Unsupervised Deep Learning for Synthetic CT Generation from CBCT Images for Proton and Carbon Ion Therapy for Paediatric Patients
Abstract
:1. Introduction
2. Materials and Methods
2.1. Patients Dataset
2.2. CBCT at CNAO
2.3. Imaging Data
2.4. Pre-Processing
- Each CBCT slice is converted from Cartesian to polar coordinates using interpolation [48]. In Cartesian coordinates, the ring manifests itself as a series of concentric rings, with the centre coinciding with the reconstructed image centre. In polar coordinates, instead, it is converted into parallel horizontal stripes [49].
- The DFT of the CBCT in polar coordinates is computed: horizontal stripes appear as a vertical band in the middle of the whole spectrum, i.e., their high-frequency information is located at the centre position of the image [49].
- Low-pass filtering in the frequency domain is used to remove artefacts while the high-frequency details of the image are preserved [49].
- The ring’s frequencies are converted back from frequency to imaging domain-polar coordinates, appearing as a series of parallel stripes. An anisotropic Gaussian filter is also applied to smooth stripes in the horizontal direction only.
- The ring in polar coordinates is converted back to Cartesian coordinates, clamping its values in the [−10 10] range, in order not to significantly affect the original CBCT grey-level distribution.
- The ring in Cartesian coordinates is algebraically subtracted to the original image to obtain a new image with the ring artefact significantly reduced.
2.5. CycleGAN
2.6. Evaluation Metrics
3. Results
3.1. Patients’ Data
3.2. Training–Testing Split
3.3. Pre-Processing Outcomes
3.4. Qualitative Evaluation
3.5. Quantitative Evaluation
4. Discussion
- Training and testing sets’ enlargement by including adult patients and employing data augmentation techniques—it is expected that by increasing the sample size, the performance of the network will improve;
- Validation on other anatomical areas, as previous investigations have demonstrated the inter-anatomic generalizability of DCNNs [17];
- Validation on pelvic and extra-pelvic images acquired in both FF and HF modalities using the newly developed system in CNAO Room 1, as soon as first data are available;
- Validation on publicly available data repositories to assess the performance of the network to generalise on data from different CBCT systems [61];
- Exploration of the 3D variants of the networks, by either decreasing the FOV and reducing computational burden or by employing powerful cloud computing services that are compliant with data protection regulation [17];
- Inclusion of the other image similarity metrics by selecting specific regions of interest (ROIs) to bypass the problems related to possible anatomical mismatch [55];
- In vivo range verification and assessment of CT number accuracy of sCT to detect failures and outliers in the generated images [64];
- Investigation of CycleGAN performances with different simulated CBCT imaging dose levels to assess the suitability of this method in low-dose paediatric protocols [68].
5. Conclusions
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
List of Abbreviations
AI | Artificial Intelligence |
APT | Adaptive PT |
c | Contrast |
C | Carbon ion |
CBCT | Cone beam CT |
CNAO | Centro Nazionale di Adroterapia Oncologica (National Centre for Oncological Hadrontherapy) |
CT | Computed tomography |
DCBCT | CBCT discriminator |
DCT | CT discriminator |
DCNN | Deep convolutional neural network |
DFT | Discrete Fourier transform |
DIR | Deformable image registration |
DL | Deep learning |
DPR | Dose Pass Rate |
DVH | Dose–volume histogram |
E | Expected value |
F | Female |
fCBCT | Filtered CBCT |
