Chunhao Wang

Overview:

  • Deep learning methods for image-based radiotherapy outcome prediction and assessment
  • Machine learning in outcome modelling
  • Automation in radiotherapy planning and delivery



Positions:

Assistant Professor of Radiation Oncology

Radiation Oncology
School of Medicine

Member of the Duke Cancer Institute

Duke Cancer Institute
School of Medicine

Education:

Ph.D. 2016

Duke University

Medical Physics Resident, Radiation Oncology Physics Division

Duke University

Medical Physics Resident, Radiation Oncology Physics Division

Duke University

Grants:

Publications:

A Dosimetric Study Comparing Two Beam Arrangement Strategies in Fractionated Thoracic Spine Stereotactic Body Radiotherapy (SBRT) Planning.

<h4>Purpose/objective(s)</h4>To compare dosimetric results of two beam arrangement strategies and their robustness to setup uncertainties in fractionated thoracic spine Stereotactic Body Radiotherapy (SBRT) MATERIALS/METHODS: Fifteen patients who received fractionated thoracic spine SBRT were retrospectively studied. All patients received simulation CT scans in body vacuum bag immobilization and multiparametric MRI exams. Clinical target volumes (CTVs) included single vertebral bodies with possible paraspinal space inclusion. Planning target volumes (PTVs) were expanded from CTVs with a 2mm margin but were cropped from spinal cord defined by MRI with a 2mm margin. Two different beam arrangement strategies of volumetric modulated arc therapy planning were studied: 1) 5 full arcs (FA) (360° each arc) with different collimator angles; and 2) 6 partial arcs (PA) (90° each arc) divided into two groups (3 arcs in each group) covering patient left-posterior-oblique (LPO) and right-posterior-oblique (RPO) regions, respectively, with orthogonal collimator angles. Both plans (Plan<sub>FA</sub> and Plan<sub>PA</sub>) were calculated as 24 Gy in 3 fractions using 6xFFF photon energy and a high-definition MLC model. During the inverse optimization of each plan for same patient, a same set of dose-volume constraints and optimization settings was used to exhaust parameter space search. Key dosimetric results of PTV as well as dose-volume parameters of relevant organs-at-risk (OARs) including spinal cord, esophagus, heart, lung and liver were evaluated. Dosimetric impact of on-board patient setup uncertainties to both plans were also simulated. All comparison results were analyzed by Wilcoxon signed-rank tests when the statistical power was sufficient.<h4>Results</h4>Both Plan<sub>FA</sub> and Plan<sub>PA</sub> achieved satisfactory spatial dose distribution. After PTV coverage normalization, Plan<sub>PA</sub> had better PTV dose uniformity (P = 0.026) and Plan<sub>FA</sub> had better dose fall-off gradient outside PTV. Plan<sub>PA</sub> had slightly better (18.1 ± 0.8 Gy vs 18.3 ± 0.9 Gy) cord max dose (D<sub>0.035cc</sub>) results (P = 0.213) and better cord low dose sparing V12 Gy (P = 0.013) results. Plan<sub>PA</sub> also achieved lower max dose (D<sub>0.035cc</sub>) of esophagus (P = 0.003) and heart, and improved low dose sparing (V5 Gy) of lung (P = 0.002) and liver. In plan parameter comparisons, Plan<sub>FA</sub> demonstrated stronger beam modulation effect (P < 0.001) but Plan<sub>PA</sub> had smaller MLC apertures (P < 0.001). In the simulated on-board scenarios with setup uncertainties, while both Plan<sub>FA</sub> and Plan<sub>PA</sub> had similar cord max dose increases with simulated pitch and/or roll in setup (< 0.1 Gy differences), Plan<sub>PA</sub> had minimal cord max dose increase (P < 0.001) with simulated anterior body weight loss.<h4>Conclusion</h4>For thoracic spine SBRT plans, the beam arrangement in Plan<sub>PA</sub> might be favored dosimetrically with better OAR sparing results and could be less sensitive to certain patient uncertainties, while the Plan<sub>FA</sub> could be acceptable with satisfactory dosimetry results.<h4>Author disclosure</h4>C. Wang: None. Y. Xie: None. Z. Hu: None. F. Yin: Research Grant; Varian Medical Systems. Teaching and mentoring graduate students. Administration of graduate program activities; Duke Kunshan University. Board of Directors Member at Large Members; AAPM. organize activities of the SANTRO; SANTRO.Z. Reitman: None. Y. Cui: None.
Authors
MLA Citation
Wang, C., et al. “A Dosimetric Study Comparing Two Beam Arrangement Strategies in Fractionated Thoracic Spine Stereotactic Body Radiotherapy (SBRT) Planning.International Journal of Radiation Oncology, Biology, Physics, vol. 111, no. 3S, 2021, p. e557. Epmc, doi:10.1016/j.ijrobp.2021.07.1510.
URI
https://scholars.duke.edu/individual/pub1503191
PMID
34701742
Source
epmc
Published In
International Journal of Radiation Oncology, Biology, Physics
Volume
111
Published Date
Start Page
e557
DOI
10.1016/j.ijrobp.2021.07.1510

