Wenjing Han
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Wenjing Han has a total of 43 co-authors. Here are the top co-authors by number of shared publications:
- Lingjing Hu
Yanjing Medical College, Capital Medical University, Beijing, China
6 shared publications Guirong Tan
Advanced Neuroimaging Laboratory, Yuebei People's Hospital Affiliated to Shantou University Medical College, Shaoguan, Guangdong Province, China; Department of Radiology, Yuebei People's Hospital Affiliated to Shantou University Medical College, Shaoguan, Guangdong Province, China
4 shared publicationsHaihui Jiang
Department of Neurosurgery. Peking University Third Hospital, Peking University, Beijing, China
3 shared publicationsMing Guo
Department of Neurosurgery, Yuebei People's Hospital Affiliated to Shantou University Medical College, Shaoguan, Guangdong Province, China
3 shared publicationsWeiyin Vivian Liu
GE Healthcare, MR Research China, Beijing, China
3 shared publications- Xiang Liu
Advanced Neuroimaging Laboratory, Yuebei People’s Hospital Affiliated to Shantou University Medical College, Shaoguan, China
3 shared publications
Latest Publications
- 17 August 2026
- Published by Elsevier in Computerized Medical Imaging and Graphics
- Vol. 135, 102809
Global–Local Feature Fusion and SIREN-enhanced geometric supervision for coronary artery segmentation in CCTA
- 17 August 2026
- Published by Elsevier in Computerized Medical Imaging and Graphics
- Vol. 135, 102809
Global–Local Feature Fusion and SIREN-enhanced geometric supervision for coronary artery segmentation in CCTA
- Open Access
- 6 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 348-355
V-MIML: A Visualization-Oriented Educational Framework for Undergraduate Medical Imaging Technology Training in Medical Image Processing and Machine Learning
- 6 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 348-355
Open AccessV-MIML: A Visualization-Oriented Educational Framework for Undergraduate Medical Imaging Technology Training in Medical Image Processing and Machine Learning
- Open Access
- 12 December 2025
- Published by Elsevier in European Journal of Surgical Oncology
- Vol. 52 (2) , 111359
Machine learning based radiomics outperforms tumor/cerebellar peduncle T2-weighted imaging intensity ratios and ROIs-based apparent diffusion coefficient measurements in the preoperative prediction of meningioma consistency
- 12 December 2025
- Published by Elsevier in European Journal of Surgical Oncology
- Vol. 52 (2) , 111359
Open AccessMachine learning based radiomics outperforms tumor/cerebellar peduncle T2-weighted imaging intensity ratios and ROIs-based apparent diffusion coefficient measurements in the preoperative prediction of meningioma consistency
- 11 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 247-251
Stable Neuroimaging Feature Selection via Multi-Knockoff Simultaneous Inference
This study addresses the critical challenge of stable and accurate prediction of cognitive and disease states in neuroimaging analysis, with a focus on early detection of neurodegenerative diseases such as Alzheimer's Disease (AD). Existing knockoff-based feature selection methods, though promising, exhibit instability and limited power to identify relevant features, restricting their practical application in neuroimaging contexts. To overcome these limitations, we propose Stable Simultaneous FDP Bounds with Multi-Knockoffs (StableSim-MK), a novel framework that integrates multi-knockoff selection with simultaneous selective inference. StableSim-MK enhances stability, significantly increases the number of reliably identified biomarkers, and provides rigorous control over the False Discovery Proportion (FDP). Experimental results demonstrate that StableSim-MK identifies a broader set of AD-associated brain regions compared to existing methods, achieving both higher stability in selection and superior predictive accuracy. Notably, the proposed method offers robust and interpretable feature selection while revealing novel insights into brain regions that are overlooked by conventional approaches, thereby establishing meaningful links to cognitive states.