FDK | Feldkamp–Davis–Kress (algorithm) |
FF | Full-fan (CBCT) |
FOV | Field of view |
Fx | Fraction |
GAN | Generative adversarial network |
GPR | Gamma pass rate |
GPU | Graphical processing unit |
Gy | Grey |
HF | Half-fan (CBCT) |
HU | Hounsfield unit |
iCBCT | Input CBCT |
l | Luminance |
L | Full objective |
LIDENTITY | Identity loss |
LGAN-F | Adversarial loss (F: CBCT generator) |
LGAN-G | Adversarial loss (G: CT generator) |
M | Male |
MAE | Mean absolute error |
MRI | Magnetic resonance imaging |
MSE | Mean square error |
OAR | Organ at risk |
P | Proton |
pCT | Planning CT |
PSNR | Peak signal-to-noise ratio |
PT | Particle therapy |
PT_ | Patient |
RBE | Relative Biological Effectiveness |
rCBCT | CBCT registered (on CT) |
RMSE | Root mean square error |
RR | Rigid registration |
s | Structure |
SSIM | Structural similarity |
sCT | Synthetic CT |
TPS | Treatment Planning System |
TR | Training |
TS | Testing |
vCT | Verification CT |
X | Discarded |
Yo | Years old |
References
- Stock, M.; Georg, D.; Ableitinger, A.; Zechner, A.; Utz, A.; Mumot, M.; Kragl, G.; Hopfgartner, J.; Gora, J.; Böhlen, T.; et al. The technological basis for adaptive ion beam therapy at MedAustron: Status and outlook. Z. Med. Phys. 2018, 28, 196–210. [Google Scholar] [CrossRef] [PubMed]
- Thummerer, A.; Seller Oria, C.; Zaffino, P.; Meijers, A.; Guterres Marmitt, G.; Wijsman, R.; Seco, J.; Langendijk, J.A.; Knopf, A.C.; Spadea, M.F.; et al. Clinical suitability of deep learning based synthetic CTs for adaptive proton therapy of lung cancer. Med. Phys. 2021, 48, 7673–7684. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Sheikh, K.; Liu, D.; Li, H.; Acharya, S.; Ladra, M.M.; Hrinivich, W.T. Dosimetric evaluation of cone-beam CT-based synthetic CTs in pediatric patients undergoing intensity-modulated proton therapy. J. Appl. Clin. Med. Phys. 2022, 23, e13604. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Taasti, V.T.; Hattu, D.; Peeters, S.; van der Salm, A.; van Loon, J.; de Ruysscher, D.; Nilsson, R.; Andersson, S.; Engwall, E.; Unipan, M.; et al. Clinical evaluation of synthetic computed tomography methods in adaptive proton therapy of lung cancer patients. Phys. Imaging Radiat. Oncol. 2023, 27, 100459. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Arai, K.; Kadoya, N.; Kato, T.; Endo, H.; Komori, S.; Abe, Y.; Nakamura, T.; Wada, H.; Kikuchi, Y.; Takai, Y.; et al. Feasibility of CBCT-based proton dose calculation using a histogram-matching algorithm in proton beam therapy. Phys. Med. 2017, 33, 68–76. [Google Scholar] [CrossRef] [PubMed]
- La Fauci, F.; Augugliaro, M.; Mazzola, G.C.; Comi, S.; Pepa, M.; Zaffaroni, M.; Vincini, M.G.; Corrao, G.; Mistretta, F.A.; Luzzago, S.; et al. Dosimetric Evaluation of the Inter-Fraction Motion of Organs at Risk in SBRT for Nodal Oligometastatic Prostate Cancer. Appl. Sci. 2022, 12, 10949. [Google Scholar] [CrossRef]
- Rosa, C.; Caravatta, L.; Di Tommaso, M.; Fasciolo, D.; Gasparini, L.; Di Guglielmo, F.C.; Augurio, A.; Vinciguerra, A.; Vecchi, C.; Genovesi, D. Cone-beam computed tomography for organ motion evaluation in locally advanced rectal cancer patients. Radiol. Med. 2021, 126, 147–154. [Google Scholar] [CrossRef] [PubMed]