Quantification of lung function on CT images based on pulmonary radiomic filtering.

PURPOSE: To develop a radiomics filtering technique for characterizing spatial-encoded regional pulmonary ventilation information on lung computed tomography (CT). METHODS: The lung volume was segmented on 46 CT images, and a 3D sliding window kernel was implemented across the lung volume to capture the spatial-encoded image information. Fifty-three radiomic features were extracted within the kernel, resulting in a fourth-order tensor object. As such, each voxel coordinate of the original lung was represented as a 53-dimensional feature vector, such that radiomic features could be viewed as feature maps within the lungs. To test the technique as a potential pulmonary ventilation biomarker, the radiomic feature maps were compared to paired functional images (Galligas PET or DTPA-SPECT) based on the Spearman correlation (ρ) analysis. RESULTS: The radiomic feature maps GLRLM-based Run-Length Non-Uniformity and GLCOM-based Sum Average are found to be highly correlated with the functional imaging. The achieved ρ (median [range]) for the two features are 0.46 [0.05, 0.67] and 0.45 [0.21, 0.65] across 46 patients and 2 functional imaging modalities, respectively. CONCLUSIONS: The results provide evidence that local regions of sparsely encoded heterogeneous lung parenchyma on CT are associated with diminished radiotracer uptake and measured lung ventilation defects on PET/SPECT imaging. These findings demonstrate the potential of radiomics to serve as a complementary tool to the current lung quantification techniques and provide hypothesis-generating data for future studies.
Authors
Yang, Z; Lafata, KJ; Chen, X; Bowsher, J; Chang, Y; Wang, C; Yin, F-F
MLA Citation
Yang, Zhenyu, et al. “Quantification of lung function on CT images based on pulmonary radiomic filtering.Med Phys, June 2022. Pubmed, doi:10.1002/mp.15837.
URI
https://scholars.duke.edu/individual/pub1525478
PMID
35770964
Source
pubmed
Published In
Med Phys
Published Date
DOI
10.1002/mp.15837

Post-Radiotherapy PET Image Outcome Prediction by Deep Learning Under Biological Model Guidance: A Feasibility Study of Oropharyngeal Cancer Application.

