Tumor-Attentive Segmentation-Guided GAN for Synthesizing Breast Contrast-Enhanced MRI Without Contrast Agents

Abstract

Objective: Breast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is a sensitive imaging technique critical for breast cancer diagnosis. However, the administration of contrast agents poses a potential risk. This can be avoided if contrast-enhanced MRI can be obtained without using contrast agents. Thus, we aimed to generate T1-weighted contrast-enhanced MRI (ceT1) images from pre-contrast T1 weighted MRI (preT1) images in the breast. Methods: We proposed a generative adversarial network to synthesize ceT1 from preT1 breast images that adopted a local discriminator and segmentation task network to focus specifically on the tumor region in addition to the whole breast. The segmentation network performed a related task of segmentation of the tumor region, which allowed important tumor-related information to be enhanced. In addition, edge maps were included to provide explicit shape and structural information. Our approach was evaluated and compared with other methods in the local (n = 306) and external validation (n = 140) cohorts. Four evaluation metrics of normalized mean squared error (NRMSE), Pearson cross-correlation coefficients (CC), peak signal-to-noise ratio (PSNR), and structural similarity index map (SSIM) for the whole breast and tumor region were measured. An ablation study was performed to evaluate the incremental benefits of various components in our approach. Results: Our approach performed the best with an NRMSE 25.65, PSNR 54.80 dB, SSIM 0.91, and CC 0.88 on average, in the local test set. Conclusion: Performance gains were replicated in the validation cohort. Significance: We hope that our method will help patients avoid potentially harmful contrast agents. Clinical and Translational Impact Statement—Contrast agents are necessary to obtain DCE-MRI which is essential in breast cancer diagnosis. However, administration of contrast agents may cause side effects such as nephrogenic systemic fibrosis and risk of toxic residue deposits. Our approach can generate DCE-MRI without contrast agents using a generative deep neural network. Thus, our approach could help patients avoid potentially harmful contrast agents resulting in an improved diagnosis and treatment workflow for breast cancer.

Quantifying the quality of GAN-synthesized images: a study on synthesizing post-contrast sequences from pre-contrast sequences in breast DCE-MRI

“…1 GANs have also been applied in the MRI modality, where segmentation-guided GANs are proposed to synthesize T-weighted contrast-enhanced MRI images from pre-contrast T1 weighted MRI images. 2 However, the majority of these studies solely depended on visual assessments or generic metrics to evaluate the quality of the synthesized images, lacking quantitative analyses of the specific effects for targeted clinical tasks. The primary metrics employed in the studies to assess the quality of synthesized images include normalized mean squared error (NRMSE), Pearson cross-correlation coefficients (CC), peak signal-to-noise ratio (PSNR), and structural similarity index map (SSIM) for both tumor and the entire breast area.…”

Pre- to post-contrast breast MRI synthesis for enhanced tumour segmentation

“…8,9,11 With recent advances in deep learning, training deep generative models to generate synthetic contrastenhanced imaging data as an alternative to contrast agent administration has been becoming a promising field of research. 12,13 For instance, Kim et al 14 provide a tumour-attentive segmentation-guided generative adversarial network (GAN) 15 that generates a contrast-enhanced T1 breast MRI image from its pre-contrast counterpart, while being guided by the predictions of a surrogate segmentation network. Similarly, Zhao et al 16 propose Tripartite-GAN to synthesise contrast-enhanced from non contrast-enhanced liver MRI with a chained tumour detection model.…”

Impact of Non-Contrast Enhanced Imaging Input Sequences on the Generation of Virtual Contrast-Enhanced Breast MRI Scans using Neural Networks

“…Given the novelty of this research field, it is expected that a high variability is yet observed in the literature with regards to the choice of input data used to train vCE imaging neural networks. T1-weighted (T1w) sequences (20,21,23) have been used for vCE breast MRI as well as combinations of T1w and T2-weighted (T2w) image acquisitions (22) or more complex protocols including diffusion-weighted imaging (DWI)(26).…”

“…The generation of virtual contrast-enhanced (vCE) MRI scans from unenhanced acquisitions using deep learning was first reported in brain studies (14)(15)(16)(17)(18)(19). However, several recent publications (20)(21)(22)(23)(24)(25)(26) have shown the technical feasibility of such approach for breast imaging. Given the novelty of this research field, it is expected that a high variability is yet observed in the literature with regards to the choice of input data used to train vCE imaging neural networks.…”