Jingwei Duan
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
Jingwei Duan has a total of 40 co-authors. Here are the top co-authors by number of shared publications:
- Quan Chen
Department of Radiation Oncology, Mayo Clinic Arizona, Phoenix, AZ 85054, USA
4 shared publications - Yi Rong
Department of Radiation Oncology, Mayo Clinic Arizona, Phoenix, AZ, USA
3 shared publications Carlos E. Cardenas
Department of Radiation Oncology University of Alabama at Birmingham Birmingham USA
2 shared publications- Joel Anthony Pogue
Department of Radiation Oncology University of Alabama at Birmingham Birmingham Alabama USA
2 shared publications Libing Zhu
Department of Radiation Oncology, Mayo Clinic Arizona, Phoenix, Arizona
2 shared publications- Libing Zhu
Department of Radiation Oncology, Mayo Clinic, Phoenix, AZ 85058, USA
2 shared publications
Latest Publications
- 18 January 2026
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 125 (4) , 1258-1270
Deformable Point Cloud Registration-Based Bidirectional Local Distance (DPCR-BLD): A Methodology for Systematic Evaluation and Visualization of Local Disagreements in Clinical Autocontouring
- 18 January 2026
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 125 (4) , 1258-1270
Deformable Point Cloud Registration-Based Bidirectional Local Distance (DPCR-BLD): A Methodology for Systematic Evaluation and Visualization of Local Disagreements in Clinical Autocontouring
- Open Access
- 12 June 2025
- Published by Wiley in Journal of Applied Clinical Medical Physics
- Vol. 26 (7) , e70137
Simulation‐free workflow for lattice radiation therapy using deep learning predicted synthetic computed tomography: A feasibility study
Purpose: Lattice radiation therapy (LRT) is a form of spatially fractionated radiation therapy that allows increased total dose delivery aiming for improved treatment response without an increase in toxicities, commonly utilized for palliation of bulky tumors. The LRT treatment planning process is complex, while eligible patients often have an urgent need for expedited treatment start. In this study, we aimed to develop a simulation‐free workflow for volumetric modulated arc therapy (VMAT)‐based LRT planning via deep learning‐predicted synthetic CT (sCT) to expedite treatment initiation. Methods: Two deep learning models were initially trained using 3D U‐Net architecture to generate sCT from diagnostic CTs (dCT) of the thoracic and abdomen regions using a training dataset of 50 patients. The models were then tested on an independent dataset of 15 patients using image similarity analysis assessing mean absolute error (MAE) and structural similarity index measure (SSIM) as metrics. VMAT‐based LRT plans were generated based on sCT and recalculated on the planning CT (pCT) for dosimetric accuracy comparison. Differences in dose volume histogram (DVH) metrics between pCT and sCT plans were assessed using the Wilcoxon signed‐rank test. Results: The final sCT prediction model demonstrated high image similarity to pCT, with a MAE and SSIM of 38.93 ± 14.79 Hounsfield Units (HU) and 0.92 ± 0.05 for the thoracic region, and 73.60 ± 22.90 HU and 0.90 ± 0.03 for the abdominal region, respectively. There were no statistically significant differences between sCT and pCT plans in terms of organ‐at‐risk and target volume DVH parameters, including maximum dose (Dmax), mean dose (Dmean), dose delivered to 90% (D90%) and 50% (D50%) of target volume, except for minimum dose (Dmin) and (D10%). Conclusion: With demonstrated high image similarity and adequate dose agreement between sCT and pCT, our study is a proof‐of‐concept for using deep learning predicted sCT for a simulation‐free treatment planning workflow for VMAT‐based LRT.
