WLWenhua Li

Wenhua Li

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

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


Wenhua Li has a total of 42 co-authors. Here are the top co-authors by number of shared publications:

  • Caiting Chu

    Department of Radiology, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200092, China

    5 shared publications
  • Chengjin Gao

    Department of Emergency, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China

    3 shared publications
  • Caiting Chu

    Department of Radiology, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China

    2 shared publications
  • Dengbin Wang

    Department of Radiology, Xinhua Hospital affiliated to Shanghai Jiao Tong University School of Medicine, 1665 Kong Jiang Road, Shanghai 200092, China

    2 shared publications
  • Hua Fan

    Department of Radiology, Xinhua Hospital affiliated to Shanghai Jiao Tong University School of Medicine, 1665 Kong Jiang Road, Shanghai 200092, China

    2 shared publications
  • Meijiao Jiang

    Department of Radiology Xinhua Hospital Shanghai Jiao Tong University School of Medicine Shanghai China

    2 shared publications

Latest Publications

  • Ovarian masses suggested for MRI examination: assessment of deep learning models based on non-contrast-enhanced MRI sequences for predicting malignancy

    by Meijiao Jiang, Chui Kong, Siwei Lu, Qingwan Li, Caiting Chu, Wenhua Li

    Purpose We aims to assessed and compare four deep learning(DL) models using non-contrast-enhanced magnetic resonance imaging(MRI) to differentiate benign from malignant ovarian tumors, considering diagnostic efficacy and associated development costs. Methods 526 patients (327 benign lesions vs 199 malignant lesions) who were recommended for MRI due to suspected ovarian masses, confirmed with histopathology, were included in this retrospective study. A training cohort (n=367) and a validation cohort (n=159) were constructed. Based on the images of T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), we evaluated the diagnostic performance of four DL models (ConvNeXt, FBNet, GhostNet, ResNet50) in distinguishing between benign and malignant ovarian tumors. Two radiologists with varying levels of experience independently reviewed all original non-contrast-enhanced MR images from the validation cohort to determine if each case was benign or malignant. The area under the receiver operating characteristic curve (AUC), confusion matrices, accuracy, sensitivity, specificity, positive predictive value(PPV) and negative predictive value(NPV) were used to compare performance. Results The study of 526 ovarian mass patients (ages 1-92) evaluated four DL models for predicting malignant tumors, with AUCs ranging from 0.8091 to 0.8572 and accuracy between 81.1% and 85.5%. An experienced radiologist achieved 86.2% accuracy, slightly surpassing the DL models, while a less experienced radiologist had 69.2% accuracy. Resnet50 had the highest sensitivity (78.3%) and NPV (87.3%), while ConvNeXt excelled in specificity and PPV (100%). GhostNet and FBNet are more parameter-efficient than other models. Conclusion The four DL models effectively distinguished between benign and malignant ovarian tumors using non-contrast MRI. These models outperformed less experienced radiologists and were slightly less accurate than experienced ones. ResNet50 had the best predictive performance, while GhostNet was highly accurate with fewer parameters. Our study indicates that DL models based on non-contrast-enhanced MRI have the potential to assist in diagnosis. Graphical Abstract


  • Open Access

    Diagnostic value of magnetic resonance imaging for malignant ovarian tumors mis-subclassified by the ultrasound-based ADNEX model

    by Meijiao Jiang, Congcong Yuan, Siwei Lu, Yunkai Zhu, Caiting Chu, Wenhua Li

    Objective: Accurately predicting metastatic cancer to the adnexa, stage I and advanced ovarian cancer before surgery is crucial. The ADNEX model, based on ultrasound, is currently the only prediction model that can differentiate between these types. This study aims to analyze MRI features and diagnostic value in malignant ovarian tumors mis-subclassified by the ADNEX model, considering their diverse histopathologic types.Methods: From January 2018 to September 2022, 164 patients with pathologically confirmed ovarian malignancies were selected from those who were examined by ultrasound. The clinical and MRI characteristics of 51 patients mis-subclassified by the ADNEX model were compared with histopathological types.Results: A total of 30 were confirmed with primary ovarian cancer (5 with HGSOC, 14 with CCC, 2 with EC, 4 with MC, 2 with GCT, 1 with YST, 1 with immature teratoma, and 1 with dysgerminoma). There were 21 patients who had metastatic ovarian tumors (10 with colorectal cancer, 4 with gastric cancer, 2 with uterine cervical cancer, 3 with endometrial cancer, 1 with breast cancer, and 1 with LAMN). The only significant difference between the two groups was in CEA. The mean diameters of the primary and metastatic ovarian tumors were 10.29 cm (range: 3.61 cm–26.02 cm) and 8.58 cm (range: 3.10 cm–20.30 cm), respectively. A total of 42 masses were lobulated (82.35%, 42/51), and 26 masses were solid-cystic (26/51, 50.98%). There was a significant difference between CCC and other tumors, with mean ADC values of 1.01 × 10−3 mm2/s (range: 0.68–1.28×10−3 mm2/s) and 0.74×10−3 mm2/s (range: 0.48–0.99×10−3 mm2/s), respectively (P=0.000). A total of 50 masses presented isointense-T1, hyperintense-T2, and hyperintense-DWI signal on MRI (50/51,98.04%). There were 33 masses that showed intensive enhancement (33/51,64.71%). There were 17 masses who had necrosis (17/51, 33.33%), with the majority being HGSOC and ovarian metastases from colorectal and gastric cancers (12/17, 70.59%). There were 19 masses that presented hemorrhage (19/51,37.25%), with the majority being CCC (10/19, 52.63%). A total of 46 masses were diagnosed correctly by MRI (46/51,90.20%). There were 35 and 15 masses that were rated as O-RADS score 5 and score 4, respectively. One mass was rated as score 3.Conclusions: DWI signal, ADC value, degree of enhancement, and characteristic components within the mass on MRI can provide supplementary information for malignant ovarian tumors mis-subclassified by the ADNEX model.