- 11 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 247-251
Stable Neuroimaging Feature Selection via Multi-Knockoff Simultaneous Inference
This study addresses the critical challenge of stable and accurate prediction of cognitive and disease states in neuroimaging analysis, with a focus on early detection of neurodegenerative diseases such as Alzheimer's Disease (AD). Existing knockoff-based feature selection methods, though promising, exhibit instability and limited power to identify relevant features, restricting their practical application in neuroimaging contexts. To overcome these limitations, we propose Stable Simultaneous FDP Bounds with Multi-Knockoffs (StableSim-MK), a novel framework that integrates multi-knockoff selection with simultaneous selective inference. StableSim-MK enhances stability, significantly increases the number of reliably identified biomarkers, and provides rigorous control over the False Discovery Proportion (FDP). Experimental results demonstrate that StableSim-MK identifies a broader set of AD-associated brain regions compared to existing methods, achieving both higher stability in selection and superior predictive accuracy. Notably, the proposed method offers robust and interpretable feature selection while revealing novel insights into brain regions that are overlooked by conventional approaches, thereby establishing meaningful links to cognitive states.
- Open Access
- 9 February 2025
- Published by Wiley in Clinical & Experimental Ophthalmology
- Vol. 53 (5) , 502-515
Detecting Glaucoma in Highly Myopic Eyes From Fundus Photographs Using Deep Convolutional Neural Networks
Background: High myopia (HM) is a major risk factor for glaucoma. However, glaucomatous optic neuropathy is often undiagnosed owing to atypical structural alterations with axial elongation. Moreover, an algorithm to detect glaucoma in highly myopic eyes has not yet been reported.Methods: We recruited 2643 colour fundus photographs to train a ResNet‐50 network for discriminating eyes with highly myopic glaucoma (HMG) from HM or glaucoma alone. We employed a 10‐fold cross‐validation strategy to evaluate the model's performance and applicability across diverse patient groups. Multiple metrics were computed to gauge the model's diagnostic process. The diagnostic ability of the model was then juxtaposed with those made by ophthalmologists to determine concordance. The gradient‐weighted class activation maps were used for visual explanations.Results: Our model demonstrated an overall accuracy of 97.7% with an area under the curve of 98.6% (sensitivity, 91.2%; specificity, 98.0%) for the differential diagnosis among HM, glaucoma, HMG and normal controls. These metrics notably outperformed the diagnostic performances of two attending ophthalmologists, who achieved accuracies of 64.7% and 69.9%. The activation maps derived from the model suggested that the most discriminative lesions for diagnosing HMG were predominantly in the disc, peripapillary area and inferior region of the disc, which are often displayed with a tessellated fundus. These results were slightly different from the understanding of the attending ophthalmologists.Conclusions: Our proposed model demonstrates high efficacy and suggests specific features for distinguishing eyes with HMG, enabling potential clinical value in assisting the intricate diagnosis of this vision‐threatening disease..
- 9 February 2025
- Published by Wiley in Clinical & Experimental Ophthalmology
- Vol. 53 (5) , 502-515
Open AccessDetecting Glaucoma in Highly Myopic Eyes From Fundus Photographs Using Deep Convolutional Neural Networks
Background: High myopia (HM) is a major risk factor for glaucoma. However, glaucomatous optic neuropathy is often undiagnosed owing to atypical structural alterations with axial elongation. Moreover, an algorithm to detect glaucoma in highly myopic eyes has not yet been reported.Methods: We recruited 2643 colour fundus photographs to train a ResNet‐50 network for discriminating eyes with highly myopic glaucoma (HMG) from HM or glaucoma alone. We employed a 10‐fold cross‐validation strategy to evaluate the model's performance and applicability across diverse patient groups. Multiple metrics were computed to gauge the model's diagnostic process. The diagnostic ability of the model was then juxtaposed with those made by ophthalmologists to determine concordance. The gradient‐weighted class activation maps were used for visual explanations.Results: Our model demonstrated an overall accuracy of 97.7% with an area under the curve of 98.6% (sensitivity, 91.2%; specificity, 98.0%) for the differential diagnosis among HM, glaucoma, HMG and normal controls. These metrics notably outperformed the diagnostic performances of two attending ophthalmologists, who achieved accuracies of 64.7% and 69.9%. The activation maps derived from the model suggested that the most discriminative lesions for diagnosing HMG were predominantly in the disc, peripapillary area and inferior region of the disc, which are often displayed with a tessellated fundus. These results were slightly different from the understanding of the attending ophthalmologists.Conclusions: Our proposed model demonstrates high efficacy and suggests specific features for distinguishing eyes with HMG, enabling potential clinical value in assisting the intricate diagnosis of this vision‐threatening disease..