- 12 June 2025
- Published by Wiley in Journal of Applied Clinical Medical Physics
- Vol. 26 (7) , e70137
Open AccessSimulation‐free workflow for lattice radiation therapy using deep learning predicted synthetic computed tomography: A feasibility study
Purpose: Lattice radiation therapy (LRT) is a form of spatially fractionated radiation therapy that allows increased total dose delivery aiming for improved treatment response without an increase in toxicities, commonly utilized for palliation of bulky tumors. The LRT treatment planning process is complex, while eligible patients often have an urgent need for expedited treatment start. In this study, we aimed to develop a simulation‐free workflow for volumetric modulated arc therapy (VMAT)‐based LRT planning via deep learning‐predicted synthetic CT (sCT) to expedite treatment initiation. Methods: Two deep learning models were initially trained using 3D U‐Net architecture to generate sCT from diagnostic CTs (dCT) of the thoracic and abdomen regions using a training dataset of 50 patients. The models were then tested on an independent dataset of 15 patients using image similarity analysis assessing mean absolute error (MAE) and structural similarity index measure (SSIM) as metrics. VMAT‐based LRT plans were generated based on sCT and recalculated on the planning CT (pCT) for dosimetric accuracy comparison. Differences in dose volume histogram (DVH) metrics between pCT and sCT plans were assessed using the Wilcoxon signed‐rank test. Results: The final sCT prediction model demonstrated high image similarity to pCT, with a MAE and SSIM of 38.93 ± 14.79 Hounsfield Units (HU) and 0.92 ± 0.05 for the thoracic region, and 73.60 ± 22.90 HU and 0.90 ± 0.03 for the abdominal region, respectively. There were no statistically significant differences between sCT and pCT plans in terms of organ‐at‐risk and target volume DVH parameters, including maximum dose (Dmax), mean dose (Dmean), dose delivered to 90% (D90%) and 50% (D50%) of target volume, except for minimum dose (Dmin) and (D10%). Conclusion: With demonstrated high image similarity and adequate dose agreement between sCT and pCT, our study is a proof‐of‐concept for using deep learning predicted sCT for a simulation‐free treatment planning workflow for VMAT‐based LRT.
- Open Access
Improving the Precision of Deep-Learning-Based Head and Neck Target Auto-Segmentation by Leveraging Radiology Reports Using a Large Language Model
Background/Objectives: The accurate delineation of primary tumors (GTVp) and metastatic lymph nodes (GTVn) in head and neck (HN) cancers is essential for effective radiation treatment planning, yet remains a challenging and laborious task. This study aims to develop a deep-learning-based auto-segmentation (DLAS) model trained on external datasets with false-positive elimination using clinical diagnosis reports. Methods: The DLAS model was trained on a multi-institutional public dataset with 882 cases. Forty-four institutional cases were randomly selected as the external testing dataset. DLAS-generated GTVp and GTVn were validated against clinical diagnosis reports to identify false-positive and false-negative segmentation errors using two large language models: ChatGPT-4 and Llama-3. False-positive ruling out was conducted by matching the centroids of AI-generated contours with the slice locations or anatomical regions described in the reports. Performance was evaluated using the Dice similarity coefficient (DSC), 95th percentile Hausdorff distance (HD95), and tumor detection precision. Results: ChatGPT-4 outperformed Llama-3 in accurately extracting tumor locations from the diagnostic reports. False-positive contours were identified in 15 out of 44 cases. The DSCmean of the DLAS contours for GTVp and GTVn increased from 0.68 to 0.75 and from 0.69 to 0.75, respectively, after the ruling-out process. Notably, the average HD95 value for GTVn decreased from 18.81 mm to 5.2 mm. Post ruling out, the model achieved 100% precision for GTVp and GTVn when compared with the results of physician-determined contours. Conclusions: The false-positive ruling-out approach based on diagnostic reports effectively enhances the precision of DLAS in the HN region. The model accurately identifies the tumor location and detects all false-negative errors.
