Ke Chen
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
Research Subjects
Ke Chen has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Ying Weng
University of Nottingham Ningbo China
9 shared publications - Juexin Zhang
University of Nottingham Ningbo China
8 shared publications Akram A. Hosseini
Nottingham University Hospitals NHS Trust, Queens Medical Center, Nottingham, United Kingdom
2 shared publicationsGuokun Zuo
The Ningbo Institute of Industrial Technology (CNITECH) of the Chinese Academy of Sciences (CAS), Ningbo, China
2 shared publications- Qifeng Zhong
University of Nottingham Ningbo China 315100 Ningbo China
2 shared publications Tom Dening
University of Nottingham, Nottingham, United Kingdom
2 shared publications
Latest Publications
- 2 February 2027
- Published by Springer Nature
- p. 455-461
Deep Learning for Glioblastoma Morpho-Pathological Features Identification: A BraTS-Pathology Challenge Solution
- 2 February 2027
- Published by Springer Nature
- p. 455-461
Deep Learning for Glioblastoma Morpho-Pathological Features Identification: A BraTS-Pathology Challenge Solution
- 2 February 2027
- Published by Springer Nature
- p. 569-575
U-Net Based Healthy 3D Brain Tissue Inpainting
- 1 April 2026
- Published by Springer Nature
- p. 102-109
Robust 3D Brain MRI Inpainting with Random Masking Augmentation
- 1 April 2026
- Published by Springer Nature
- p. 102-109
Robust 3D Brain MRI Inpainting with Random Masking Augmentation
- 1 April 2026
- Published by Springer Nature
- p. 193-202
Patch-Level Glioblastoma Subregion Classification with a Contrastive Learning-Based Encoder
- 1 April 2026
- Published by Springer Nature
- p. 193-202
Patch-Level Glioblastoma Subregion Classification with a Contrastive Learning-Based Encoder
- Open Access
- 24 December 2025
- Published by Wiley in Alzheimer's & Dementia
- Vol. 21 (S2) , e097585
Ensemble Deep Learning Models for Brain Connectivity in Alzheimer's Disease Diagnosis via Different Perspectives
Background: Accurate diagnosis of Alzheimer's disease (AD) remains a significant challenge in neuroimaging research. Combining anatomical and structural connectivity features offers a promising avenue for improving diagnostic performance. This study introduces two advanced ensemble methods, Brain Ensemble Network (BrainEnsNet) and Population Ensemble Network (PopEnsNet), designed to leverage multimodal brain connectivity for AD diagnosis. Method: BrainEnsNet integrates anatomical features with brain connectivity graphs, optimizing individual sub‐model performance. PopEnsNet enhances this framework by constructing population‐level relationships and employing node correlation‐guided aggregation to achieve superior node classification. Performance was evaluated across multiple scales using metrics such as accuracy, precision, recall, and F1‐score. Result: With features derived from two scales of Lausanne2018 atlas, BrainEnsNet achieved accuracies of 80.82% (Scale 1) and 81.50% (Scale 2) for AD/Mild Cognitive Impairment (MCI)/Normal Cognitive (NC), consistently outperforming its sub‐models. PopEnsNet further improved diagnostic outcomes with accuracies of 81.40% (Scale 1) and 82.03% (Scale 2), demonstrating its effectiveness in incorporating population‐level relationships. Conclusion: The results highlight the potential of multimodal ensemble networks in neuroimaging‐based AD diagnosis. By integrating population‐level relationships and robust feature representations, BrainEnsNet and PopEnsNet set a foundation for advancing diagnostic accuracy. Future research could incorporate longitudinal data and additional imaging modalities to enhance clinical utility and generalizability.