- Open Access
- 17 December 2024
- Published by Springer Nature in Journal of Neuro-Oncology
- Vol. 172 (1) , 239-247
MR perfusion characteristics of pseudoprogression in brain tumors treated with immunotherapy – a comparative study with chemo-radiation induced pseudoprogression and radiation necrosis
Purpose Pseudoprogression is an atypical imaging pattern of response to immunotherapy in patients with brain tumors. MR perfusion studies in this field are limited. The purpose of our study is to compare the perfusion features between pseudoprogression lesions in malignant gliomas and brain metastases treated with immunotherapy (iPsP) and the pseudoprogression after chemo-radiation therapy and radiation necrosis after radiation treatment (ChR-PsP & RN). Methods We retrospectively reviewed 25 iPsP lesions in 16 brain tumor patients and 48 ChR-PsP & RN lesions in 35 patients. The cerebral blood volume (CBV) of MR dynamic susceptibility contrast (DSC) perfusion weighted imaging (PWI) was analyzed, and the mean and maximal values of the ratio of CBV (rCBVmean and rCBVmax) of iPsPs and ChR-PsP & RNs were calculated and compared between these two groups using the Mann-Whitney U test. A receiver operating characteristic curve analysis was conducted, and the optimal cutoff of perfusion parameters were determined using the area under the curve, sensitivity, and specificity. Results The medians of rCBVmean and rCBVmax in iPsP group were significantly higher (0.94 and 1.39 respectively) than ChR-PsP & RN group (0.67, p < 0.01 and 1.1, p = 0.01 respectively). The rCBVmean value of 0.69 can differentiate the iPsP from ChR-PsP & RN with the area under the curve of 0.71, sensitivity of 0.72, and specificity of 0.56. Conclusion These findings may suggest immunotherapy-induced higher perfusion in the iPsP lesions compared to ChR-PsP & RN lesions in primary and metastatic brain tumors..
- 17 December 2024
- Published by Springer Nature in Journal of Neuro-Oncology
- Vol. 172 (1) , 239-247
Open AccessMR perfusion characteristics of pseudoprogression in brain tumors treated with immunotherapy – a comparative study with chemo-radiation induced pseudoprogression and radiation necrosis
Purpose Pseudoprogression is an atypical imaging pattern of response to immunotherapy in patients with brain tumors. MR perfusion studies in this field are limited. The purpose of our study is to compare the perfusion features between pseudoprogression lesions in malignant gliomas and brain metastases treated with immunotherapy (iPsP) and the pseudoprogression after chemo-radiation therapy and radiation necrosis after radiation treatment (ChR-PsP & RN). Methods We retrospectively reviewed 25 iPsP lesions in 16 brain tumor patients and 48 ChR-PsP & RN lesions in 35 patients. The cerebral blood volume (CBV) of MR dynamic susceptibility contrast (DSC) perfusion weighted imaging (PWI) was analyzed, and the mean and maximal values of the ratio of CBV (rCBVmean and rCBVmax) of iPsPs and ChR-PsP & RNs were calculated and compared between these two groups using the Mann-Whitney U test. A receiver operating characteristic curve analysis was conducted, and the optimal cutoff of perfusion parameters were determined using the area under the curve, sensitivity, and specificity. Results The medians of rCBVmean and rCBVmax in iPsP group were significantly higher (0.94 and 1.39 respectively) than ChR-PsP & RN group (0.67, p < 0.01 and 1.1, p = 0.01 respectively). The rCBVmean value of 0.69 can differentiate the iPsP from ChR-PsP & RN with the area under the curve of 0.71, sensitivity of 0.72, and specificity of 0.56. Conclusion These findings may suggest immunotherapy-induced higher perfusion in the iPsP lesions compared to ChR-PsP & RN lesions in primary and metastatic brain tumors..