Open AccessImproving the Precision of Deep-Learning-Based Head and Neck Target Auto-Segmentation by Leveraging Radiology Reports Using a Large Language Model
Background/Objectives: The accurate delineation of primary tumors (GTVp) and metastatic lymph nodes (GTVn) in head and neck (HN) cancers is essential for effective radiation treatment planning, yet remains a challenging and laborious task. This study aims to develop a deep-learning-based auto-segmentation (DLAS) model trained on external datasets with false-positive elimination using clinical diagnosis reports. Methods: The DLAS model was trained on a multi-institutional public dataset with 882 cases. Forty-four institutional cases were randomly selected as the external testing dataset. DLAS-generated GTVp and GTVn were validated against clinical diagnosis reports to identify false-positive and false-negative segmentation errors using two large language models: ChatGPT-4 and Llama-3. False-positive ruling out was conducted by matching the centroids of AI-generated contours with the slice locations or anatomical regions described in the reports. Performance was evaluated using the Dice similarity coefficient (DSC), 95th percentile Hausdorff distance (HD95), and tumor detection precision. Results: ChatGPT-4 outperformed Llama-3 in accurately extracting tumor locations from the diagnostic reports. False-positive contours were identified in 15 out of 44 cases. The DSCmean of the DLAS contours for GTVp and GTVn increased from 0.68 to 0.75 and from 0.69 to 0.75, respectively, after the ruling-out process. Notably, the average HD95 value for GTVn decreased from 18.81 mm to 5.2 mm. Post ruling out, the model achieved 100% precision for GTVp and GTVn when compared with the results of physician-determined contours. Conclusions: The false-positive ruling-out approach based on diagnostic reports effectively enhances the precision of DLAS in the HN region. The model accurately identifies the tumor location and detects all false-negative errors.
- Open Access
Deep Learning-Based Synthetic CT for Personalized Treatment Modality Selection Between Proton and Photon Therapy in Thoracic Cancer
Objectives: Identifying patients’ advantageous radiotherapy modalities prior to CT simulation is challenging. This study aimed to develop a workflow using deep learning (DL)-predicted synthetic CT (sCT) for treatment modality comparison based solely on a diagnostic CT (dCT). Methods: A DL network, U-Net, was trained utilizing 46 thoracic cases from a public database to generate sCT images predicting planning CT (pCT) scans based on the latest dCT, and tested on 15 institutional patients. The sCT accuracy was evaluated against the corresponding pCT and a commercial algorithm deformed CT (MdCT) based on Mean Absolute Error (MAE) and Universal Quality Index (UQI). To determine advantageous treatment modality, clinical dose-volume histogram (DVH) metrics and Normal Tissue Complication Probability (NTCP) differences between proton and photon treatment plans were analyzed on the sCTs via concordance correlation coefficient (CCC). Results: The AI-generated sCTs closely resembled those of the commercial deformation algorithm in the tested cases. The differences in MAE and UQI values between the sCT-vs-pCT and MdCT-vs-pCT were 19.38 HU and 0.06, respectively. The mean absolute NTCP deviation between sCT and pCT was 1.54%, 0.21%, and 2.36% for esophagus perforation, lung pneumonitis, and heart pericarditis, respectively. The CCC between sCT and pCT was 0.90 for DVH metrics and 0.97 for NTCP, indicating moderate agreement for DVH metrics and substantial agreement. Conclusions: Radiation oncologists can potentially utilize this personalized sCT based approach as a clinical support tool to rapidly compare the treatment modality benefit during patient consultation and facilitate in-depth discussion on potential toxicities at a patient-specific level.
