Multi‐sequence MR image‐based synthetic CT generation using a generative adversarial network for head and neck MRI‐only radiotherapy

Abstract

Purpose: The purpose of this study is to investigate the effect of different magnetic resonance (MR) sequences on the accuracy of deep learning-based synthetic computed tomography (sCT) generation in the complex head and neck region. Methods: Four sequences of MR images (T1, T2, T1C, and T1DixonC-water) were collected from 45 patients with nasopharyngeal carcinoma. Seven conditional generative adversarial network (cGAN) models were trained with different sequences (single channel) and different combinations (multi-channel) as inputs. To further verify the cGAN performance, we also used a U-net network as a comparison. Mean absolute error, structural similarity index, peak signal-to-noise ratio, dice similarity coefficient, and dose distribution were evaluated between the actual CTs and sCTs generated from different models. Results: The results show that the cGAN model with multi-channel (i.e., T1 + T2 + T1C + T1DixonC-water) as input to predict sCT achieves higher accuracy than any single MR sequence model. The T1-weighted MR model achieves better results than T2, T1C, and T1DixonC-water models. The comparison between cGAN and U-net shows that the sCTs predicted by cGAN retains additional image details are less blurred and more similar to the actual CT. Conclusions: Conditional generative adversarial network with multiple MR sequences as model input shows the best accuracy. The T1-weighted MR images provide sufficient image information and are suitable for sCT prediction in clinical scenarios with limited acquisition sequences or limited acquisition time.

Comparison of CBCT‐based dose calculation methods in head and neck cancer radiotherapy: from Hounsfield unit to density calibration curve to deep learning

“…Such methods have already been used for pCT generation from MRI in the context of an MRI-only workflow. 52 The two recent MRI H&N studies using conditional GAN 53 or U-Net 54 provided an MAE of approximately 70-75 HU in the entire body contour. However, the aim of pCT generation from CBCT images is different.…”

A review on medical imaging synthesis using deep learning and its clinical applications

“…Qi et al proposed a four‐channel input composed of T1, T2, contrast‐enhanced T1, and contrast‐enhanced T1 Dixon water images. Compared with the results from fewer channels, the four‐channel results demonstrated lower MAE 70 . Florkow et al investigated single and multichannel input using magnitude MR images and Dixon‐reconstructed water, fat, in‐phase and opposed‐phase images obtained from a single T1 multi‐echo gradient‐echo acquisition 71 .…”

Multi-Parametric Deep Learning Model for Prediction of Overall Survival after Postoperative Concurrent Chemoradiotherapy in Glioblastoma Patients

“…CNN is a shift- or space-invariant deep learning model, which can automatically extract optimal features by itself from the given data to achieve the best performance. Thus, it is suitable for OS prediction considering positional relationship of features between MRI series [ 29 ]. In this study, we calculated the mean weight from the optimized weight of one-by-one convolution filter to analyze the weight of four pulse sequence MRI series.…”