- Open Access
- 1 November 2024
- Published by Springer Nature in Cancer Imaging
- Vol. 24 (1) , 1-13
Multi-parameter MRI radiomics model in predicting postoperative progressive cerebral edema and hemorrhage after resection of meningioma
Background: Postoperative progressive cerebral edema and hemorrhage (PPCEH) are major complications after meningioma resection, yet their preoperative predictive studies are limited. The aim is to develop and validate a multiparametric MRI machine learning model to predict PPCEH after meningioma resection. Methods: This retrospective study included 148 patients with meningioma. A stratified three-fold cross-validation was used to split the dataset into training and validation sets. Radiomics features from the tumor enhancement (TE) and peritumoral brain edema (PTBE) regions were extracted from T1WI, T2WI, and ADC maps. Support vector machine constructed different radiomics models, and logistic regression explored clinical risk factors. Prediction models, integrating clinical and radiomics features, were evaluated using the area under the curve (AUC), visualized in a nomogram. Results: The radiomics model based on TE and PTBE regions (training set mean AUC: 0.85 (95% CI: 0.78–0.93), validation set mean AUC: 0.77 (95%CI: 0.63–0.90)) outperformed the model with TE region solely (training set mean AUC: 0.83 (95% CI: 0.76–0.91), validation set mean AUC: 0.73 (95% CI: 0.58–0.87)). Furthermore, the combined model incorporating radiomics features, and clinical features of preoperative peritumoral edema and tumor boundary adhesion, had the best predictive performance, with AUC values of 0.87 (95% CI: 0.80–0.94) and 0.84 (95% CI: 0.72–0.95) for the training and validation set. Conclusions: We developed a novel model based on clinical characteristics and multiparametric radiomics features derived from TE and PTBE regions, which can accurately and non-invasively predict PPCEH after meningioma resection. Additionally, our findings suggest the crucial role of PTBE radiomics features in understanding the potential mechanisms of PPCEH.
- 1 November 2024
- Published by Springer Nature in Cancer Imaging
- Vol. 24 (1) , 1-13
Open AccessMulti-parameter MRI radiomics model in predicting postoperative progressive cerebral edema and hemorrhage after resection of meningioma
Background: Postoperative progressive cerebral edema and hemorrhage (PPCEH) are major complications after meningioma resection, yet their preoperative predictive studies are limited. The aim is to develop and validate a multiparametric MRI machine learning model to predict PPCEH after meningioma resection. Methods: This retrospective study included 148 patients with meningioma. A stratified three-fold cross-validation was used to split the dataset into training and validation sets. Radiomics features from the tumor enhancement (TE) and peritumoral brain edema (PTBE) regions were extracted from T1WI, T2WI, and ADC maps. Support vector machine constructed different radiomics models, and logistic regression explored clinical risk factors. Prediction models, integrating clinical and radiomics features, were evaluated using the area under the curve (AUC), visualized in a nomogram. Results: The radiomics model based on TE and PTBE regions (training set mean AUC: 0.85 (95% CI: 0.78–0.93), validation set mean AUC: 0.77 (95%CI: 0.63–0.90)) outperformed the model with TE region solely (training set mean AUC: 0.83 (95% CI: 0.76–0.91), validation set mean AUC: 0.73 (95% CI: 0.58–0.87)). Furthermore, the combined model incorporating radiomics features, and clinical features of preoperative peritumoral edema and tumor boundary adhesion, had the best predictive performance, with AUC values of 0.87 (95% CI: 0.80–0.94) and 0.84 (95% CI: 0.72–0.95) for the training and validation set. Conclusions: We developed a novel model based on clinical characteristics and multiparametric radiomics features derived from TE and PTBE regions, which can accurately and non-invasively predict PPCEH after meningioma resection. Additionally, our findings suggest the crucial role of PTBE radiomics features in understanding the potential mechanisms of PPCEH.