- 24 December 2025
- Published by Wiley in Alzheimer's & Dementia
- Vol. 21 (S2) , e097585
Open AccessEnsemble Deep Learning Models for Brain Connectivity in Alzheimer's Disease Diagnosis via Different Perspectives
Background: Accurate diagnosis of Alzheimer's disease (AD) remains a significant challenge in neuroimaging research. Combining anatomical and structural connectivity features offers a promising avenue for improving diagnostic performance. This study introduces two advanced ensemble methods, Brain Ensemble Network (BrainEnsNet) and Population Ensemble Network (PopEnsNet), designed to leverage multimodal brain connectivity for AD diagnosis. Method: BrainEnsNet integrates anatomical features with brain connectivity graphs, optimizing individual sub‐model performance. PopEnsNet enhances this framework by constructing population‐level relationships and employing node correlation‐guided aggregation to achieve superior node classification. Performance was evaluated across multiple scales using metrics such as accuracy, precision, recall, and F1‐score. Result: With features derived from two scales of Lausanne2018 atlas, BrainEnsNet achieved accuracies of 80.82% (Scale 1) and 81.50% (Scale 2) for AD/Mild Cognitive Impairment (MCI)/Normal Cognitive (NC), consistently outperforming its sub‐models. PopEnsNet further improved diagnostic outcomes with accuracies of 81.40% (Scale 1) and 82.03% (Scale 2), demonstrating its effectiveness in incorporating population‐level relationships. Conclusion: The results highlight the potential of multimodal ensemble networks in neuroimaging‐based AD diagnosis. By integrating population‐level relationships and robust feature representations, BrainEnsNet and PopEnsNet set a foundation for advancing diagnostic accuracy. Future research could incorporate longitudinal data and additional imaging modalities to enhance clinical utility and generalizability.
- Open Access
Robust 3D Brain MRI Inpainting with Random Masking Augmentation
The ASNR-MICCAI BraTS-Inpainting Challenge was established to mitigate dataset biases that limit deep learning models in the quantitative analysis of brain tumor MRI. This paper details our submission to the 2025 challenge, a novel deep learning framework for synthesizing healthy tissue in 3D scans. The core of our method is a U-Net architecture trained to inpaint synthetically corrupted regions, enhanced with a random masking augmentation strategy to improve generalization. Quantitative evaluation confirmed the efficacy of our approach, yielding an SSIM of 0.873$\pm$0.004, a PSNR of 24.996$\pm$4.694, and an MSE of 0.005$\pm$0.087 on the validation set. On the final online test set, our method achieved an SSIM of 0.919$\pm$0.088, a PSNR of 26.932$\pm$5.057, and an RMSE of 0.052$\pm$0.026. This performance secured first place in the BraTS-Inpainting 2025 challenge and surpassed the winning solutions from the 2023 and 2024 competitions on the official leaderboard.
Open AccessRobust 3D Brain MRI Inpainting with Random Masking Augmentation
The ASNR-MICCAI BraTS-Inpainting Challenge was established to mitigate dataset biases that limit deep learning models in the quantitative analysis of brain tumor MRI. This paper details our submission to the 2025 challenge, a novel deep learning framework for synthesizing healthy tissue in 3D scans. The core of our method is a U-Net architecture trained to inpaint synthetically corrupted regions, enhanced with a random masking augmentation strategy to improve generalization. Quantitative evaluation confirmed the efficacy of our approach, yielding an SSIM of 0.873$\pm$0.004, a PSNR of 24.996$\pm$4.694, and an MSE of 0.005$\pm$0.087 on the validation set. On the final online test set, our method achieved an SSIM of 0.919$\pm$0.088, a PSNR of 26.932$\pm$5.057, and an RMSE of 0.052$\pm$0.026. This performance secured first place in the BraTS-Inpainting 2025 challenge and surpassed the winning solutions from the 2023 and 2024 competitions on the official leaderboard.
- Open Access
Patch-Level Glioblastoma Subregion Classification with a Contrastive Learning-Based Encoder
The significant molecular and pathological heterogeneity of glioblastoma, an aggressive brain tumor, complicates diagnosis and patient stratification. While traditional histopathological assessment remains the standard, deep learning offers a promising path toward objective and automated analysis of whole slide images. For the BraTS-Path 2025 Challenge, we developed a method that fine-tunes a pre-trained Vision Transformer (ViT) encoder with a dedicated classification head on the official training dataset. Our model's performance on the online validation set, evaluated via the Synapse platform, yielded a Matthews Correlation Coefficient (MCC) of 0.7064 and an F1-score of 0.7676. On the final test set, the model achieved an MCC of 0.6509 and an F1-score of 0.5330, which secured our team second place in the BraTS-Pathology 2025 Challenge. Our results establish a solid baseline for ViT-based histopathological analysis, and future efforts will focus on bridging the performance gap observed on the unseen validation data.
