Dynamic CT Reconstruction from Limited Views with Implicit Neural Representations and Parametric Motion Fields

Reed, Albert; Kim, Hyojin; Mohan, K. Aditya; Champley, Kyle; Kang, Jingu; Jayasuriya, Suren

doi:10.1109/iccv48922.2021.00226

Cited by 29 publications

(26 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…INRs, sometimes called coordinate-based networks, are a class of neural networks constructed from fully-connected neuron layers that are tasked with learning a function that maps input coordinates to physical properties in the scene (e.g., the acoustic reflectivity at (x, y)). INRs have also recently been applied to tomographic imaging problems such as computed tomography [57], which inspired our motivation for applying them to CSAS imaging. Key to their success is a positional encoding layer [58] that maps input coordinates to a high dimensional space using a Fourier basis and enables regularizing the spatial frequencies reconstructed by the network.…”

Section: Implicit Neural Representationsmentioning

confidence: 99%

SINR: Deconvolving Circular SAS Images Using Implicit Neural Representations

Reed¹,

Blanford²,

Brown³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Circular Synthetic aperture sonars (CSAS) capture multiple observations of a scene to reconstruct high-resolution images. We can characterize resolution by modeling CSAS imaging as the convolution between a scene's underlying point scattering distribution and a system-dependent point spread function (PSF). The PSF is a function of the transmitted waveform's bandwidth and determines a fixed degree of blurring on reconstructed imagery. In theory, deconvolution overcomes bandwidth limitations by reversing the PSF-induced blur and recovering the scene's scattering distribution. However, deconvolution is an illposed inverse problem and sensitive to noise. We propose a selfsupervised pipeline (does not require training data) that leverages an implicit neural representation (INR) for deconvolving CSAS images. We highlight the performance of our SAS INR pipeline, which we call SINR, by implementing and comparing to existing deconvolution methods. Additionally, prior SAS deconvolution methods assume a spatially-invariant PSF, which we demonstrate yields subpar performance in practice. We provide theory and methods to account for a spatially-varying CSAS PSF, and demonstrate that doing so enables SINR to achieve superior deconvolution performance on simulated and real acoustic SAS data. We provide code 1 to encourage reproducibility of research.

show abstract

Section: Implicit Neural Representationsmentioning

confidence: 99%

SINR: Deconvolving Circular SAS Images Using Implicit Neural Representations

Reed¹,

Blanford²,

Brown³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…We benefit from developments in the related applications of neural scene representation [8,9] and, more broadly, implicit neural representation [10,11]. Implicit neural representations have seen some exploration for scientific imaging, including for CT reconstruction [12,13] and for synthetic aperture sonar reconstruction [14].…”

Section: Simulation and Reconstructionmentioning

confidence: 99%

Novel Light Field Imaging Device with Enhanced Light Collection for Cold Atom Clouds

Schwartzman

2022

Preprint

View full text Add to dashboard Cite

We present a light field imaging system that captures multiple views of an object with a single shot. The system is designed to maximize the total light collection by accepting a larger solid angle of light than a conventional lens with equivalent depth of field. This is achieved by populating a plane of virtual objects using mirrors and fully utilizing the available field of view and depth of field. Simulation results demonstrate that this design is capable of single-shot tomography of objects of size O(1 mm 3 ), reconstructing the 3-dimensional (3D) distribution and features not accessible from any single view angle in isolation. In particular, for atom clouds used in atom interferometry experiments, the system can reconstruct 3D fringe patterns with size O(100 µm). We also demonstrate this system with a 3D-printed prototype. The prototype is used to take images of O(1 mm 3 ) sized objects, and 3D reconstruction algorithms running on a single-shot image successfully reconstruct O(100 µm) internal features. The prototype also shows that the system can be built with 3D printing technology and hence can be deployed quickly and cost-effectively in experiments with needs for enhanced light collection or 3D reconstruction. Imaging of cold atom clouds in atom interferometry is a key application of this new type of imaging device where enhanced light collection, high depth of field, and 3D tomographic reconstruction can provide new handles to characterize the atom clouds.

show abstract

“…and Mescheder et al., 42,43 NIRs have been used for novel view synthesis 13,44–46 and multi‐view reconstruction 47–49 . Recently, NIRs have been used to improve CT 50–55 and MR 56 imaging. The use of NIRs to help solve inverse problem of CT reconstruction was first shown by Sun 50 .…”

Section: Introductionmentioning

confidence: 99%

“…In contrast, the focus of this paper is to reconstruct objects that move during data acquisition. NIRs have been extended to dynamic scenes by warping a learned template scene with estimated motion field 55 . Due to the reliance on a template, this method fails to handle complex topology changes.…”

Section: Introductionmentioning

confidence: 99%

DiFiR‐CT: Distance field representation to resolve motion artifacts in computed tomography

et al. 2023

View full text Add to dashboard Cite

Background: Motion during data acquisition leads to artifacts in computed tomography (CT) reconstructions. In cases such as cardiac imaging, not only is motion unavoidable, but evaluating the motion of the object is of clinical interest. Reducing motion artifacts has typically been achieved by developing systems with faster gantry rotation or via algorithms which measure and/or estimate the displacement. However, these approaches have had limited success due to both physical constraints as well as the challenge of estimating non-rigid, temporally varying, and patient-specific motion fields. Purpose: To develop a novel reconstruction method which generates timeresolved, artifact-free images without estimation or explicit modeling of the motion. Methods:We describe an analysis-by-synthesis approach which progressively regresses a solution consistent with the acquired sinogram. In our method, we focus on the movement of object boundaries. Not only are the boundaries the source of image artifacts, but object boundaries can simultaneously be used to represent both the object as well as its motion over time without the need for an explicit motion model. We represent the object boundaries via a signed distance function (SDF) which can be efficiently modeled using neural networks. As a result, optimization can be performed under spatial and temporal smoothness constraints without the need for explicit motion estimation. Results: We illustrate the utility of DiFiR-CT in three imaging scenarios with increasing motion complexity: translation of a small circle, heart-like change in an ellipse's diameter, and a complex topological deformation. Compared to filtered backprojection, DiFiR-CT provides high quality image reconstruction for all three motions without hyperparameter tuning or change to the architecture. We also evaluate DiFiR-CT's robustness to noise in the acquired sinogram and found its reconstruction to be accurate across a wide range of noise levels. Lastly, we demonstrate how the approach could be used for multi-intensity scenes and illustrate the importance of the initial segmentation providing a realistic initialization. Code and supplemental movies are available at https://kunalmgupta.github.io/projects/DiFiR-CT.html. Conclusions: Projection data can be used to accurately estimate a temporallyevolving scene without the need for explicit motion estimation using a neural implicit representation and analysis-by-synthesis approach.

show abstract

Dynamic CT Reconstruction from Limited Views with Implicit Neural Representations and Parametric Motion Fields

Cited by 29 publications

References 38 publications

SINR: Deconvolving Circular SAS Images Using Implicit Neural Representations

SINR: Deconvolving Circular SAS Images Using Implicit Neural Representations

Novel Light Field Imaging Device with Enhanced Light Collection for Cold Atom Clouds

DiFiR‐CT: Distance field representation to resolve motion artifacts in computed tomography

Contact Info

Product

Resources

About