Open AccessDeep Learning-Based Synthetic CT for Personalized Treatment Modality Selection Between Proton and Photon Therapy in Thoracic Cancer
Objectives: Identifying patients’ advantageous radiotherapy modalities prior to CT simulation is challenging. This study aimed to develop a workflow using deep learning (DL)-predicted synthetic CT (sCT) for treatment modality comparison based solely on a diagnostic CT (dCT). Methods: A DL network, U-Net, was trained utilizing 46 thoracic cases from a public database to generate sCT images predicting planning CT (pCT) scans based on the latest dCT, and tested on 15 institutional patients. The sCT accuracy was evaluated against the corresponding pCT and a commercial algorithm deformed CT (MdCT) based on Mean Absolute Error (MAE) and Universal Quality Index (UQI). To determine advantageous treatment modality, clinical dose-volume histogram (DVH) metrics and Normal Tissue Complication Probability (NTCP) differences between proton and photon treatment plans were analyzed on the sCTs via concordance correlation coefficient (CCC). Results: The AI-generated sCTs closely resembled those of the commercial deformation algorithm in the tested cases. The differences in MAE and UQI values between the sCT-vs-pCT and MdCT-vs-pCT were 19.38 HU and 0.06, respectively. The mean absolute NTCP deviation between sCT and pCT was 1.54%, 0.21%, and 2.36% for esophagus perforation, lung pneumonitis, and heart pericarditis, respectively. The CCC between sCT and pCT was 0.90 for DVH metrics and 0.97 for NTCP, indicating moderate agreement for DVH metrics and substantial agreement. Conclusions: Radiation oncologists can potentially utilize this personalized sCT based approach as a clinical support tool to rapidly compare the treatment modality benefit during patient consultation and facilitate in-depth discussion on potential toxicities at a patient-specific level.
- Top Cited PapersOpen Access
- 3 March 2025
- Published by Wiley in Journal of Applied Clinical Medical Physics
- Vol. 26 (5) , e70038
Assessing HyperSight iterative CBCT for dose calculation in online adaptive radiotherapy for pelvis and breast patients compared to synthetic CT
Purpose/objectives: Recent technological advancements have increased efficiency for clinical deliverability of online‐adaptive‐radiotherapy (oART). Previous cone‐beam‐computed‐tomography (CBCT) generations lacked the ability to provide reliable Hounsfield‐units (HU), thus requiring oART workflows to rely on synthetic‐CT (sCT) images derived through deformable‐image‐registration (DIR) between the planning CT (pCT) and the daily CBCT. These sCTs are prone to errors stemming from DIR, potentially contributing to dosimetric errors. This study aims to evaluate the capability of direct dose calculation using a novel CBCT platform, HyperSight (Varian‐Medical‐Systems), as an alternative for sCT.Methods/materials: To validate the HyperSight iterative CBCT (HS‐iCBCT) HU accuracy, 125 kV and 140 kV HS‐iCBCT calibration curves were benchmarked against a pCT calibration curve. To determine the clinical impact of HS‐iCBCT compared to sCT, daily adaptive sessions from 47 oART fractions from 10 patients were analyzed. For these patients, HS‐iCBCT was acquired for daily adaption, and sCT was generated as part of the standard adaptive workflow. After daily adaption, dose was recalculated directly using the HS‐iCBCT, and the HS‐iCBCT and sCT dose distributions were compared by γ‐index and dose‐volume‐histogram (DVH) analysis.Results: The mean HU differences of pCT minus HS‐iCBCT (140/125 kV) were −40.97/−57.79, 9.86/21.74, and 87.22/158.20 for lung, water, and bone. In the patient cohort, the median gamma passing rates between HS‐iCBCT and sCT‐based dose calculations for 3%/2 mm and 1%/1 mm were 99.57 and 96.45% with 10% threshold, and 99.92% and 86.15% with 80% threshold. Dosimetric deviations in high dose regions were concentrated in areas with larger deformation, that is, surface change and variable bladder/bowel filling. The median (min‐max) D98%/V100% absolute deviations were 0.3(0.0–1.6)/0.0(0.0–13.7) and 0.4(0.0–1.4)/0.5(0.0–17.5) for CTVs and PTVs.Conclusions: The HS‐iCBCT platform with iterative reconstruction provides dose calculation comparable to sCT for pelvis and breast patients. However, acceptable, yet noticeable, dose discrepancies between HS‐iCBCT and sCT exist, particularly in high‐dose regions. Further investigations are needed to benchmark these methods against ground truth measurements.
