound Predicts Immunohistochemistry Marker Expression in Hepatocellular Carcinoma

    by Chen Li, Yuan Liu, Mingxiao Wu, Weide Dai, Jinghai Song, Yong Wang

    Purpose: Few studies have explored the value of radiomics signatures in predicting immunohistochemical (IHC) staining markers. This study aimed to investigate and validate radiomics models based on the Kupffer phase of Sonazoid contrast‐enhanced intraoperative ultrasonography (S‐CEUS) images for predicting IHC marker expression in hepatocellular carcinoma (HCC). Method: Overall, 113 consecutive patients diagnosed with HCC between November 2019 and May 2023 were retrospectively analyzed. Histopathological assessment included IHC staining for GS, CD10, GPC3, and HSP70. Radiomic features extracted from S‐CEUS images were selected and analyzed. A Naïve Bayes classifier was employed to predict IHC marker expression in HCC, using selected clinical biomarkers and radiomic features. Results: For GPC3, the radiomics classifier achieved a macro‐average area under the receiver operating characteristic curve (AUC) of 0.700, indicating strong performance. For GS, both radiomics and combined clinical‐radiomics classifiers exhibited strong discrimination (AUCs: 0.870 and 0.882, respectively). The radiomics classifier outperformed clinical biomarkers (total and direct bilirubin) in predicting CD10, with a macro‐average AUC of 0.834. However, its accuracy decreased for higher HSP70 marker expression levels (AUC: 0.694). These findings underscore the consistent effectiveness of radiomics across different IHC markers when compared to traditional clinical approaches. Conclusions: The Kupffer phase in the S‐CEUS‐based radiomics signature is an excellent biomarker for predicting IHC marker expression in patients with HCC.


  • Open Access

    A Foundational Generative Model for Breast Ultrasound Image Analysis

    by Liwei Wang, Haojun Yu, Youcheng Li, Nan Zhang, Zihan Niu, Xuantong Gong, Yanwen Luo, Haotian Ye, Siyu He, Quanlin Wu, et al.

    Foundational models have emerged as powerful tools for addressing various tasks in clinical settings. However, their potential development to breast ultrasound analysis remains untapped. In this paper, we present BUSGen, the first foundational generative model specifically designed for breast ultrasound image analysis. Pretrained on over 3.5 million breast ultrasound images, BUSGen has acquired extensive knowledge of breast structures, pathological features, and clinical variations. With few-shot adaptation, BUSGen can generate repositories of realistic and informative task-specific data, facilitating the development of models for a wide range of downstream tasks. Extensive experiments highlight BUSGen's exceptional adaptability, significantly exceeding real-data-trained foundational models in breast cancer screening, diagnosis, and prognosis. In breast cancer early diagnosis, our approach outperformed all board-certified radiologists (n=9), achieving an average sensitivity improvement of 16.5% (P-value<0.0001). Additionally, we characterized the scaling effect of using generated data which was as effective as the collected real-world data for training diagnostic models. Moreover, extensive experiments demonstrated that our approach improved the generalization ability of downstream models. Importantly, BUSGen protected patient privacy by enabling fully de-identified data sharing, making progress forward in secure medical data utilization. An online demo of BUSGen is available at https://aibus.bio.

  • Open Access

    A Foundational Generative Model for Breast Ultrasound Image Analysis

    by Haojun Yu, Youcheng Li, Nan Zhang, Zihan Niu, Xuantong Gong, Yanwen Luo, Haotian Ye, Siyu He, Quanlin Wu, Wangyan Qin, et al.

    Foundational models have emerged as powerful tools for addressing various tasks in clinical settings. However, their potential development to breast ultrasound analysis remains untapped. In this paper, we present BUSGen, the first foundational generative model specifically designed for breast ultrasound image analysis. Pretrained on over 3.5 million breast ultrasound images, BUSGen has acquired extensive knowledge of breast structures, pathological features, and clinical variations. With few-shot adaptation, BUSGen can generate repositories of realistic and informative task-specific data, facilitating the development of models for a wide range of downstream tasks. Extensive experiments highlight BUSGen's exceptional adaptability, significantly exceeding real-data-trained foundational models in breast cancer screening, diagnosis, and prognosis. In breast cancer early diagnosis, our approach outperformed all board-certified radiologists (n=9), achieving an average sensitivity improvement of 16.5% (P-value<0.0001). Additionally, we characterized the scaling effect of using generated data which was as effective as the collected real-world data for training diagnostic models. Moreover, extensive experiments demonstrated that our approach improved the generalization ability of downstream models. Importantly, BUSGen protected patient privacy by enabling fully de-identified data sharing, making progress forward in secure medical data utilization. An online demo of BUSGen is available at https://aibus.bio.