Purpose: To develop a method of biologically guided deep learning for post-radiation 18FDG-PET image outcome prediction based on pre-radiation images and radiotherapy dose information. Methods: Based on the classic reaction-diffusion mechanism, a novel biological model was proposed using a partial differential equation that incorporates spatial radiation dose distribution as a patient-specific treatment information variable. A 7-layer encoder-decoder-based convolutional neural network (CNN) was designed and trained to learn the proposed biological model. As such, the model could generate post-radiation 18FDG-PET image outcome predictions with breakdown biological components for enhanced explainability. The proposed method was developed using 64 oropharyngeal patients with paired 18FDG-PET studies before and after 20-Gy delivery (2 Gy/day fraction) by intensity-modulated radiotherapy (IMRT). In a two-branch deep learning execution, the proposed CNN learns specific terms in the biological model from paired 18FDG-PET images and spatial dose distribution in one branch, and the biological model generates post-20-Gy 18FDG-PET image prediction in the other branch. As in 2D execution, 718/233/230 axial slices from 38/13/13 patients were used for training/validation/independent test. The prediction image results in test cases were compared with the ground-truth results quantitatively. Results: The proposed method successfully generated post-20-Gy 18FDG-PET image outcome prediction with breakdown illustrations of biological model components. Standardized uptake value (SUV) mean values in 18FDG high-uptake regions of predicted images (2.45 ± 0.25) were similar to ground-truth results (2.51 ± 0.33). In 2D-based Gamma analysis, the median/mean Gamma Index (<1) passing rate of test images was 96.5%/92.8% using the 5%/5 mm criterion; such result was improved to 99.9%/99.6% when 10%/10 mm was adopted. Conclusion: The developed biologically guided deep learning method achieved post-20-Gy 18FDG-PET image outcome predictions in good agreement with ground-truth results. With the breakdown biological modeling components, the outcome image predictions could be used in adaptive radiotherapy decision-making to optimize personalized plans for the best outcome in the future.
Authors
Ji, H; Lafata, K; Mowery, Y; Brizel, D; Bertozzi, AL; Yin, F-F; Wang, C
MLA Citation
Ji, Hangjie, et al. “Post-Radiotherapy PET Image Outcome Prediction by Deep Learning Under Biological Model Guidance: A Feasibility Study of Oropharyngeal Cancer Application.Front Oncol, vol. 12, 2022, p. 895544. Pubmed, doi:10.3389/fonc.2022.895544.
URI
https://scholars.duke.edu/individual/pub1523882
PMID
35646643
Source
pubmed
Published In
Frontiers in Oncology
Volume
12
Published Date
Start Page
895544
DOI
10.3389/fonc.2022.895544

Insights of an AI agent via analysis of prediction errors: a case study of fluence map prediction for radiation therapy planning.

Purpose.We have previously reported an artificial intelligence (AI) agent that automatically generates intensity-modulated radiation therapy (IMRT) plans via fluence map prediction, by-passing inverse planning. This AI agent achieved clinically comparable quality for prostate cases, but its performance on head-and-neck patients leaves room for improvement. This study aims to collect insights of the deep-learning-based (DL-based) fluence map prediction model by systematically analyzing its prediction errors.Methods.From the modeling perspective, the DL model's output is the fluence maps of IMRT plans. However, from the clinical planning perspective, the plan quality evaluation should be based on the clinical dosimetric criteria such as dose-volume histograms. To account for the complex and non-intuitive relationships between fluence map prediction errors and the corresponding dose distribution changes, we propose a novel error analysis approach that systematically examines plan dosimetric changes that are induced by varying amounts of fluence prediction errors. We investigated four decomposition modes of model prediction errors. The two spatial domain decompositions are based on fluence intensity and fluence gradient. The two frequency domain decompositions are based on Fourier-space banded frequency rings and Fourier-space truncated low-frequency disks. The decomposed error was analyzed for its impact on the resulting plans' dosimetric metrics. The analysis was conducted on 15 test cases spared from the 200 training and 16 validation cases used to train the model.Results.Most planning target volume metrics were significantly correlated with most error decompositions. The Fourier space disk radii had the largest Spearman's coefficients. The low-frequency region within a disk of ∼20% Fourier space contained most of errors that impact overall plan quality.Conclusions.This study demonstrates the feasibility of using fluence map prediction error analysis to understand the AI agent's performance. Such insights will help fine-tune the DL models in architecture design and loss function selection.
Authors
Li, X; Wu, QJ; Wu, Q; Wang, C; Sheng, Y; Wang, W; Stephens, H; Yin, F-F; Ge, Y
MLA Citation
Li, Xinyi, et al. “Insights of an AI agent via analysis of prediction errors: a case study of fluence map prediction for radiation therapy planning.Phys Med Biol, vol. 66, no. 23, Nov. 2021. Pubmed, doi:10.1088/1361-6560/ac3841.
URI
https://scholars.duke.edu/individual/pub1501370
PMID
34757945
Source
pubmed
Published In
Phys Med Biol
Volume
66
Published Date
DOI
10.1088/1361-6560/ac3841