- Szmul, A.; Taylor, S.; Lim, P.; Cantwell, J.; Moreira, I.; Zhang, Y.; D’Souza, D.; Moinuddin, S.; Gaze, M.N.; Gains, J.; et al. Deep learning based synthetic CT from cone beam CT generation for abdominal paediatric radiotherapy. Phys. Med. Biol. 2023, 68, 105006. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Taylor, S.; Lim, P.; Cantwell, J.; D’Souza, D.; Moinuddin, S.; Chang, Y.C.; Gaze, M.N.; Gains, J.; Veiga, C. Image guidance and interfractional anatomical variation in paediatric abdominal radiotherapy. Br. J. Radiol. 2023, 96, 20230058. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Lalonde, A.; Bobić, M.; Winey, B.; Verburg, J.; Sharp, G.C.; Paganetti, H. Anatomic changes in head and neck intensity-modulated proton therapy: Comparison between robust optimization and online adaptation. Radiother. Oncol. 2021, 159, 39–47. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Thummerer, A.; de Jong, B.A.; Zaffino, P.; Meijers, A.; Marmitt, G.G.; Seco, J.; Steenbakkers, R.J.H.M.; Langendijk, J.A.; Both, S.; Spadea, M.F.; et al. Comparison of the suitability of CBCT- and MR-based synthetic CTs for daily adaptive proton therapy in head and neck patients. Phys. Med. Biol. 2020, 65, 235036. [Google Scholar] [CrossRef] [PubMed]
- Yao, W.; Zhang, B.; Han, D.; Polf, J.; Vedam, S.; Lasio, G.; Yi, B. Use of CBCT plus plan robustness for reducing QACT frequency in intensity-modulated proton therapy: Head-and-neck cases. Med. Phys. 2022, 49, 6794–6801. [Google Scholar] [CrossRef] [PubMed]
- Dzierma, Y.; Mikulla, K.; Richter, P.; Bell, K.; Melchior, P.; Nuesken, F.; Rübe, C. Imaging dose and secondary cancer risk in image-guided radiotherapy of pediatric patients. Radiat. Oncol. 2018, 13, 168. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Pham, T.T.; Whelan, B.; Oborn, B.M.; Delaney, G.P.; Vinod, S.; Brighi, C.; Barton, M.; Keall, P. Magnetic resonance imaging (MRI) guided proton therapy: A review of the clinical challenges, potential benefits and pathway to implementation. Radiother. Oncol. 2022, 170, 37–47. [Google Scholar] [CrossRef] [PubMed]
- Stanforth, A.; Lin, L.; Beitler, J.J.; Janopaul-Naylor, J.R.; Chang, C.W.; Press, R.H.; Patel, S.A.; Zhao, J.; Eaton, B.; Schreibmann, E.E.; et al. Onboard cone-beam CT-based replan evaluation for head and neck proton therapy. J. Appl. Clin. Med. Phys. 2022, 23, e13550. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Nesteruk, K.P.; Bobić, M.; Lalonde, A.; Winey, B.A.; Lomax, A.J.; Paganetti, H. CT-on-Rails Versus In-Room CBCT for Online Daily Adaptive Proton Therapy of Head-and-Neck Cancers. Cancers 2021, 13, 5991. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Uh, J.; Wang, C.; Acharya, S.; Krasin, M.J.; Hua, C.H. Training a deep neural network coping with diversities in abdominal and pelvic images of children and young adults for CBCT-based adaptive proton therapy. Radiother. Oncol. 2021, 160, 250–258. [Google Scholar] [CrossRef] [PubMed]
- Harms, J.; Lei, Y.; Wang, T.; McDonald, M.; Ghavidel, B.; Stokes, W.; Curran, W.J.; Zhou, J.; Liu, T.; Yang, X. Cone-beam CT-derived relative stopping power map generation via deep learning for proton radiotherapy. Med. Phys. 2020, 47, 4416–4427. [Google Scholar] [CrossRef] [PubMed]
- Maspero, M.; Houweling, A.C.; Savenije, M.H.F.; van Heijst, T.C.F.; Verhoeff, J.J.C.; Kotte, A.N.T.J.; van den Berg, C.A.T. A single neural network for cone-beam computed tomography-based radiotherapy of head-and-neck, lung and breast cancer. Phys. Imaging Radiat. Oncol. 2020, 14, 24–31. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Schulze, R.; Heil, U.; Gross, D.; Bruellmann, D.D.; Dranischnikow, E.; Schwanecke, U.; Schoemer, E. Artefacts in CBCT: A review. Dentomaxillofac Radiol. 2011, 40, 265–273. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Neppl, S.; Kurz, C.; Köpl, D.; Yohannes, I.; Schneider, M.; Bondesson, D.; Rabe, M.; Belka, C.; Dietrich, O.; Landry, G.; et al. Measurement-based range evaluation for quality assurance of CBCT-based dose calculations in adaptive proton therapy. Med. Phys. 2021, 48, 4148–4159. [Google Scholar] [CrossRef] [PubMed]