Open AccessPatch-Level Glioblastoma Subregion Classification with a Contrastive Learning-Based Encoder
The significant molecular and pathological heterogeneity of glioblastoma, an aggressive brain tumor, complicates diagnosis and patient stratification. While traditional histopathological assessment remains the standard, deep learning offers a promising path toward objective and automated analysis of whole slide images. For the BraTS-Path 2025 Challenge, we developed a method that fine-tunes a pre-trained Vision Transformer (ViT) encoder with a dedicated classification head on the official training dataset. Our model's performance on the online validation set, evaluated via the Synapse platform, yielded a Matthews Correlation Coefficient (MCC) of 0.7064 and an F1-score of 0.7676. On the final test set, the model achieved an MCC of 0.6509 and an F1-score of 0.5330, which secured our team second place in the BraTS-Pathology 2025 Challenge. Our results establish a solid baseline for ViT-based histopathological analysis, and future efforts will focus on bridging the performance gap observed on the unseen validation data.
- Open Access
Deep Learning for Glioblastoma Morpho-pathological Features Identification: A BraTS-Pathology Challenge Solution
Glioblastoma, a highly aggressive brain tumor with diverse molecular and pathological features, poses a diagnostic challenge due to its heterogeneity. Accurate diagnosis and assessment of this heterogeneity are essential for choosing the right treatment and improving patient outcomes. Traditional methods rely on identifying specific features in tissue samples, but deep learning offers a promising approach for improved glioblastoma diagnosis. In this paper, we present our approach to the BraTS-Path Challenge 2024. We leverage a pre-trained model and fine-tune it on the BraTS-Path training dataset. Our model demonstrates poor performance on the challenging BraTS-Path validation set, as rigorously assessed by the Synapse online platform. The model achieves an accuracy of 0.392229, a recall of 0.392229, and a F1-score of 0.392229, indicating a consistent ability to correctly identify instances under the target condition. Notably, our model exhibits perfect specificity of 0.898704, showing an exceptional capacity to correctly classify negative cases. Moreover, a Matthews Correlation Coefficient (MCC) of 0.255267 is calculated, to signify a limited positive correlation between predicted and actual values and highlight our model's overall predictive power. Our solution also achieves the second place during the testing phase.
Open AccessDeep Learning for Glioblastoma Morpho-pathological Features Identification: A BraTS-Pathology Challenge Solution
Glioblastoma, a highly aggressive brain tumor with diverse molecular and pathological features, poses a diagnostic challenge due to its heterogeneity. Accurate diagnosis and assessment of this heterogeneity are essential for choosing the right treatment and improving patient outcomes. Traditional methods rely on identifying specific features in tissue samples, but deep learning offers a promising approach for improved glioblastoma diagnosis. In this paper, we present our approach to the BraTS-Path Challenge 2024. We leverage a pre-trained model and fine-tune it on the BraTS-Path training dataset. Our model demonstrates poor performance on the challenging BraTS-Path validation set, as rigorously assessed by the Synapse online platform. The model achieves an accuracy of 0.392229, a recall of 0.392229, and a F1-score of 0.392229, indicating a consistent ability to correctly identify instances under the target condition. Notably, our model exhibits perfect specificity of 0.898704, showing an exceptional capacity to correctly classify negative cases. Moreover, a Matthews Correlation Coefficient (MCC) of 0.255267 is calculated, to signify a limited positive correlation between predicted and actual values and highlight our model's overall predictive power. Our solution also achieves the second place during the testing phase.
