3D conditional generative adversarial networks for high-quality PET image estimation at low dose

Abstract

Positron emission tomography (PET) is a widely used imaging modality, providing insight into both the biochemical and physiological processes of human body. Usually, a full dose radioactive tracer is required to obtain high-quality PET images for clinical needs. This inevitably raises concerns about potential health hazards. On the other hand, dose reduction may cause the increased noise in the reconstructed PET images, which impacts the image quality to a certain extent. In this paper, in order to reduce the radiation exposure while maintaining the high quality of PET images, we propose a novel method based on 3D conditional generative adversarial networks (3D c-GANs) to estimate the high-quality full-dose PET images from low-dose ones. Generative adversarial networks (GANs) include a generator network and a discriminator network which are trained simultaneously with the goal of one beating the other. Similar to GANs, in the proposed 3D c-GANs, we condition the model on an input low-dose PET image and generate a corresponding output full-dose PET image. Specifically, to render the same underlying information between the low-dose and full-dose PET images, a 3D U-net-like deep architecture which can combine hierarchical features by using skip connection is designed as the generator network to synthesize the full-dose image. In order to guarantee the synthesized PET image to be close to the real one, we take into account of the estimation error loss in addition to the discriminator feedback to train the generator network. Furthermore, a concatenated 3D c-GANs based progressive refinement scheme is also proposed to further improve the quality of estimated images. Validation was done on a real human brain dataset including both the normal subjects and the subjects diagnosed as mild cognitive impairment (MCI). Experimental results show that our proposed 3D c-GANs method outperforms the benchmark methods and achieves much better performance than the state-of-the-art methods in both qualitative and quantitative measures.

Introduction to Radiomics

“…Recently, deep learning for PET reconstruction was proposed, either for post-processing of conventionally reconstructed images (33), within an iterative reconstruction framework (34), or to directly map PET data into images (35). These methods have been able to restore or reconstruct PET images of higher quality compared to conventional OSEM, making images less noisy without sacrificing resolution.…”

Improving the Quality of Synthetic FLAIR Images with Deep Learning Using a Conditional Generative Adversarial Network for Pixel-by-Pixel Image Translation

“…Notably, images with a higher PSNR and lower NRMSE theoretically exhibit higher image quality. 29 Synthetic T1-weighted images, which were inherently aligned to synthetic FLAIR images, were skull-stripped by the intracranial masks and segmented into GM, WM, and CSF using the FMRIB Automated Segmentation Tool (FAST; http://fsl.fmrib.ox. ac.uk/fsl/fslwiki/fast).…”

A neural lens for super-resolution biological imaging

“…While parametric modelling approaches have been applied in order to enhance medical imaging [8], recent work using deep learning approaches has also shown the potential for the application of neural networks (NNs) for enhancing image resolution. Examples for such an application include, a 4x upscaling on photographic images [9], optical microscopy (improving the resolution from 40x to 100x) [10], dental imaging [11], phase imaging [12], fluorescence microscopy [13], magnetic resonance imaging [14], SEM imaging [15,16], positronemission tomography [17], stochastic optical reconstruction microscopy [18], and ultrasound imaging [19]. NNs are also well-suited to the classification of objects in images, and accordingly the classification of biological, pollution and colloidal particles from images and scattering patterns has also been demonstration [20][21][22][23][24].…”