- Qiu, R.L.J.; Lei, Y.; Shelton, J.; Higgins, K.; Bradley, J.D.; Curran, W.J.; Liu, T.; Kesarwala, A.H.; Yang, X. Deep learning-based thoracic CBCT correction with histogram matching. Biomed. Phys. Eng. Express 2021, 7, 065040. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Chang, Y.; Liang, Y.; Yang, B.; Qiu, J.; Pei, X.; Xu, X.G. Dosimetric comparison of deformable image registration and synthetic CT generation based on CBCT images for organs at risk in cervical cancer radiotherapy. Radiat. Oncol. 2023, 18, 3. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Sun, H.; Fan, R.; Li, C.; Lu, Z.; Xie, K.; Ni, X.; Yang, J. Imaging Study of Pseudo-CT Synthesized from Cone-Beam CT Based on 3D CycleGAN in Radiotherapy. Front. Oncol. 2021, 11, 603844. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Landry, G.; Nijhuis, R.; Dedes, G.; Handrack, J.; Thieke, C.; Janssens, G.; Orban de Xivry, J.; Reiner, M.; Kamp, F.; Wilkens, J.J.; et al. Investigating CT to CBCT image registration for head and neck proton therapy as a tool for daily dose recalculation. Med. Phys. 2015, 42, 1354–1366. [Google Scholar] [CrossRef] [PubMed]
- Landry, G.; Dedes, G.; Zöllner, C.; Handrack, J.; Janssens, G.; Orban de Xivry, J.; Reiner, M.; Paganelli, C.; Riboldi, M.; Kamp, F.; et al. Phantom based evaluation of CT to CBCT image registration for proton therapy dose recalculation. Phys. Med. Biol. 2015, 60, 595–613. [Google Scholar] [CrossRef] [PubMed]
- Kidar, H.S.; Azizi, H. Assessing the impact of choosing different deformable registration algorithms on cone-beam CT enhancement by histogram matching. Radiat. Oncol. 2018, 13, 217. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Liu, B.; Lerma, F.A.; Wu, J.Z.; Yi, B.Y.; Yu, C. Tissue Density Mapping of Cone Beam CT Images for Accurate Dose Calculations. Int. J. Med. Phys. 2015, 4, 162–171. [Google Scholar] [CrossRef]
- Landry, G.; Kurz, C.; Traverso, A. The role of artificial intelligence in radiotherapy clinical practice. BJR Open. 2023, 5, 20230030. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Dayarathna, S.; Islam, K.T.; Uribe, S.; Yang, G.; Hayat, M.; Chen, Z. Deep learning based synthesis of MRI, CT and PET: Review and analysis. Med. Image Anal. 2024, 92, 103046. [Google Scholar] [CrossRef] [PubMed]
- Bengio, Y.; Goodfellow, I.; Courville, A. Deep Learning; MIT Press: Cambridge, MA, USA, 2017; Volume 1. [Google Scholar]
- Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; Bengio, Y. Generative adversarial networks. Commun. ACM 2020, 63, 139–144. [Google Scholar] [CrossRef]
- Amyar, A.; Ruan, S.; Vera, P.; Decazes, P.; Modzelewski, R. RADIOGAN: Deep convolutional conditional generative adversarial network to generate PET images. In Proceedings of the 7th International Conference on Bioinformatics Research and Applications, Berlin, Germany, 13–15 September 2020; pp. 28–33. [Google Scholar]
- Gatys, L.A. A neural algorithm of artistic style. arXiv 2015, arXiv:1508.06576. [Google Scholar] [CrossRef]
- Isola, P.; Zhu, J.Y.; Zhou, T.; Efros, A.A. Image-to-image translation with conditional adversarial networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA, 21–26 July 2017; pp. 1125–1134. [Google Scholar]
- Park, T.; Efros, A.A.; Zhang, R.; Zhu, J.Y. Contrastive learning for unpaired image-to-image translation. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, 23–28 August 2020, Proceedings, Part. IX 16; Springer International Publishing: New York, NY, USA, 2020; pp. 319–345. [Google Scholar]
- Rusanov, B.; Hassan, G.M.; Reynolds, M.; Sabet, M.; Kendrick, J.; Rowshanfarzad, P.; Ebert, M. Deep learning methods for enhancing cone-beam CT image quality toward adaptive radiation therapy: A systematic review. Med. Phys. 2022, 49, 6019–6054. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Deng, L.; Ji, Y.; Huang, S.; Yang, X.; Wang, J. Synthetic CT generation from CBCT using double-chain-CycleGAN. Comput. Biol. Med. 2023, 161, 106889. [Google Scholar] [CrossRef] [PubMed]