  • Open Access

    A prospective study on using shear wave elastography to predict the ypT0 stage of rectal cancer after neoadjuvant therapy: a new support for the watch-and-wait approach?

    by Mengjia Liu, Ningyi Cui, Chao Sun, Xuantong Gong, Bo Wang, Di Yang, Yong Wang

    Introduction: The diagnostic accuracy of traditional imaging examination in predicting ypT stage of rectal cancer after neoadjuvant therapy is significantly reduced, which would affect patients’ subsequent treatment choices. This study aimed to investigate the use of endorectal shear wave elastography (SWE) for diagnosing ypT0 stage of rectal cancer after neoadjuvant chemoradiotherapy (nCRT).Methods: Sixty patients with rectal cancer were prospectively recruited in this study. Data on endorectal ultrasound (ERUS) and SWE parameters were collected before nCRT and 6–8 weeks after nCRT. Postoperative pathological results were the gold standard for evaluating the diagnostic accuracy of SWE and ERUS in predicting the ypT0 stage of rectal cancer after nCRT. Receiver operating characteristic (ROC) curve analysis was used to determine the cut-off values of the SWE parameters that best corresponded to the ypT0 stage and analyze the sensitivity, specificity, and accuracy.Results: The diagnostic accuracies of using ERUS to predict the ypT and ypT0 stages of rectal cancer after nCRT were 58.1% (18/31) and 64.3% (9/14), respectively. The ROC curve was constructed with the lesion’s Emean, Emean corrected (EC), Emean difference (ED), Emean corrected differencede (ECD), Emean descendding rate (EDR) and Emean corrected descendding rate (ECDR) values after nCRT, the cut-off values of diagnosing the ypT0 stage were 64.40 kPa, 55.45 kPa, 72.55 kPa, 73.75 kPa, 50.15%, and 55.93%, respectively; the area under the curve (AUC) for diagnosing the ypT0 stage was 0.924, 0.933, 0.748, 0.729, 0.857 and 0.861, respectively. The EC value showed the best diagnostic performance.Conclusion: SWE could improve the accuracy of conventional ERUS in diagnosing the ypT0 stage of rectal cancer after nCRT. It is expected to become a new method to help predict pathological complete responses in clinical practice and provide new evidence for the watch-and-wait approach.


  • Manipulating Redox Homeostasis of Cancer Stem Cells Overcome Chemotherapeutic Resistance through Photoactivatable Biomimetic Nanodiscs

    by Bo Wang, Wuwan Wang, Yunxue Xu, Renfa Liu, Rui Li, Peipei Yang, Chenyang Zhao, Zhifei Dai, Yong Wang

    Tumor heterogeneity remains a significant obstacle in cancer therapy due to diverse cells with varying treatment responses. Cancer stem‐like cells (CSCs) contribute significantly to intratumor heterogeneity, characterized by high tumorigenicity and chemoresistance. CSCs reside in the depth of the tumor, possessing low reactive oxygen species (ROS) levels and robust antioxidant defense systems to maintain self‐renewal and stemness. A nanotherapeutic strategy is developed using tumor‐penetrating peptide iRGD‐modified high‐density lipoprotein (HDL)‐mimetic nanodiscs (IPCND) that ingeniously loaded with pyropheophorbide‐a (Ppa), bis (2‐hydroxyethyl) disulfide (S‐S), and camptothecin (CPT) by synthesizing two amphiphilic drug‐conjugated sphingomyelin derivatives. Photoactivatable Ppa can generate massive ROS which as intracellular signaling molecules effectively shut down self‐renewal and trigger differentiation of the CSCs, while S‐S is utilized to deplete GSH and sustainably imbalance redox homeostasis by reducing ROS clearance. Simultaneously, the depletion of GSH is accompanied by the release of CPT, which leads to subsequent cell death. This dual strategy successfully disturbed the redox equilibrium of CSCs, prompting their differentiation and boosting the ability of CPT to kill CSCs upon laser irradiation. Additionally, it demonstrated a synergistic anti‐cancer effect by concurrently eliminating therapeutically resistant CSCs and bulk tumor cells, effectively suppressing tumor growth in CSC‐enriched heterogeneous colon tumor mouse models.


