arXiv subjects

Genoveva Burca

Publications and source records attributed to Genoveva Burca.

5 recordsLinked to original sources

arXiv2023

Shot Noise Reduction in Radiographic and Tomographic Multi-Channel Imaging with Self-Supervised Deep Learning

Noise is an important issue for radiographic and tomographic imaging techniques. It becomes particularly critical in applications where additional constraints force a strong reduction of the Signal-to-Noise Ratio (SNR) per image. These constraints may result from limitations on the maximum available flux or permissible dose and the associated restriction on exposure time. Often, a high SNR per image is traded for the ability to distribute a given total exposure capacity per pixel over multiple channels, thus obtaining additional information about the object by the same total exposure time. These can be energy channels in the case of spectroscopic imaging or time channels in the case of time-resolved imaging. In this paper, we report on a method for improving the quality of noisy multi-channel (time or energy-resolved) imaging datasets. The method relies on the recent Noise2Noise (N2N) self-supervised denoising approach that learns to predict a noise-free signal without access to noise-free data. N2N in turn requires drawing pairs of samples from a data distribution sharing identical signals while being exposed to different samples of random noise. The method is applicable if adjacent channels share enough information to provide images with similar enough information but independent noise. We demonstrate several representative case studies, namely spectroscopic (k-edge) X-ray tomography, in vivo X-ray cine-radiography, and energy-dispersive (Bragg edge) neutron tomography. In all cases, the N2N method shows dramatic improvement and outperforms conventional denoising methods. For such imaging techniques, the method can therefore significantly improve image quality, or maintain image quality with further reduced exposure time per image.

cs.CV↗

arXiv2022

Low-rank flat-field correction for artifact reduction in spectral computed tomography

Spectral computed tomography has received considerable interest in recent years since spectral measurements contain much richer information about the object of interest. In spectral computed tomography, we are interested in the energy channel-wise reconstructions of the object. However, such reconstructions suffer from low signal-to-noise ratio and share the challenges of conventional low-dose computed tomography such as ring artifacts. Ring artifacts arise from errors in the flat-field correction and can significantly degrade the quality of the reconstruction. We propose an extended flat-field model that exploits high correlation in the spectral flat-fields to reduce ring artifacts in the channel-wise reconstructions. The extended model relies on the assumption that the spectral flat-fields can be well-approximated by a low-rank matrix. Our proposed model works directly on the spectral flat-fields and can be combined with any existing reconstruction model, e.g., filtered back projection and iterative methods. The proposed model is validated on a neutron data set. The results show that our method successfully diminishes ring artifacts and improves the quality of the reconstructions. Moreover, the results indicate that our method is robust; it only needs a single spectral flat-field image, whereas existing methods need multiple spectral flat-field images to reach a similar level of ring reduction.

math.NA↗

arXiv2022

Recovering the second moment of the strain distribution from neutron Bragg edge data

Point by point strain scanning is often used to map the residual stress (strain) in engineering materials and components. However, the gauge volume and hence spatial resolution is limited by the beam defining apertures and can be anisotropic for very low and high diffraction (scattering) angles. Alternatively, wavelength resolved neutron transmission imaging has a potential to retrieve information tomographically about residual strain induced within materials through measurement in transmission of Bragg edges - crystallographic fingerprints whose locations and shapes depend on microstructure and strain distribution. In such a case the spatial resolution is determined by the geometrical blurring of the measurement setup and the detector point spread function. Mathematically, reconstruction of strain tensor field is described by the longitudinal ray transform; this transform has a non-trivial null-space, making direct inversion impossible. A combination of the longitudinal ray transform with physical constraints was used to reconstruct strain tensor fields in convex objects. To relax physical constraints and generalise reconstruction, a recently introduced concept of histogram tomography can be employed. Histogram tomography relies on our ability to resolve the distribution of strain in the beam direction, as we discuss in the paper. More specifically, Bragg edge strain tomography requires extraction of the second moment (variance about zero) of the strain distribution which has not yet been demonstrated in practice. In this paper we verify experimentally that the second moment can be reliably measured for a previously well characterised aluminium ring and plug sample. We compare experimental measurements against numerical calculation and further support our conclusions by rigorous uncertainty quantification of the estimated mean and variance of the strain distribution.

cond-mat.mtrl-sci↗

arXiv2021

Crystalline phase discriminating neutron tomography using advanced reconstruction methods

Time-of-flight neutron imaging offers complementary attenuation contrast to X-ray computed tomography (CT), coupled with the ability to extract additional information from the variation in attenuation as a function of neutron energy (time of flight) at every point (voxel) in the image. In particular Bragg edge positions provide crystallographic information and therefore enable the identification of crystalline phases directly. Here we demonstrate Bragg edge tomography with high spatial and spectral resolution. We propose a new iterative tomographic reconstruction method with a tailored regularisation term to achieve high quality reconstruction from low-count data, where conventional filtered back-projection (FBP) fails. The regularisation acts in a separated mode for spatial and spectral dimensions and favours characteristic piece-wise constant and piece-wise smooth behaviour in the respective dimensions. The proposed method is compared against FBP and a state-of-the-art regulariser for multi-channel tomography on a multi-material phantom. The proposed new regulariser which accommodates specific image properties outperforms both conventional and state-of-the-art methods and therefore facilitates Bragg edge fitting at the voxel level. The proposed method requires significantly shorter exposure to retrieve features of interest. This in turn facilitates more efficient usage of expensive neutron beamline time and enables the full utilisation of state-of-the-art high resolution detectors.

physics.med-ph↗

arXiv2021

Core Imaging Library -- Part I: a versatile Python framework for tomographic imaging

We present the Core Imaging Library (CIL), an open-source Python framework for tomographic imaging with particular emphasis on reconstruction of challenging datasets. Conventional filtered back-projection reconstruction tends to be insufficient for highly noisy, incomplete, non-standard or multi-channel data arising for example in dynamic, spectral and in situ tomography. CIL provides an extensive modular optimisation framework for prototyping reconstruction methods including sparsity and total variation regularisation, as well as tools for loading, preprocessing and visualising tomographic data. The capabilities of CIL are demonstrated on a synchrotron example dataset and three challenging cases spanning golden-ratio neutron tomography, cone-beam X-ray laminography and positron emission tomography.

math.OC↗