Collect Insights of an H&N IMRT Planning AI Agent Through Analyzing Relationships Between Fluence Map Prediction Error and the Corresponding Dosimetric Impacts.

<h4>Purpose/objective(s)</h4>With many deep learning (DL) models being developed for clinical applications, it is important to understand their behavior and clinical consequence. This study aims to collect insights of the relationship between fluence map prediction error and its dosimetric impacts in a DL-based AI agent for H&N IMRT planning.<h4>Materials/methods</h4>An AI agent has been implemented to generate IMRT plans via fluence map prediction, bypassing inverse optimization. While the prostate IMRT plans generated by the agent were comparable to clinical plans in quality, the application into H&N patients exhibited large variations in the plan quality due to higher anatomy complexity. As the DL model's output is fluence maps of an IMRT plan, standard error analyses were focused on the differences between the predicted and ground truth fluence maps, i.e., prediction error. However, the ultimate plan evaluation is based on clinical criteria such as DVHs and dose distributions. Therefore, the AI agent's performance in clinics is subjected to complex and non-intuitive relationships between fluence map prediction error and corresponding dose distribution changes, and warrants thorough investigation. In this study, a series of tests were designed to collect insights of the impact of DL model performance on plan's dosimetric quality. The fluence map prediction error was analyzed for its dosimetric effects using five error decomposition modes:1) ground truth fluence intensity bands in 5 threshold levels, 2) predicted fluence intensity bands in 5 threshold levels, 3) ground truth fluence gradient bands (high and low), 4) Fourier space bands (frequency bands) in 8 threshold levels, and 5) Fourier space circles (below certain frequency) in 8 threshold levels. The DL model was trained with 216 cases and tested with 15 additional cases. PTV and OAR dosimetric metrics were analyzed by Spearman's rank tests (P = 0.05).<h4>Results</h4>Most PTV-related metrics were significantly correlated with the error components. Among the different decomposition modes, the Fourier space circle radii have large Spearman's coefficients with PTV metrics, suggesting that they were best able to extract error components that reveal plan quality impacts. The low-frequency error within a Fourier space circle of radius = 32 pixels (20% of Fourier space) had the most significant impact on overall plan quality and PTV heterogeneity.<h4>Conclusion</h4>The fluence map prediction error analysis is critical to evaluate the AI agent performance. Such insight will help with fine-tuning the DL models in architecture design and loss function selection.
Authors
Li, X; Wu, QJJ; Wu, Q; Wang, C; Sheng, Y; Wang, W; Stephens, H; Yin, FF; Ge, Y
MLA Citation
Li, X., et al. “Collect Insights of an H&N IMRT Planning AI Agent Through Analyzing Relationships Between Fluence Map Prediction Error and the Corresponding Dosimetric Impacts.International Journal of Radiation Oncology, Biology, Physics, vol. 111, no. 3S, 2021, p. e94. Epmc, doi:10.1016/j.ijrobp.2021.07.479.
URI
https://scholars.duke.edu/individual/pub1502629
PMID
34702003
Source
epmc
Published In
International Journal of Radiation Oncology, Biology, Physics
Volume
111
Published Date
Start Page
e94
DOI
10.1016/j.ijrobp.2021.07.479