- 28 December 2024
- Published by Springer Nature
- p. 233-240
Synthesis of Healthy Tissue Within Tumor Area via U-Net
- 28 December 2024
- Published by Springer Nature
- p. 233-240
Synthesis of Healthy Tissue Within Tumor Area via U-Net
- Open Access
- 1 December 2024
- Published by Wiley in Alzheimer's & Dementia
- Vol. 20 (S1) , e090237
Brain Connectome Graphs based Machine Learning for Diagnosis of Alzheimer’s Disease
Background: Alzheimer’s Disease (AD) is a neurodegenerative disorder characterized by progressive cognitive decline and memory loss. Early and accurate diagnosis of AD is crucial for patient information, advance planning, and potentially effective intervention and treatment. The integration of machine learning techniques with brain connectome graphs, encompassing both structural and functional brain connectomes, can enhance the accuracy and efficiency of AD diagnosis.Method: We propose a framework for AD diagnosis using both structural and functional brain connectome graphs with machine learning techniques. Our framework comprises three stages: image pre‐processing, brain connectome graph construction, and machine learning‐based AD diagnosis. We use PANDA and fMRIPrep for image pre‐processing and brain connectome graph construction. The two types of derived graphs, namely brain structural and functional connectome graphs, are then used as joint inputs for graph neural network (GNN)‐based models for AD prediction, including its early stage, and mild cognitive impairment (MCI).Result: The experiments are performed on diffusion magnetic resonance imaging (dMRI) and functional MRI (fMRI) obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) datasets, displaying promising results with high performance in identifying AD, (MCI), and cognitively normal (CN) patients. Explainable artificial intelligence algorithms are also applied to the model’s predictions for visualizing decision strategies.Conclusion: The findings of our research contribute to the burgeoning field of neuroinformatics by offering a novel and effective approach to AD diagnosis. The integration of machine learning with brain connectome graphs has potential to provide early and accurate identification of individuals at risk of AD, and pave the way for timely interventions and personalized treatment strategies. Moreover, this study sheds light on the intricate connectome‐level changes associated with AD and fosters a deeper understanding of the disease’s pathophysiology. Ultimately, this research has established a significant step towards leveraging advanced computational techniques to enhance our ability to diagnose and manage AD.
- 1 December 2024
- Published by Wiley in Alzheimer's & Dementia
- Vol. 20 (S1) , e090237
Open AccessBrain Connectome Graphs based Machine Learning for Diagnosis of Alzheimer’s Disease
Background: Alzheimer’s Disease (AD) is a neurodegenerative disorder characterized by progressive cognitive decline and memory loss. Early and accurate diagnosis of AD is crucial for patient information, advance planning, and potentially effective intervention and treatment. The integration of machine learning techniques with brain connectome graphs, encompassing both structural and functional brain connectomes, can enhance the accuracy and efficiency of AD diagnosis.Method: We propose a framework for AD diagnosis using both structural and functional brain connectome graphs with machine learning techniques. Our framework comprises three stages: image pre‐processing, brain connectome graph construction, and machine learning‐based AD diagnosis. We use PANDA and fMRIPrep for image pre‐processing and brain connectome graph construction. The two types of derived graphs, namely brain structural and functional connectome graphs, are then used as joint inputs for graph neural network (GNN)‐based models for AD prediction, including its early stage, and mild cognitive impairment (MCI).Result: The experiments are performed on diffusion magnetic resonance imaging (dMRI) and functional MRI (fMRI) obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) datasets, displaying promising results with high performance in identifying AD, (MCI), and cognitively normal (CN) patients. Explainable artificial intelligence algorithms are also applied to the model’s predictions for visualizing decision strategies.Conclusion: The findings of our research contribute to the burgeoning field of neuroinformatics by offering a novel and effective approach to AD diagnosis. The integration of machine learning with brain connectome graphs has potential to provide early and accurate identification of individuals at risk of AD, and pave the way for timely interventions and personalized treatment strategies. Moreover, this study sheds light on the intricate connectome‐level changes associated with AD and fosters a deeper understanding of the disease’s pathophysiology. Ultimately, this research has established a significant step towards leveraging advanced computational techniques to enhance our ability to diagnose and manage AD.