- Yang, B.; Chang, Y.; Liang, Y.; Wang, Z.; Pei, X.; Xu, X.G.; Qiu, J. A Comparison Study Between CNN-Based Deformed Planning CT and CycleGAN-Based Synthetic CT Methods for Improving iCBCT Image Quality. Front. Oncol. 2022, 12, 896795. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Liang, X.; Zhang, Z.; Niu, T.; Yu, S.; Wu, S.; Li, Z.; Zhang, H.; Xie, Y. Iterative image-domain ring artifact removal in cone-beam CT. Phys. Med. Biol. 2017, 62, 5276–5292. [Google Scholar] [CrossRef] [PubMed]
- Liang, X.; Chen, L.; Nguyen, D.; Zhou, Z.; Gu, X.; Yang, M.; Wang, J.; Jiang, S. Generating synthesized computed tomography (CT) from cone-beam computed tomography (CBCT) using CycleGAN for adaptive radiation therapy. Phys. Med. Biol. 2019, 64, 125002. [Google Scholar] [CrossRef] [PubMed]
- Rusanov, B.; Hassan, G.M.; Reynolds, M.; Sabet, M.; Rowshanfarzad, P.; Bucknell, N.; Gill, S.; Dass, J.; Ebert, M. Transformer CycleGAN with uncertainty estimation for CBCT based synthetic CT in adaptive radiotherapy. Phys. Med. Biol. 2024, 69, 035014. [Google Scholar] [CrossRef] [PubMed]
- Rossi, S. The National Centre for Oncological Hadrontherapy (CNAO): Status and perspectives. Phys. Med. 2015, 31, 333–351. [Google Scholar] [CrossRef] [PubMed]
- Pella, A.; Riboldi, M.; Tagaste, B.; Bianculli, D.; Desplanques, M.; Fontana, G.; Cerveri, P.; Seregni, M.; Fattori, G.; Orecchia, R.; et al. Commissioning and quality assurance of an integrated system for patient positioning and setup verification in particle therapy. Technol. Cancer Res. Treat. 2014, 13, 303–314. [Google Scholar] [CrossRef] [PubMed]
- Fattori, G.; Riboldi, M.; Pella, A.; Peroni, M.; Cerveri, P.; Desplanques, M.; Fontana, G.; Tagaste, B.; Valvo, F.; Orecchia, R.; et al. Image guided particle therapy in CNAO room 2: Implementation and clinical validation. Phys. Med. 2015, 31, 9–15. [Google Scholar] [CrossRef] [PubMed]
- Belotti, G.; Rossi, M.; Pella, A.; Cerveri, P.; Baroni, G. A new system for in-room image guidance in particle therapy at CNAO. Phys. Med. 2023, 114, 103162. [Google Scholar] [CrossRef] [PubMed]
- Rossi, M.; Belotti, G.; Paganelli, C.; Pella, A.; Barcellini, A.; Cerveri, P.; Baroni, G. Image-based shading correction for narrow-FOV truncated pelvic CBCT with deep convolutional neural networks and transfer learning. Med. Phys. 2021, 48, 7112–7126. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Yang, Y.; Zhang, D.; Yang, F.; Teng, M.; Du, Y.; Huang, K. Post-processing method for the removal of mixed ring artifacts in CT images. Opt. Express 2020, 28, 30362–30378. [Google Scholar] [CrossRef] [PubMed]
- Huo, Q.; Li, J.; Lu, Y.; Yan, Z. Removing ring artifacts in CBCT images via L0 smoothing. Int. J. Imaging Syst. Technol. 2016, 26, 284–294. [Google Scholar] [CrossRef]
- Chen, L.; Liang, X.; Shen, C.; Nguyen, D.; Jiang, S.; Wang, J. Synthetic CT generation from CBCT images via unsupervised deep learning. Phys. Med. Biol. 2021, 66, 115019. [Google Scholar] [CrossRef] [PubMed]
- Zhu, J.Y.; Park, T.; Isola, P.; Efros, A.A. Unpaired image-to-image translation using cycle-consistent adversarial networks. In Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy, 22–29 October 2017; pp. 2223–2232. [Google Scholar]
- Amyar, A.; Guo, R.; Cai, X.; Assana, S.; Chow, K.; Rodriguez, J.; Yankama, T.; Cirillo, J.; Pierce, P.; Goddu, B.; et al. Impact of deep learning architectures on accelerated cardiac T1 mapping using MyoMapNet. NMR Biomed. 2022, 35, e4794. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Sara, U.; Akter, M.; Uddin, M. Image Quality Assessment through FSIM, SSIM, MSE and PSNR—A Comparative Study. J. Comput. Commun. 2019, 7, 8–18. [Google Scholar] [CrossRef]