  • Open Access

    Chest computed tomography-based artificial intelligence-aided latent class analysis for diagnosis of severe pneumonia

    by Caiting Chu, Yiran Guo, Zhenghai Lu, Ting Gui, Shuhui Zhao, Xuee Cui, Siwei Lu, Meijiao Jiang, Wenhua Li, Chengjin Gao

    Background: There is little literature describing the artificial intelligence (AI)-aided diagnosis of severe pneumonia (SP) subphenotypes and the association of the subphenotypes with the ventilatory treatment efficacy. The aim of our study is to illustrate whether clinical and biological heterogeneity, such as ventilation and gas-exchange, exists among patients with SP using chest computed tomography (CT)-based AI-aided latent class analysis (LCA). Methods: This retrospective study included 413 patients hospitalized at Xinhua Hospital diagnosed with SP from June 1, 2015 to May 30, 2020. AI quantification results of chest CT and their combination with additional clinical variables were used to develop LCA models in an SP population. The optimal subphenotypes were determined though evaluating statistical indicators of all the LCA models, and clinical implications of them such as guiding ventilation strategies were further explored by statistical methods. Results: The two-class LCA model based on AI quantification results of chest CT can describe the biological characteristics of the SP population well and hence yielded the two clinical subphenotypes. Patients with subphenotype-1 had milder infections (P P P P = 0.001) and 2-year (P P <0.001). Compared with patients with subphenotype-2, those with subphenotype-1 showed a relatively better match between CT-based AI metrics of the non-infected region and oxygenation, and their clinical outcomes were effectively improved after receiving invasive ventilation treatment. Conclusions: A two-class LCA model based on AI quantification results of chest CT in the SP population particularly revealed clinical heterogeneity of lung function. Identifying the degree of match between ventilation and gas-exchange may help guide decisions about assisted ventilation.


  • Open Access

    MRI-Based Radiomics Analysis for Intraoperative Risk Assessment in Gravid Patients at High Risk with Placenta Accreta Spectrum

    by Caiting Chu, Ming Liu, Yuzhen Zhang, Shuhui Zhao, Yaqiong Ge, Wenhua Li, Chengjin Gao

    Background: Gravid patients at high risk with placenta accreta spectrum (PAS) face life-threatening risk at delivery. Intraoperative risk assessment for patients is currently insufficient. We aimed to develop an assessment system of intraoperative risks through MRI-based radiomics. Methods: A total of 131 patients enrolled were randomly grouped according to a ratio of 7:3. Clinical data were analyzed retrospectively. Radiomic features were extracted from sagittal Fast Imaging Employing State-sate Acquisition images. Univariate and multivariate regression analyses were performed to build models using R software. A receiver operating characteristic curve and decision curve analysis (DCA) were performed to determine the predictive performance of models. Results: Six radiomic features and two clinical variables were used to construct the combined model for selection of removal protocols of the placenta, with an area under the curve (AUC) of 0.90 and 0.91 in the training and test cohorts, respectively. Nine radiomic features and two clinical variables were obtained to establish the combined model for prediction of intraoperative blood loss, with an AUC of 0.90 and 0.88 in the both cohorts, respectively. The DCA confirmed the clinical utility of the combined model. Conclusion: The analysis of combined MRI-based radiomics with clinics could be clinically beneficial for patients.


