Dose evaluation of fast synthetic-CT generation using a generative adversarial network for general pelvis MR-only radiotherapy

Matteo Maspero,Peter R Seevinck,Mark H F Savenije,Ina M Jurgenliemk-Schulz,Martijn P W Intven,Linda G W Kerkmeijer,Anna M Dinkla,Cornelis A T Van Den Berg

doi:10.1088/1361-6560/aada6d

Matteo Maspero, Peter R Seevinck + Show 6 more

Open Access

https://doi.org/10.1088/1361-6560/aada6d

Copy DOI

Abstract

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 for accurate MR-based dose calculations in the entire pelvis.A study was conducted on data of 91 patients with prostate (59), rectal (18) and cervical (14) 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%. The average gamma pass rates using the 3%, 3 mm and 2%, 2 mm criteria were above 97 and 91%, respectively, for all volumes of interests considered. All DVH points calculated on sCT differed less than ±2.5% from the corresponding points on CT.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 for integration in an MR-guided radiotherapy workflow.

Highlights

Results suggest that accurate magnetic resonance (MR)-based dose calculation using synthetic CT (sCT) images generated with a conditional generative adversarial network (cGAN) trained on prostate cancer patients is feasible for the entire pelvis
After training, inferring the generator network to obtain sCT images was performed in 5.6 s per patient on graphical processing unit (GPU) and four times longer (21 s) on central processing unit (CPU)
We showed that training a cGAN on prostate cancer patients results in MR-based dose calculation within an average dose difference of 0.5% compared to CT-based calculations (CT–sCT)

Summary

Introduction

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. The sCT generation required 5.6 s and 21 s for a single patient volume on a GPU and CPU, respectively. SCT images resulted in a higher dose to the target of maximum 0.3%. The average gamma pass rates using the 3%, 3 mm and 2%, 2 mm criteria were above 97 and 91%, respectively, for all volumes of interests considered. All DVH points calculated on sCT differed less than ±2.5% from the corresponding points on CT

Objectives

Methods

Results

Discussion

Conclusion