- Wang, Z.; Bovik, A.C.; Sheikh, H.R.; Simoncelli, E.P. Image quality assessment: From error visibility to structural similarity. IEEE Trans. Image Process. 2004, 13, 600–612. [Google Scholar] [CrossRef] [PubMed]
- Kida, S.; Kaji, S.; Nawa, K.; Imae, T.; Nakamoto, T.; Ozaki, S.; Ohta, T.; Nozawa, Y.; Nakagawa, K. Visual enhancement of Cone-beam CT by use of CycleGAN. Med. Phys. 2020, 47, 998–1010. [Google Scholar] [CrossRef] [PubMed]
- Reiners, K.; Dagan, R.; Holtzman, A.; Bryant, C.; Andersson, S.; Nilsson, R.; Hong, L.; Johnson, P.; Zhang, Y. CBCT-Based Dose Monitoring and Adaptive Planning Triggers in Head and Neck PBS Proton Therapy. Cancers 2023, 15, 3881. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Liu, Y.; Lei, Y.; Wang, T.; Fu, Y.; Tang, X.; Curran, W.J.; Liu, T.; Patel, P.; Yang, X. CBCT-based synthetic CT generation using deep-attention CycleGAN for pancreatic adaptive radiotherapy. Med. Phys. 2020, 47, 2472–2483. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Thummerer, A.; Zaffino, P.; Meijers, A.; Marmitt, G.G.; Seco, J.; Steenbakkers, R.J.H.M.; Langendijk, J.A.; Both, S.; Spadea, M.F.; Knopf, A.C. Comparison of CBCT based synthetic CT methods suitable for proton dose calculations in adaptive proton therapy. Phys. Med. Biol. 2020, 65, 095002. [Google Scholar] [CrossRef] [PubMed]
- Ates, O.; Uh, J.; Pirlepesov, F.; Hua, C.H.; Merchant, T.E.; Krasin, M.J. Monitoring of Interfractional Proton Range Verification and Dosimetric Impact Based on Daily CBCT for Pediatric Patients with Pelvic Tumors. Cancers 2023, 15, 4200. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Parrella, G.; Vai, A.; Nakas, A.; Garau, N.; Meschini, G.; Camagni, F.; Molinelli, S.; Barcellini, A.; Pella, A.; Ciocca, M.; et al. Synthetic CT in Carbon Ion Radiotherapy of the Abdominal Site. Bioengineering 2023, 10, 250. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Rossi, M.; Cerveri, P. Comparison of Supervised and Unsupervised Approaches for the Generation of Synthetic CT from Cone-Beam CT. Diagnostics 2021, 11, 1435. [Google Scholar] [CrossRef] [PubMed]
- Giacometti, V.; Hounsell, A.R.; McGarry, C.K. A review of dose calculation approaches with cone beam CT in photon and proton therapy. Phys. Med. 2020, 76, 243–276. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Thing, R.S.; Nilsson, R.; Andersson, S.; Berg, M.; Lund, M.D. Evaluation of CBCT based dose calculation in the thorax and pelvis using two generic algorithms. Phys. Med. 2022, 103, 157–165. [Google Scholar] [CrossRef] [PubMed]
- Seller Oria, C.; Thummerer, A.; Free, J.; Langendijk, J.A.; Both, S.; Knopf, A.C.; Meijers, A. Range probing as a quality control tool for CBCT-based synthetic CTs: In vivo application for head and neck cancer patients. Med. Phys. 2021, 48, 4498–4505. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Sibolt, P.; Andersson, L.M.; Calmels, L.; Sjöström, D.; Bjelkengren, U.; Geertsen, P.; Behrens, C.F. Clinical implementation of artificial intelligence-driven cone-beam computed tomography-guided online adaptive radiotherapy in the pelvic region. Phys. Imaging Radiat. Oncol. 2020, 17, 1–7. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Kaushik, S.; Ödén, J.; Sharma, D.S.; Fredriksson, A.; Toma-Dasu, I. Generation and evaluation of anatomy-preserving virtual CT for online adaptive proton therapy. Med. Phys. 2024, 51, 1536–1546. [Google Scholar] [CrossRef] [PubMed]
