Contrast agent dose reduction in computed tomography with deep learning using a conditional generative adversarial network

Abstract

Objectives To reduce the dose of intravenous iodine-based contrast media (ICM) in CT through virtual contrast-enhanced images using generative adversarial networks. Methods Dual-energy CTs in the arterial phase of 85 patients were randomly split into an 80/20 train/test collective. Four different generative adversarial networks (GANs) based on image pairs, which comprised one image with virtually reduced ICM and the original full ICM CT slice, were trained, testing two input formats (2D and 2.5D) and two reduced ICM dose levels (−50% and −80%). The amount of intravenous ICM was reduced by creating virtual non-contrast series using dual-energy and adding the corresponding percentage of the iodine map. The evaluation was based on different scores (L1 loss, SSIM, PSNR, FID), which evaluate the image quality and similarity. Additionally, a visual Turing test (VTT) with three radiologists was used to assess the similarity and pathological consistency. Results The −80% models reach an SSIM of > 98%, PSNR of > 48, L1 of between 7.5 and 8, and an FID of between 1.6 and 1.7. In comparison, the −50% models reach a SSIM of > 99%, PSNR of > 51, L1 of between 6.0 and 6.1, and an FID between 0.8 and 0.95. For the crucial question of pathological consistency, only the 50% ICM reduction networks achieved 100% consistency, which is required for clinical use. Conclusions The required amount of ICM for CT can be reduced by 50% while maintaining image quality and diagnostic accuracy using GANs. Further phantom studies and animal experiments are required to confirm these initial results. Key Points • The amount of contrast media required for CT can be reduced by 50% using generative adversarial networks. • Not only the image quality but especially the pathological consistency must be evaluated to assess safety. • A too pronounced contrast media reduction could influence the pathological consistency in our collective at 80%.

Contrast Media Reduction in Computed Tomography With Deep Learning Using a Generative Adversarial Network in an Experimental Animal Study

“…In this context, it was shown that pathologies in artificially inhomogeneous collectives could be inserted or removed by GANs 25 . To avoid this, close monitoring of the training collective was ensured by pairing and coregistering the animals with themselves and checking pathological consistency, as described in our preliminary work 14 . In context of the present study pathological consistency means that no structure has been added or removed, all vessels are contrasted as expected, and no pathology has been added.…”

“…14 Here, networks are trained to perform an image-to-image conversion from a low contrast dose images to a normal contrast dose images to reduce the required amount of contrast media by 50% to 80%. 14 So far, this method could only be evaluated for CT on virtual contrast-reduced images, based on a combination of dual-energy derived virtual noncontrast and iodine images. However, with these, it is uncertain whether contrast information remains in the image that the network can use for contrast enhancement.…”

An Overview of Artificial Intelligence in Oncology

“…Also, generative networks, including additional types of architectures (e.g., autoencoders [AEs] and variational autoencoders [VAEs]), have the capability of improving the acquisition of multimodality imaging, such as MRI and CT scans, reducting radiation dose and use of intravenous contrast [ 29–31 ]. Since oncology patients must do routine scans for tumor staging, AE and VAE have the potential to reduce healthcare costs while improving patient safety.…”

A Review of Generative Adversarial Networks in Cancer Imaging: New Applications, New Solutions

“…Conditional GANs offer the possibility to generate one modality from another, alleviating the need to actually perform the potentially more harmful screenings-i.e. high-dose CT, PET-that expose patients to radiation, or require invasive contrast agents such as intravenous iodine-based contrast media (ICM) in CT [102], gadolinium-based contrast agents in MRI [103](in Table 5) or radioactive tracers in PET [104,105]. Furthermore, extending the acquisition modalities used in a given task would also enhance the performance and generalisability of AI models, allowing them to learn shared representations among these imaging modalities [24,106].…”