  • Quantifying magnetic resonance imaging features to classify placenta accreta spectrum (PAS) in high-risk gravid patients

    by Caiting Chu, Ming Liu, Yuzheng Zhang, Lingwei Yu, Dengbin Wang, Chengjin Gao, Wenhua Li


  • A combined radiomics and clinical variables model for prediction of malignancy in T2 hyperintense uterine mesenchymal tumors on MRI

    by Tingting Wang, Jing Gong, Qiao Li, Caiting Chu, Wenbin Shen, Weijun Peng, Yajia Gu, Wenhua Li

    Objective This study aims to develop a machine learning model for prediction of malignancy in T2 hyperintense mesenchymal uterine tumors based on T2-weighted image (T2WI) features and clinical information. Methods This retrospective study included 134 patients with T2 hyperintense uterine mesenchymal tumors (104 patients in training cohort and 30 in testing cohort). A total of 960 radiomics features were initially computed and extracted from each 3D segmented tumor depicting on T2WI. The support vector machine (SVM) classifier was applied to build computer-aided diagnosis (CAD) models by using selected clinical and radiomics features, respectively. Finally, an observer study was conducted by comparing with two radiologists to evaluate the diagnostic performance. The area under the receiver operating characteristic (ROC) curve (AUC) was computed to assess the performance of each model. Results Comparing with the T2WI-based radiomics model (AUC: 0.76 ± 0.09) and the clinical model (AUC: 0.79 ± 0.09), the combined model significantly improved the AUC value to 0.91 ± 0.05 (p < 0.05). The clinical-radiomics combined model yielded equivalent or higher performance than two radiologists (AUC: 0.78 vs. 0.91, p = 0.03; 0.90 vs.0.91, p = 0.13). There was a significant difference between the AUC values of two radiologists (p < 0.05). Conclusions It is feasible to predict malignancy risk of T2 hyperintense uterine mesenchymal tumors by combining clinical variables and T2WI-based radiomics features. Machine learning–based classification model may be useful to assist radiologists in decision-making. Key Points • Radiomics approach has the potential to distinguish between benign and malignant mesenchymal uterine tumors. • T2WI-based radiomics analysis combined with clinical variables performed well in predicting malignancy risk of T2 hyperintense uterine mesenchymal tumors. • Machine learning–based classification model may be useful to assist radiologists in characterization of a T2 hyperintense uterine mesenchymal tumor.


  • Open Access

    Combining Clinical Characteristics and Specific Magnetic Resonance Imaging Features to Predict Placenta Accreta

    by Caiting Chu, Shuhui Zhao, Ming Ding, Ming Liu, Yuzheng Zhang, Lei Bao, Dengbin Wang, Wenhua Li

    Objective: The aim of this study was to explore the independent clinical and magnetic resonance imaging (MRI) performance risk factors for predicting placenta accreta. Methods: From January 2012 to December 2015, we retrospectively reviewed the clinical characteristics and MRI features of 97 patients. Of these, 42 were confirmed to be placenta accreta by pathological results or cesarean delivery findings. We tried to identify the independent risk factors by multivariate logistic regression model for significant differences in variables determined by univariate analysis. Results: The multivariate logistic regression model indicated that 2 or more instances of previous cesarean deliveries and/or abortions, placenta previa, and placenta-myometrial interface interruption were independent risk factors for placenta accreta. The odd ratios were 3.79 for patients who had 2 or more instances of previous cesarean deliveries and/or abortions, 0.04 for marginal/partial placenta previa, 0.024 for complete placenta previa, and 6.56 for placenta-myometrial interface interruption. The values of accuracy and positive prediction by combination of a single clinical risk factor and placenta-myometrial interface interruption and of positive prediction by a combination of all 3 risk factors for predicting placenta accreta were raised to 83.5%, 75%, and 92.9%, respectively. We obtained 3 different risk groups by different combinations of all 3 risk factors. Conclusions: The study suggested that 2 or more instances of previous cesarean deliveries and/or abortion, placenta previa, and placenta-myometrial interface interruption were independent risk factors for placenta accreta. A combination of a single clinical risk factor and an MRI risk factor can improve the diagnosis of placenta accreta, and a combination of all 3 risk factors could help recognize patients with placenta accreta.


  • Open Access

    Characterization of tubo-ovarian abscess mimicking adnexal masses: Comparison between contrast-enhanced CT, 18F–FDG PET/CT and MRI

    by Hua Fan, Ting-Ting Wang, Gang Ren, Hong-Liang Fu, Xiang-Ru Wu, Cai-Ting Chu, Wen-Hua Li


  • Open Access

    The Correlation of Pirani and Dimeglio Scoring Systems for Ponseti Management at Different Levels of Deformity Severity

    by Hua Fan, Yubin Liu, Li Zhao, Caiting Chu, Yongyu An, Tingting Wang, Wenhua Li

    The Pirani and Dimeglio scoring systems both have excellent inter-observer and intra-observer reliability, but no research has been conducted to determine their inter-observer reliability and their relationship at different levels of deformity. A total of 173 idiopathic clubfoot cases were reviewed using Pirani and Dimeglio scoring systems, and the number of casts needed was also recorded. For clubfeet with a cast number equal to 2 or 7 and 8, the inter-observer reliability of the two scoring systems was poor or moderate, and there was no correlation between the two scoring systems. There was also no correlation between the Dimeglio scoring score with the number of casts for grade II or IV clubfeet. A binary regression of the number of casts on initial Pirani or Dimeglio scores showed that there was a Quadratic or Cubic relation between the scores and the cast numbers. In conclusion, in the case of mild and very severe clubfoot deformity, the interobserver reliability and its ability to predict the number of casts needed for clubfoot deformity correction was poor. A more objective evaluation system may be required.