- Liu, H.; Schaal, D.; Curry, H.; Clark, R.; Magliari, A.; Kupelian, P.; Khuntia, D.; Beriwal, S. Review of cone beam computed tomography based online adaptive radiotherapy: Current trend and future direction. Radiat. Oncol. 2023, 18, 144. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Chan, Y.C.I.; Li, M.; Thummerer, A.; Parodi, K.; Belka, C.; Kurz, C.; Landry, G. Minimum imaging dose for deep learning-based pelvic synthetic computed tomography generation from cone beam images. Phys. Imaging Radiat. Oncol. 2024, 30, 100569. [Google Scholar] [CrossRef] [PubMed]
- Spadea, M.F.; Maspero, M.; Zaffino, P.; Seco, J. Deep learning based synthetic-CT generation in radiotherapy and PET: A review. Med. Phys. 2021, 48, 6537–6566. [Google Scholar] [CrossRef] [PubMed]
- Trnkova, P.; Zhang, Y.; Toshito, T.; Heijmen, B.; Richter, C.; Aznar, M.C.; Albertini, F.; Bolsi, A.; Daartz, J.; Knopf, A.C.; et al. A survey of practice patterns for adaptive particle therapy for interfractional changes. Phys. Imaging Radiat. Oncol. 2023, 26, 100442. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Tsai, P.; Tseng, Y.L.; Shen, B.; Ackerman, C.; Zhai, H.A.; Yu, F.; Simone, C.B., 2nd; Choi, J.I.; Lee, N.Y.; Kabarriti, R.; et al. The Applications and Pitfalls of Cone-Beam Computed Tomography-Based Synthetic Computed Tomography for Adaptive Evaluation in Pencil-Beam Scanning Proton Therapy. Cancers 2023, 15, 5101. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Paganetti, H.; Botas, P.; Sharp, G.C.; Winey, B. Adaptive proton therapy. Phys. Med. Biol. 2021, 66, 22TR01. [Google Scholar] [CrossRef] [PubMed] [PubMed Central]
- Nenoff, L.; Matter, M.; Charmillot, M.; Krier, S.; Uher, K.; Weber, D.C.; Lomax, A.J.; Albertini, F. Experimental validation of daily adaptive proton therapy. Phys. Med. Biol. 2021, 66, 205010. [Google Scholar] [CrossRef] [PubMed]
Clinical Data | Treatment Data | Imaging Data | Training Configuration | |||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Study ID | Sex | Age at PT (Years) | Classification * | Treatment Site | Diagnosis | Particle | Nr. Fx | Dose Per Fx (GyRBE/fx) | Total Dose (GyRBE) | Nr. CTs | Nr CBCTs ** | Nr. CBCTs Included | TR | TS | Discarded | Reason for Discarding |
PT_01 | F | 15 | Paediatric | Sacrum | Ewing’s Sarcoma | P | 33 | 1.8 × 28 + 1.8 × 5 (boost) | 59.4 | 7 | 6 | 6 | 6 | |||
PT_02 | M | 16 | Paediatric | Right Sacrum | Osteosarcoma | C | 16 | 4.6 × 9 + 4.6 × 7 (boost) | 73.6 | 3 | 2 | 0 | 2 | CT import | ||
PT_03 | M | 10 | Paediatric | Right Iliac Crest | Ewing’s Sarcoma | P | 33 | 1.8 × 28 + 1.8 × 5 (boost) | 59.4 | 5 | 4 | 4 | 4 | |||
PT_04 | M | 15 | Paediatric | Left Pelvis | Osteosarcoma (relapse) | C | 16 | 4.2 × 16 | 67.2 | 5 | 3 | 3 | 3 | |||
PT_05 | M | 15 | Paediatric | Left Pelvis | Osteosarcoma (relapse) | C | 15 | 4 × 15 | 60 | 3 | 2 | 2 | 2 | |||
PT_06 | M | 15 | Paediatric | Sacrum-Coccyx | Ewing’s Sarcoma | P | 33 | 1.8 × 30 + 1.8 × 3 (boost) | 59.4 | 7 | 6 | 5 | 5 | 1 | Registration | |
PT_07 | M | 14 | Paediatric | Sacrum-Coccyx | Ewing’s Sarcoma | P | 33 | 1.8 × 30 + 1.8 × 3 (boost) | 59.4 | 7 | 6 | 6 | 6 | |||
PT_08 | F | 16 | Paediatric | Left Gluteus | Rhabdomyosarcoma | P | 31 | 1.8 × 28 + 1.8 × 3 (boost) | 55.8 | 4 | 3 | 3 | 3 | |||
PT_09 | M | 7 | Paediatric | Prostate | Rhabdomyosarcoma | P | 30 | 1.8 × 30 | 54 | 3 | 3 | 0 | 3 | Image quality | ||
PT_10 | F | 16 | Paediatric | Right Gluteus | Sarcoma | P | 28 | 1.8 × 28 | 50.4 | 4 | 3 | 3 | 3 | |||
PT_11 | F | 17 | Paediatric | Sacrum | Ewing’s sarcoma (relapse) | C | 15 | 4 × 15 | 60 | 2 | 1 | 1 | 1 | |||
PT_12 | M | 17 | Paediatric | Right Pelvis | Ewing’s Sarcoma | P | 33 | 1.8 × 33 | 59.4 | 6 | 5 | 5 | 5 | |||
PT_13 | F | 20 | Young Adult | Sacrum/Left Iliac Crest | Osteosarcoma | C | 16 | 4.6 × 16 | 73.6 | 2 | 1 | 1 | 1 | |||
PT_14 | F | 20 | Young Adult | Right Sciatic Nerve | MPNST | C | 16 | 4.6 × 16 | 73.6 | 2 | 1 | 0 | 1 | CT import | ||