- Open Access
- 17 August 2024
- Published by Springer Nature in Research Square
Pre-operative Non-invasive Prediction of Meningioma Consistency -- A Comparative Study Between MR Radiomics Model and ROI Based ADC Measurement
Background Meningioma is the most common intracranial tumor in the world, and total resection is the standard treatment approach. The status of meningiomas consistency is one of the most critical factors affecting the difficulty of surgery, which results in preoperative non-invasive prediction of tumor consistency is crucial for the design of surgical plan. This study aims to compare the performance of novel radiomics method and traditional apparent diffusion coefficient (ADC) values measured using regions of interest (ROI) in predicting the consistency of meningiomas. Methods and Materials: 148 patients with pathology-diagnosed meningiomas were enrolled in this retrospective study. Meningioma consistency was categorized into three grades: soft, moderate, and hard. ADC method: 3–5 ROIs (20-30mm2 each) within enhancing mass were placed on ADC to calculate the minimum, maximum, mean ADC values, and their ratios. Radiomics method: radiomics features were extracted from tumor enhancement region based on contrast-enhanced T1WI images, ADC maps, and T2WI images. Mann-Whitney U test and the least absolute shrinkage and selection operator (LASSO) algorithm were performed for potential radiomics feature selection. ADC measurement models and radiomics models were constructed separately using logistic regression. The receiver operating characteristic (ROC) and the area under curve (AUC) analysis were used to evaluate the predictive performance. Results 23 meningiomas were classified as soft, 98 were classified as moderate, and 27 as hard. For “soft” prediction, the highest AUC values for the training and validation sets are 0.54 and 0.54 respectively base on ADCmax. In contrast, AUC values were 0.82 in the training set and 0.78 in the validation set from radiomics models. For “hard” prediction, the best AUC values based on ADC measurement models were 0.58 and 0.58 respectively for the training and validation sets. The radiomics models achieved higher AUC values of 0.84 in the training set and 0.86 in the validation set. Conclusion Radiomics models outperform manual ADC measurement models in predicting meningioma consistency. Clinical Relevance Statement: Radiomics method offer better preoperative prediction of the meningioma consistency compared to manual ADC measurement methods. Therefore, the integration of artificial intelligence-based radiomics model to predict tumor consistency should be incorporated into the clinical preoperative management protocol for meningiomas.
- 17 August 2024
- Published by Springer Nature in Research Square
Open AccessPre-operative Non-invasive Prediction of Meningioma Consistency -- A Comparative Study Between MR Radiomics Model and ROI Based ADC Measurement
Background Meningioma is the most common intracranial tumor in the world, and total resection is the standard treatment approach. The status of meningiomas consistency is one of the most critical factors affecting the difficulty of surgery, which results in preoperative non-invasive prediction of tumor consistency is crucial for the design of surgical plan. This study aims to compare the performance of novel radiomics method and traditional apparent diffusion coefficient (ADC) values measured using regions of interest (ROI) in predicting the consistency of meningiomas. Methods and Materials: 148 patients with pathology-diagnosed meningiomas were enrolled in this retrospective study. Meningioma consistency was categorized into three grades: soft, moderate, and hard. ADC method: 3–5 ROIs (20-30mm2 each) within enhancing mass were placed on ADC to calculate the minimum, maximum, mean ADC values, and their ratios. Radiomics method: radiomics features were extracted from tumor enhancement region based on contrast-enhanced T1WI images, ADC maps, and T2WI images. Mann-Whitney U test and the least absolute shrinkage and selection operator (LASSO) algorithm were performed for potential radiomics feature selection. ADC measurement models and radiomics models were constructed separately using logistic regression. The receiver operating characteristic (ROC) and the area under curve (AUC) analysis were used to evaluate the predictive performance. Results 23 meningiomas were classified as soft, 98 were classified as moderate, and 27 as hard. For “soft” prediction, the highest AUC values for the training and validation sets are 0.54 and 0.54 respectively base on ADCmax. In contrast, AUC values were 0.82 in the training set and 0.78 in the validation set from radiomics models. For “hard” prediction, the best AUC values based on ADC measurement models were 0.58 and 0.58 respectively for the training and validation sets. The radiomics models achieved higher AUC values of 0.84 in the training set and 0.86 in the validation set. Conclusion Radiomics models outperform manual ADC measurement models in predicting meningioma consistency. Clinical Relevance Statement: Radiomics method offer better preoperative prediction of the meningioma consistency compared to manual ADC measurement methods. Therefore, the integration of artificial intelligence-based radiomics model to predict tumor consistency should be incorporated into the clinical preoperative management protocol for meningiomas.