Generation and Evaluation of Synthetic Computed Tomography (CT) from Cone-Beam CT (CBCT) by Incorporating Feature-Driven Loss into Intensity-Based Loss Functions in Deep Convolutional Neural Network

Abstract

Deep convolutional neural network (CNN) helped enhance image quality of cone-beam computed tomography (CBCT) by generating synthetic CT. Most of the previous works, however, trained network by intensity-based loss functions, possibly undermining to promote image feature similarity. The verifications were not sufficient to demonstrate clinical applicability, either. This work investigated the effect of variable loss functions combining feature- and intensity-driven losses in synthetic CT generation, followed by strengthening the verification of generated images in both image similarity and dosimetry accuracy. The proposed strategy highlighted the feature-driven quantification in (1) training the network by perceptual loss, besides L1 and structural similarity (SSIM) losses regarding anatomical similarity, and (2) evaluating image similarity by feature mapping ratio (FMR), besides conventional metrics. In addition, the synthetic CT images were assessed in terms of dose calculating accuracy by a commercial Monte-Carlo algorithm. The network was trained with 50 paired CBCT-CT scans acquired at the same CT simulator and treatment unit to constrain environmental factors any other than loss functions. For 10 independent cases, incorporating perceptual loss into L1 and SSIM losses outperformed the other combinations, which enhanced FMR of image similarity by 10%, and the dose calculating accuracy by 1–2% of gamma passing rate in 1%/1mm criterion.

Supervised deep learning-based synthetic computed tomography from kilovoltage cone-beam computed tomography images for adaptive radiation therapy in head and neck cancer

“…Jihong et al [24] reported a rate of 95.7% ± 1.9% for sCT images with uncorrected CBCT and 97.1% ± 1.9% for sCT images with corrected CBCT. Meanwhile, Yoo et al [25] achieved an impressive result of 99.7% ± 0.0% for sCT using a combination loss functions for model training. Both studies utilized advanced deep learning techniques, with Jihong et al [24] employing unsupervised learning via CycleGAN with HU correction, and Yoo et al [25] enhancing performance through the integration of perceptual loss into L1 and structural similarity loss functions during model training.…”

“…Meanwhile, Yoo et al [25] achieved an impressive result of 99.7% ± 0.0% for sCT using a combination loss functions for model training. Both studies utilized advanced deep learning techniques, with Jihong et al [24] employing unsupervised learning via CycleGAN with HU correction, and Yoo et al [25] enhancing performance through the integration of perceptual loss into L1 and structural similarity loss functions during model training. These findings suggest that unsupervised deep learning and specialized loss functions can enhance the quality of sCT images, and preprocessing techniques such as HU correction can further improve outcomes.…”

“…These anatomical variations could potentially impact the training of the U-Net algorithm, which relies on paired image data. Therefore, exploring unsupervised models such as GANs could enhance the model's performance, particularly when dealing with unpaired image data (intra-individual co-registration) [18,[24][25][26][27][28][29][30]. Furthermore, the process of rigidly registering the CT and CBCT images might not have been adequate in establishing the required image similarity for network training.…”

Enhancing adaptive proton therapy through CBCT images: Synthetic head and neck CT generation based on 3D vision transformers

“…The initial efforts using deep‐learning‐based solutions were based on U‐Nets 8,13 . Currently, some studies propose the use of DenseNet, 15 while others use conditional generative adversarial networks (cGANs) to generate pCT images 9–12,16 …”

CBCT-to-CT Synthesis for Cervical Cancer Adaptive Radiotherapy via U-Net-Based Model Hierarchically Trained with Hybrid Dataset

“…Hardware-based methods attempt to decrease the influence caused by scattering by utilizing an anti-scatter grid when acquiring CBCT images [15,16]. Image post-processing methods mainly consist of deformation of pCT [17][18][19][20][21], an estimation of scatter kernels [22], Monte Carlo simulations of scatter distribution [23,24], histogram matching [25], and deep learning-based methods [26][27][28][29][30][31]. Deformation of pCT is one of the commonly used methods, which is based on deformable registration between pCT and CBCT.…”