• Corpus ID: 232307516

Scatter Correction in X-ray CT by Physics-Inspired Deep Learning

@article{Iskender2021ScatterCI,
  title={Scatter Correction in X-ray CT by Physics-Inspired Deep Learning},
  author={Berk Iskender and Yoram Bresler},
  journal={ArXiv},
  year={2021},
  volume={abs/2103.11509}
}
Scatter due to interaction of photons with the imaged object is a fundamental problem in X-ray Computed Tomography (CT). It manifests as various artifacts in the reconstruction, making its abatement or correction critical for image quality. Despite success in specific settings, hardwarebased methods require modification in the hardware, or increase in the scan time or dose. This accounts for the great interest in software-based methods, including Monte-Carlo based scatter estimation, analytical… 

References

SHOWING 1-10 OF 42 REFERENCES
A Physics-Motivated DNN for X-Ray CT Scatter Correction
TLDR
A new physics-motivated deep-learning-based method to estimate scatter and correct for it in the acquired projection measurements, which incorporates both an initial reconstruction and the scatter-corrupted measurements.
A Deep Learning-Based Scatter Correction of Simulated X-ray Images
TLDR
A deep learning-based scatter correction method, which adopts a convolutional neural network (CNN) for restoration of degraded images, and shows better results in projected images.
Deep Scatter Estimation (DSE): Accurate Real-Time Scatter Estimation for X-Ray CT Using a Deep Convolutional Neural Network
TLDR
To correct for scatter in real-time, the deep scatter estimation (DSE) is developed which uses a deep convolutional neural network which is trained to reproduce the output of MC simulations using only the acquired projection data as input.
X-RAY CT SCATTER CORRECTION BY A PHYSICS-MOTIVATED DNN WITH OPPOSITE VIEW PROCESSING
Photon scattering in X-ray CT creates streaks, cupping, shading artifacts and decreased contrast in the reconstructions. We describe a novel physics-motivated deep-learning-based method using
Real‐time scatter estimation for medical CT using the deep scatter estimation: Method and robustness analysis with respect to different anatomies, dose levels, tube voltages, and data truncation
TLDR
The deep scatter estimation (DSE) is proposed: a deep convolutional neural network that is trained to predict MC scatter estimates based on the acquired projection data to derive highly accurate scatter estimates in real time.
Library based x-ray scatter correction for dedicated cone beam breast CT.
TLDR
The authors find that a scatter library on a simple breast model with only one input parameter, i.e., the breast diameter, sufficiently guarantees improvements in SNU and CDR and improves over the existing methods on implementation simplicity and computational efficiency.
Patient-specific scatter correction for flat-panel detector-based cone-beam CT imaging.
TLDR
A novel method has been proposed and validated to generate the needed coarse scatter profile for parameter calibration in the convolution method and can significantly reduce scatter artefacts and recover the correct CT number.
Scatter correction of cone-beam CT using a deep residual convolution neural network (DRCNN)
TLDR
The proposed DRCNN has a higher accuracy of scatter correction than the networks without the RLF or the dropout operations and is effective, efficient and robust as a solution to the CBCT scatter correction.
Hybrid scatter correction for CT imaging
TLDR
The results show that the hybrid scatter correction algorithm (HSC) can significantly reduce scatter artifacts and reduce the total runtime of the scatter correction by about two orders of magnitude for the cases investigated here when using the HSC instead of a low-noise Monte Carlo simulation for scatter correction.
Scatter Correction for Industrial Cone-Beam Computed Tomography (CBCT) Using 3D VSHARP, a fast GPU-Based Linear Boltzmann Transport Equation Solver
The use of cone-beam computed tomography (CBCT) to inspect objects containing aluminum, Inconel and other metals is a challenge due to high amounts of scattered radiation in the projection data.
...
...