Dose evaluation of fast synthetic-CT generation using a generative adversarial network for general pelvis MR-only radiotherapy
| Autor: | Linda G W Kerkmeijer, Peter R. Seevinck, Matteo Maspero, Mark H. F. Savenije, Martijn Intven, Cornelis A. T. van den Berg, Anna M. Dinkla, Ina M. Jürgenliemk-Schulz |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2018 |
| Předmět: |
Male
Computer science medicine.medical_treatment Radiotherapy Planning Tomography X-Ray Computed/methods Uterine Cervical Neoplasms Radiotherapy Image-Guided/methods computer.software_genre pseudo-CT 030218 nuclear medicine & medical imaging Magnetic Resonance Imaging/methods Prostate cancer 0302 clinical medicine Voxel Prostate Computer-Assisted/methods Non-U.S. Gov't Tomography MR-only Radiotherapy medicine.diagnostic_test Radiological and Ultrasound Technology Research Support Non-U.S. Gov't Intensity-Modulated/methods dose calculations Radiotherapy Dosage X-Ray Computed/methods medicine.anatomical_structure Radiology Nuclear Medicine and imaging 030220 oncology & carcinogenesis Female MRI CT neural network medical imaging FOS: Physical sciences Research Support Pelvis/diagnostic imaging Pelvis 03 medical and health sciences pseudo CT Magnetic resonance imaging Histogram medicine Journal Article cancer Humans Radiology Nuclear Medicine and imaging External beam radiotherapy Radiotherapy Planning Computer-Assisted/methods Radiotherapy business.industry Radiotherapy Planning Computer-Assisted Image-Guided/methods generative adversarial network Prostatic Neoplasms Deep learning Radiotherapy Intensity-Modulated/methods medicine.disease Physics - Medical Physics Prostatic Neoplasms/radiotherapy Radiation therapy Uterine Cervical Neoplasms/radiotherapy Medical Physics (physics.med-ph) Radiotherapy Intensity-Modulated Nuclear medicine business Tomography X-Ray Computed computer Radiotherapy Image-Guided |
| Zdroj: | Physics in Medicine and Biology, 63(18), 185001. IOP Publishing Ltd. |
| ISSN: | 0031-9155 |
| Popis: | To enable magnetic resonance (MR)-only radiotherapy and facilitate modelling of radiation attenuation in humans, synthetic-CT (sCT) images need to be generated. Considering the application of MR-guided radiotherapy and online adaptive replanning, sCT generation should occur within minutes. This work aims at assessing whether an existing deep learning network can rapidly generate sCT images to be used for accurate MR-based dose calculations in the entire pelvis. A study was conducted on data of 91 patients with prostate, rectal and cervical cancer who underwent external beam radiotherapy acquiring both CT and MRI for patients' simulation. Dixon reconstructed water, fat and in-phase images obtained from a conventional dual gradient-recalled echo sequence were used to generate sCT images. A conditional generative adversarial network (cGAN) was trained in a paired fashion on 2D transverse slices of 32 prostate cancer patients. The trained network was tested on the remaining patients to generate sCT images. For 30 patients in the test set, dose recalculations of the clinical plan were performed on sCT images. Dose distributions were evaluated comparing voxel-based dose differences, gamma and dose-volume histogram (DVH) analysis. The sCT generation required 5.6 s and 21 s for a single patient volume on a GPU and CPU, respectively. On average, sCT images resulted in a higher dose to the target of maximum 0.3%. Results suggest that accurate MR-based dose calculation using sCT images generated with a cGAN trained on prostate cancer patients is feasible for the entire pelvis. The sCT generation was sufficiently fast to be integrated into an MR-guided radiotherapy workflow. Accepted for publication in Physics in Medicine and Biology, in press (2018) |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|