- 3 March 2025
- Published by Wiley in Journal of Applied Clinical Medical Physics
- Vol. 26 (5) , e70038
Top Cited PapersOpen AccessAssessing HyperSight iterative CBCT for dose calculation in online adaptive radiotherapy for pelvis and breast patients compared to synthetic CT
Purpose/objectives: Recent technological advancements have increased efficiency for clinical deliverability of online‐adaptive‐radiotherapy (oART). Previous cone‐beam‐computed‐tomography (CBCT) generations lacked the ability to provide reliable Hounsfield‐units (HU), thus requiring oART workflows to rely on synthetic‐CT (sCT) images derived through deformable‐image‐registration (DIR) between the planning CT (pCT) and the daily CBCT. These sCTs are prone to errors stemming from DIR, potentially contributing to dosimetric errors. This study aims to evaluate the capability of direct dose calculation using a novel CBCT platform, HyperSight (Varian‐Medical‐Systems), as an alternative for sCT.Methods/materials: To validate the HyperSight iterative CBCT (HS‐iCBCT) HU accuracy, 125 kV and 140 kV HS‐iCBCT calibration curves were benchmarked against a pCT calibration curve. To determine the clinical impact of HS‐iCBCT compared to sCT, daily adaptive sessions from 47 oART fractions from 10 patients were analyzed. For these patients, HS‐iCBCT was acquired for daily adaption, and sCT was generated as part of the standard adaptive workflow. After daily adaption, dose was recalculated directly using the HS‐iCBCT, and the HS‐iCBCT and sCT dose distributions were compared by γ‐index and dose‐volume‐histogram (DVH) analysis.Results: The mean HU differences of pCT minus HS‐iCBCT (140/125 kV) were −40.97/−57.79, 9.86/21.74, and 87.22/158.20 for lung, water, and bone. In the patient cohort, the median gamma passing rates between HS‐iCBCT and sCT‐based dose calculations for 3%/2 mm and 1%/1 mm were 99.57 and 96.45% with 10% threshold, and 99.92% and 86.15% with 80% threshold. Dosimetric deviations in high dose regions were concentrated in areas with larger deformation, that is, surface change and variable bladder/bowel filling. The median (min‐max) D98%/V100% absolute deviations were 0.3(0.0–1.6)/0.0(0.0–13.7) and 0.4(0.0–1.4)/0.5(0.0–17.5) for CTVs and PTVs.Conclusions: The HS‐iCBCT platform with iterative reconstruction provides dose calculation comparable to sCT for pelvis and breast patients. However, acceptable, yet noticeable, dose discrepancies between HS‐iCBCT and sCT exist, particularly in high‐dose regions. Further investigations are needed to benchmark these methods against ground truth measurements.