PT_15 | M | 24 | Young Adult | Sacrum-Coccyx | Cordoma | C | 16 | 4.8 × 13 + 4.8 × 3 (boost) | 76.8 | 2 | 1 | 1 | 1 | |||
PT_16 | F | 27 | Young Adult | Right Iliac Crest | Condrosarcoma | C | 16 | 4.6 × 16 | 73.6 | 2 | 1 | 1 | 1 | |||
PT_17 | M | 27 | Young Adult | Pelvis | Sarcoma | C | 15 | - | - | 3 | 0 | 0 | ||||
PT_18 | F | 27 | Young Adult | Left Pelvis | Sarcoma | C | 12 | 4.6 × 12 | 55.2 | 2 | 1 | 0 | 1 | Registration | ||
PT_19 | F | 28 | Young Adult | Vagina | Adenocarcinoma | C | 8 | 4.8 × 8 | 38.4 | 3 | 2 | 2 | 2 | |||
PT_20 | F | 30 | Young Adult | Sacrum | Emangioma | P | 30 | 2 × 30 | 60 | 4 | 3 | 1 | 1 | 2 | ||
PT_21 | M | 30 | Young Adult | Pelvis | Osteosarcoma | C | 16 | 4.8 × 13 | 62.4 | 3 | 1 | 0 | 1 | Registration | ||
TOT | 79 | 55 | 44 | 36 | 8 | 11 |
Res-Net | U-Net | |||||
---|---|---|---|---|---|---|
SSIM (Median, IQR) | PSNR (Median, IQR) | SSIM (Median, IQR) | PSNR (Median, IQR) | |||
PT_05 | V1 | Pre | 0.73, 0.03 | 49.09, 2.88 | 0.73, 0.03 | 49.09, 2.88 |
Post | 0.79, 0.05 | 51.97, 2.96 | 0.75, 0.04 | 51.39, 3.67 | ||
p-value | <0.05 | <0.05 | <0.05 | <0.05 | ||
V2 | Pre | 0.69, 0.05 | 48.48, 2.70 | 0.69, 0.05 | 48.48, 2.70 | |
Post | 0.73, 0.07 | 50.68, 3.85 | 0.72, 0.06 | 50.27, 3.94 | ||
p-value | <0.05 | <0.05 | <0.05 | <0.05 | ||
PT_08 | V1 | Pre | 0.58, 0.17 | 49.85, 1.93 | 0.58, 0.17 | 49.85, 1.93 |
Post | 0.57, 0.20 | 50.90, 1.60 | 0.57, 0.21 | 50.28, 2.25 | ||
p-value | <0.05 | <0.05 | 0.19 | <0.05 | ||
V2 | Pre | 0.54, 0.18 | 49.69, 1.82 | 0.54, 0.18 | 49.69, 1.82 | |
Post | 0.54, 0.19 | 50.42, 0.72 | 0.54, 0.21 | 50.32, 1.33 | ||
p-value | <0.05 | <0.05 | 0.41 | <0.05 | ||
V3 | Pre | 0.55, 0.17 | 49.02, 1.31 | 0.55, 0.17 | 49.02, 1.31 | |
Post | 0.54, 0.20 | 49.46, 0.21 | 0.54, 0.20 | 49.48, 1.00 | ||
p-value | <0.05 | <0.05 | 0.90 | <0.05 | ||
PT_13 | Pre | 0.75, 0.04 | 48.66, 2.19 | 0.75, 0.04 | 48.66, 2.19 | |
Post | 0.83, 0.02 | 52.96, 3.47 | 0.78, 0.04 | 50.73, 2.34 | ||
p-value | <0.05 | <0.05 | <0.05 | <0.05 | ||
PT_15 | Pre | 0.75, 0.03 | 48.24, 2.14 | 0.75, 0.03 | 48.24, 2.14 | |
Post | 0.80, 0.03 | 52.95, 3.02 | 0.78, 0.05 | 50.48, 1.60 | ||
p-value | <0.05 | <0.05 | <0.05 | <0.05 | ||
PT_20 | Pre | 0.68, 0.07 | 44.40, 2.10 | 0.68, 0.07 | 44.40, 2.10 | |
Post | 0.73, 0.08 | 48.98, 2.89 | 0.73, 0.06 | 48.17, 2.05 | ||
p-value | <0.05 | <0.05 | <0.05 | <0.05 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Pepa, M.; Taleghani, S.; Sellaro, G.; Mirandola, A.; Colombo, F.; Vennarini, S.; Ciocca, M.; Paganelli, C.; Orlandi, E.; Baroni, G.; et al. Unsupervised Deep Learning for Synthetic CT Generation from CBCT Images for Proton and Carbon Ion Therapy for Paediatric Patients. Sensors 2024, 24, 7460. https://doi.org/10.3390/s24237460
Pepa M, Taleghani S, Sellaro G, Mirandola A, Colombo F, Vennarini S, Ciocca M, Paganelli C, Orlandi E, Baroni G, et al. Unsupervised Deep Learning for Synthetic CT Generation from CBCT Images for Proton and Carbon Ion Therapy for Paediatric Patients. Sensors. 2024; 24(23):7460. https://doi.org/10.3390/s24237460
Chicago/Turabian StylePepa, Matteo, Siavash Taleghani, Giulia Sellaro, Alfredo Mirandola, Francesca Colombo, Sabina Vennarini, Mario Ciocca, Chiara Paganelli, Ester Orlandi, Guido Baroni, and et al. 2024. "Unsupervised Deep Learning for Synthetic CT Generation from CBCT Images for Proton and Carbon Ion Therapy for Paediatric Patients" Sensors 24, no. 23: 7460. https://doi.org/10.3390/s24237460
APA StylePepa, M., Taleghani, S., Sellaro, G., Mirandola, A., Colombo, F., Vennarini, S., Ciocca, M., Paganelli, C., Orlandi, E., Baroni, G., & Pella, A. (2024). Unsupervised Deep Learning for Synthetic CT Generation from CBCT Images for Proton and Carbon Ion Therapy for Paediatric Patients. Sensors, 24(23), 7460. https://doi.org/10.3390/s24237460