A Generative Adversarial Network (GAN)-Based Technique for Synthesizing On-Board Imaging Artifacts in the Extended Cardiac-Torso (XCAT) Phantoms
Y Chang1*, Z Zhang1, Z Jiang1, Y Lai2, L Ren1, (1) Duke University, Durham, NC, (2) University of Texas Southwestern Medical Center, Dallas, TX
Presentations
WE-IePD-TRACK 1-7 (Wednesday, 7/28/2021) 3:00 PM - 3:30 PM [Eastern Time (GMT-4)]
Purpose: Simulating on-board imaging in digital phantoms is valuable for evaluating and optimizing of on-board treatment techniques like adaptive therapy. This study aims to synthesize CBCT imaging artifacts for different acquisition conditions in XCAT phantoms by developing a GAN-based deep-learning technique.
Methods: An artifact simulation model is developed using CT images and image acquisition conditions as inputs, and is trained to generate CBCT images with the conditions-specific imaging artifacts. Then, XCAT phantoms are input to the model with user-desired image acquisition conditions to simulate CBCT artifacts in XCAT. The model was trained in two steps: (1)simulate cone and under-sampling artifacts for different projection numbers (450, 150, or 100); (2)simulate scatter and beam-hardening artifacts. Cone-beam projections of the CT volumes were generated by ray-tracing and used to reconstruct CBCTs using FDK-backprojection, which were used for the output for step 1 and input for step 2. The second-step model outputs corresponding CBCTs reconstructed using Monte-Carlo(MC) projections. We incorporated 15 patients and 5 XCAT phantoms for model training, validation, and testing. We evaluated the model performance with peak-signal-to-noise-ratio(PSNR), structural-similarity(SSIM), cross-correlation and NPS(noise power spectrum) correlation coefficient(NCC).
Results: For the under-sampling and cone artifacts simulation, the PSNR reached 38.58/37.24/36.92, and the SSIM reached 0.987/0.978/0.966 for the testing patient with 450/225/100 projections, respectively. The cross-correlation and NCC achieved at least 0.99 for all projection numbers. Qualitatively, the model successfully simulated the streak artifact due to under-sampling and cone artifact in the XCAT phantom. For scatter simulation, the PSNR reached 28.05, and SSIM reached 0.904 for the testing patient. The cross-correlation was 0.927, and NCC was 0.984.
Conclusion: The results demonstrated the feasibility of synthesizing CBCT artifacts in the 4D-XCAT phantom using the proposed method. This crucial development enhances the value of XCAT phantoms for evaluating and optimizing various on-board imaging and treatment delivery techniques.
Funding Support, Disclosures, and Conflict of Interest: This work is supported by the National Institutes of Health under Grant No. R01-CA184173 and R01-EB028324.
ePosters
Keywords
Phantoms, Image Artifacts, Cone-beam CT
Taxonomy
IM/TH- Cone Beam CT: Phantoms – digital
Contact Email