- Open Access
- 17 February 2025
- Published by Elsevier in Advances in Radiation Oncology
- Vol. 10 (4) , 101740
Assessing Dosimetric Benefits of Cone Beam Computed Tomography-Guided Online Adaptive Radiation Treatment Frequencies for Lung Cancer
- 17 February 2025
- Published by Elsevier in Advances in Radiation Oncology
- Vol. 10 (4) , 101740
Open AccessAssessing Dosimetric Benefits of Cone Beam Computed Tomography-Guided Online Adaptive Radiation Treatment Frequencies for Lung Cancer
- Open Access
- 25 February 2023
- Published by Wiley in Medical Physics
- Vol. 50 (5) , 2715-2732
Contouring quality assurance methodology based on multiple geometric features against deep learning auto‐segmentation
Background: Contouring error is one of the top failure modes in radiation treatment. Multiple efforts have been made to develop tools to automatically detect segmentation errors. Deep learning‐based auto‐segmentation (DLAS) has been used as a baseline for flagging manual segmentation errors, but those efforts are limited to using only one or two contour comparison metrics.Purpose: The purpose of this research is to develop an improved contouring quality assurance system to identify and flag manual contouring errors.Methods and materials: DLAS contours were used as a reference to compare with manually segmented contours. A total of 27 geometric agreement metrics were determined from the comparisons between the two segmentation approaches. Feature selection was performed to optimize the training of a machine learning classification model to identify potential contouring errors. A public dataset with 339 cases was used to train and test the classifier. Four independent classifiers were trained using five‐fold cross validation, and the predictions from each classifier were ensembled using soft voting. The trained model was validated on a held‐out testing dataset. An additional independent clinical dataset with 60 cases was used to test the generalizability of the model. Model predictions were reviewed by an expert to confirm or reject the findings.Results: The proposed machine learning multiple features (ML‐MF) approach outperformed traditional nonmachine‐learning‐based approaches that are based on only one or two geometric agreement metrics. The machine learning model achieved recall (precision) values of 0.842 (0.899), 0.762 (0.762), 0.727 (0.842), and 0.773 (0.773) for Brainstem, Parotid_L, Parotid_R, and mandible contours, respectively compared to 0.526 (0.909), 0.619 (0.765), 0.682 (0.882), 0.773 (0.568) for an approach based solely on Dice similarity coefficient values. In the external validation dataset, 66.7, 93.3, 94.1, and 58.8% of flagged cases were confirmed to have contouring errors by an expert for Brainstem, Parotid_L, Parotid_R, and mandible contours, respectively.Conclusions: The proposed ML‐MF approach, which includes multiple geometric agreement metrics to flag manual contouring errors, demonstrated superior performance in comparison to traditional methods. This method is easy to implement in clinical practice and can help to reduce the significant time and labor costs associated with manual segmentation and review.
- 25 February 2023
- Published by Wiley in Medical Physics
- Vol. 50 (5) , 2715-2732
Open AccessContouring quality assurance methodology based on multiple geometric features against deep learning auto‐segmentation
Background: Contouring error is one of the top failure modes in radiation treatment. Multiple efforts have been made to develop tools to automatically detect segmentation errors. Deep learning‐based auto‐segmentation (DLAS) has been used as a baseline for flagging manual segmentation errors, but those efforts are limited to using only one or two contour comparison metrics.Purpose: The purpose of this research is to develop an improved contouring quality assurance system to identify and flag manual contouring errors.Methods and materials: DLAS contours were used as a reference to compare with manually segmented contours. A total of 27 geometric agreement metrics were determined from the comparisons between the two segmentation approaches. Feature selection was performed to optimize the training of a machine learning classification model to identify potential contouring errors. A public dataset with 339 cases was used to train and test the classifier. Four independent classifiers were trained using five‐fold cross validation, and the predictions from each classifier were ensembled using soft voting. The trained model was validated on a held‐out testing dataset. An additional independent clinical dataset with 60 cases was used to test the generalizability of the model. Model predictions were reviewed by an expert to confirm or reject the findings.Results: The proposed machine learning multiple features (ML‐MF) approach outperformed traditional nonmachine‐learning‐based approaches that are based on only one or two geometric agreement metrics. The machine learning model achieved recall (precision) values of 0.842 (0.899), 0.762 (0.762), 0.727 (0.842), and 0.773 (0.773) for Brainstem, Parotid_L, Parotid_R, and mandible contours, respectively compared to 0.526 (0.909), 0.619 (0.765), 0.682 (0.882), 0.773 (0.568) for an approach based solely on Dice similarity coefficient values. In the external validation dataset, 66.7, 93.3, 94.1, and 58.8% of flagged cases were confirmed to have contouring errors by an expert for Brainstem, Parotid_L, Parotid_R, and mandible contours, respectively.Conclusions: The proposed ML‐MF approach, which includes multiple geometric agreement metrics to flag manual contouring errors, demonstrated superior performance in comparison to traditional methods. This method is easy to implement in clinical practice and can help to reduce the significant time and labor costs associated with manual segmentation and review.