TomoGAN: Low-Dose Synchrotron X-Ray Tomography with Generative Adversarial Networks

Abstract

Synchrotron-based x-ray tomography is a noninvasive imaging technique that allows for reconstructing the internal structure of materials at high spatial resolutions from tens of micrometers to a few nanometers. In order to resolve sample features at smaller length scales, however, a higher radiation dose is required. Therefore, the limitation on the achievable resolution is set primarily by noise at these length scales. We present \TOMOGAN{}, a denoising technique based on generative adversarial networks, for improving the quality of reconstructed images for low-dose imaging conditions. We evaluate our approach in two photon-budget-limited experimental conditions: (1) sufficient number of low-dose projections (based on Nyquist sampling), and (2) insufficient or limited number of high-dose projections. In both cases the angular sampling is assumed to be isotropic, and the photon budget throughout the experiment is fixed based on the maximum allowable radiation dose on the sample. Evaluation with both simulated and experimental datasets shows that our approach can significantly reduce noise in reconstructed images, improving the structural similarity score of simulation and experimental data from 0.18 to 0.9 and from 0.18 to 0.41, respectively. Furthermore, the quality of the reconstructed images with filtered back projection followed by our denoising approach exceeds that of reconstructions with the simultaneous iterative reconstruction technique, showing the computational superiority of our approach.

References in corpus (16)

• Adam: A Method for Stochastic Optimization
• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• scikit-image: Image processing in Python
• Deep Convolutional Neural Network for Inverse Problems in Imaging
• Low Dose CT Image Denoising Using a Generative Adversarial Network with Wasserstein Distance and Perceptual Loss
• Improved Training of Wasserstein GANs
• 3D Convolutional Encoder-Decoder Network for Low-Dose CT via Transfer Learning from a 2D Trained Network
• Model based learning for accelerated, limited-view 3D photoacoustic tomography
• Loss Functions for Neural Networks for Image Processing
• A Deep Journey into Super-resolution: A survey
• Deep Bayesian Inversion
• Denoising Imaging Polarimetry by an Adapted BM3D Method
• X-ray tomography of extended objects: a comparison of data acquisition approaches
• Deep Learning Based Computed Tomography Whys and Wherefores
• Deep Learning Accelerated Light Source Experiments

Cited by in corpus (23)

• Generative Adversarial Network in Medical Imaging: A Review
• Quantum Machine Learning for Chemistry and Physics
• Unsupervised Deep Learning Methods for Biological Image Reconstruction and Enhancement
• Deep Encoder-decoder Adversarial Reconstruction (DEAR) Network for 3D CT from Few-view Data
• An Unsupervised Reconstruction Method For Low-Dose CT Using Deep Generative Regularization Prior
• Artifact Identification in X-ray Diffraction Data using Machine Learning Methods
• Distributed optimization for nonrigid nano-tomography
• Algorithm-driven Advances for Scientific CT Instruments: From Model-based to Deep Learning-based Approaches
• Physics-informed Score-based Diffusion Model for Limited-angle Reconstruction of Cardiac Computed Tomography
• Quantum State Tomography with Conditional Generative Adversarial Networks
• Physics-assisted Generative Adversarial Network for X-Ray Tomography
• BraggNN: Fast X-ray Bragg Peak Analysis Using Deep Learning
• Medical Image Generation using Generative Adversarial Networks
• Classification and reconstruction of optical quantum states with deep neural networks
• Neural Network Methods for Radiation Detectors and Imaging
• Smart-PGSim: Using Neural Network to Accelerate AC-OPF Power Grid Simulation
• HAGAN: Hybrid Augmented Generative Adversarial Network for Medical Image Synthesis
• Deep Learning Accelerated Light Source Experiments
• Deep learning based low-dose synchrotron radiation CT reconstruction
• Joint ptycho-tomography with deep generative priors
• TextureWGAN: Texture Preserving WGAN with MLE Regularizer for Inverse Problems
• Super time-resolved tomography
• StreamGuard: Low-Overhead Resilience for Real-time HPC Data Streams