  • The GRAPHS-CRAFITY score: a novel efficacy predictive tool for unresectable hepatocellular carcinoma treated with immunotherapy

    by Yi Yang, Lu Li, Ying Xu, Jingzhong Ouyang, Yanzhao Zhou, Feng Ye, Zhen Huang, Wen Zhang, Aiping Zhou, Xinming Zhao, et al.

    Objectives To investigate MR features associated with prognosis of unresectable HCC receiving immunotherapy and establish a MR feature-based scoring system to predict efficacy of immunotherapy. Methods This retrospective study included patients with unresectable HCC who received immunotherapy at 2 hospitals between August 2018 and February 2022. The last follow-up was October 2022. Clinical variables and MR features were assessed using univariate and multivariate Cox regression analyses. A new scoring system was constructed based on independent risk factors and the CRAFITY score consisting of AFP (≥ 100 ng/ml) and CRP (≥ 1 mg/dl). And the predictive performance of CRAFITY core and new score were compared by receiver-operating-characteristics curves (ROCs), area under ROCs (AUCs), and calibration curves. Results A total of 166 patients (55.6 ± 10.4 years) were included in training cohort and 77 patients (55.4 ± 10.7 years) were included in validation cohort. There were significant differences in BCLC stage, max size, macrovascular invasion, intratumoral artery, and enhancing capsule between the 2 groups. Based on independent risk factors (gross GRowtH type, intratumoral fAt, enhancing tumor caPsule, Sex and CRAFITY score), a novel efficacy predictive tool named the GRAPHS-CRAFITY score was developed to predict OS. The OS was significantly different among the 3 groups according to GRAPHS-CRAFITY score (p value < 0.001). The GRAPHS-CRAFITY score could predict tumor response and disease control (p value < 0.001, p value < 0.001). Conclusions The GRAPHS-CRAFITY score is a reliable and easily applicable tool to predict the efficacy of unresectable HCC receiving immunotherapy.


  • The MAPS-CRAFITY score: a novel efficacy predictive tool for unresectable hepatocellular carcinoma treated with targeted therapy plus immunotherapy

    by Jingzhong Ouyang, Yi Yang, Yanzhao Zhou, Feng Ye, Zhengzheng Wang, Qingjun Li, Ying Xu, Lu Li, Xinming Zhao, Wen Zhang, et al.

    Background Body composition parameters (BCPs) are associated with mortality in patients with hepatocellular carcinoma (HCC). Our purpose was to develop a practical scoring model by BCP and the CRAFITY score to predict the overall survival (OS) and tumor response of patients with HCC who received targeted therapy plus immunotherapy. Methods This retrospective study included 265 patients with HCC who received targeted therapy plus immunotherapy at 2 centers in China from August 2018 to February 2022. Univariate and multivariate Cox regression analyses were applied to analyze clinical factors and BCP. A scoring model based on independent risk factors was developed to predict OS and tumor response. Moreover, the model’s prediction was further validated by an external cohort. Results A total of 150 patients (55.5 ± 10.8 years) and 115 patients (55.0 ± 8.9 years) treated with lenvatinib or bevacizumab biosimilar plus anti-programmed death-1 (PD-1) antibody were included in training and validation cohorts, respectively. In the training cohort, independent predictive factors for OS included macrovascular invasion (p = 0.016), Child‒Pugh class (A vs. B, p = 0.001; A vs. C, p < 0.001), sarcopenia (p = 0.034), and the CRAFITY score (p = 0.011). Based on independent risk factors (MAcrovascular invasion, Child‒Pugh class, Sarcopenia, and the CRAFITY score) identified by multivariate analysis, a novel efficacy predictive tool named the MAPS-CRAFITY score was developed to predict OS. In all the training and validation cohorts, the OS differed significantly across the three groups based on the MAPS-CRAFITY score (< 2.1, 2.1–2.3, ≥ 2.4; p < 0.001). Moreover, the C-index of the MAPS-CRAFITY score was 0.720 and 0.761 in the training and validation cohorts, respectively. In both the validation and training cohorts, the MAPS-CRAFITY score was predictive of tumor response and disease control (p < 0.001). The AUCs of the MAPS-CRAFITY score for predicting disease control were 0.752 in the training cohort and 0.836 in the validation cohort. Conclusions The MAPS-CRAFITY score based on sarcopenia and the CRAFITY score is a reliable and practical tool for predicting the efficacy of targeted therapy plus immunotherapy in patients with unresectable HCC, and may help hepatologists and oncologists in clinical decision-making. Graphical abstract