Abstract : Mathematical imaging and modeling are two key challenges in the imaging modality magnetic particle imaging (MPI). MPI provides reconstructions of the concentration of magnetic nanoparticles in 4D. To address this inverse problem properly, various dynamics need to be taken into account, e.g., the particles’ magnetization behavior and the dynamics in the fluid tracer. MPI provides challenging problems in imaging, modeling, and parameter identification. In this mini-symposium, we aim at bringing together researchers working on MPI and related mathematical fields. We cover theoretical and practical topics focusing on mathematical and physical as well as algorithmic and computational issues.
[03274] The image reconstruction problem in magnetic particle imaging and an application of the deep image prior
Format : Talk at Waseda University
Author(s) :
Tobias Kluth (University of Bremen)
Abstract : Magnetic particle imaging (MPI) is a tracer-based imaging modality detecting the concentration of superparamagnetic iron oxide nanoparticles. The imaging problem is a linear inverse problem given by a Fredholm integral equation of the first kind describing the concentration-to-voltage mapping. The talk provides a general introduction to MPI and the imaging problem. We further investigate general deep image prior concepts for inverse problems and their application to image reconstruction in MPI.
[05253] A hybrid model for image reconstruction in MPI using a FFL
Format : Online Talk on Zoom
Author(s) :
Jürgen Frikel (OTH Regensburg)
Abstract : In Magnetic Particle Imaging (MPI), most model-based reconstruction methods rely on idealized assumptions, such as an ideal field-free line (FFL) topology. However, real MPI scanners generate magnetic fields with distortions that often lead to inaccurate reconstructions and artifacts. To improve the reconstruction quality in MPI, it is essential to develop more realistic models. In this talk, we present a hybrid MPI model that can integrate real measurements of the applied magnetic fields into a physical model. Based on this model, we introduce a novel calibration procedure that allows the acquisition of the required magnetic field measurements, independent of the resolution, and with significantly less time consumption than a measurement-based model. In addition, we present a discretization strategy for the model that can be used in algebraic reconstructions.
[04125] MPI using an FFL-scanner: Radon-based image reconstruction for realistic setup assumptions
Format : Talk at Waseda University
Author(s) :
Stephanie Blanke (Universität Hamburg)
Christina Brandt (Universität Hamburg)
Abstract : Magnetic particle imaging is a tracer-based imaging modality exploiting the nonlinear magnetization response of magnetic particles to changing magnetic fields. We adapt forward model and reconstruction methods towards more realistic setup assumptions for the specific choice of using a field-free line scanner. In this case, the scanning geometry resembles those of computerized tomography and we are able to jointly reconstruct particle concentration and corresponding Radon data by means of total variation regularization.
[04651] Parameter estimation for modeling of nanoparticle dynamics
Format : Online Talk on Zoom
Author(s) :
Hannes Albers (Universität Bremen)
Tobias Kluth (University of Bremen)
Abstract : In order to overcome the limitations of needing full delta probe calibrations for MPI, accurate and fast model-based image reconstruction with as few as possible calibration measurements are highly desirable. We discuss methods for estimating particle parameters from calibration measurements and subsequently applying them to dynamic particle models, such as the Néel relaxation model, to obtain modeled system matrices of high quality for reconstruction.
[05030] Implicit neural representations for super-resolution in magnetic particle imaging
Format : Talk at Waseda University
Author(s) :
Franziska Schrank (RWTH Aachen University)
Volkmar Schulz (RWTH Aachen University)
Abstract : Magnetic particle imaging is a medical imaging technology based on the non-linear magnetization of magnetic nanoparticles. For image reconstruction, the received signal from the excitation of the nanoparticles is converted into the particles’ concentration distribution via the system matrix, which is commonly measured in a calibration scan. We propose to parametrize this system matrix using implicit neural representations, enabling to super-resolve it or to reduce the matrix’ acquisition time by processing an undersampled matrix.
[05026] Reducing displacement artifacts in multi-patch magnetic particle imaging
Format : Talk at Waseda University
Author(s) :
Marija Boberg (University Medical Center Hamburg-Eppendorf)
Tobias Knopp (University Medical Center Hamburg-Eppendorf)
Martin Möddel (University Medical Center Hamburg-Eppendorf)
Abstract : Magnetic particle imaging determines the spatial distribution of superparamagnetic nanoparticles within a small field-of-view. Multi-patch approaches can expand the field-of-view at the cost of artifacts caused by field imperfections. Time-consuming calibration scans can reduce these displacement artifacts by measuring system matrices for each patch. In this contribution, only one central system matrix is used, which is warped according to the underlying magnetic fields, resulting in low calibration times and higher image quality.
[05158] Deconvolution of direct Chebyshev reconstructions in MPI with neural networks
Format : Online Talk on Zoom
Author(s) :
Mathias Eulers (Universität zu Lübeck)
Marco Maass (Universität zu Lübeck)
Christine Droigk (Universität zu Lübeck)
Alfred Mertins (Universität zu Lübeck)
Abstract : Recently, a direct reconstruction method using Chebyshev polynomials for multi-dimensional MPI has been proposed. The reconstruction method weights and sums the frequency components of the voltage signals with tensor products Chebyshev polynomials, followed by a deconvolution step to perform a very fast image reconstruction. Unfortunately, the method is degraded by image artifacts. In this presentation, the method itself will be explained and a data-driven deconvolution model is presented which improves the image quality.
[04792] A Flexible Approach to Model-Based Reconstruction in Magnetic Particle Imaging
Format : Talk at Waseda University
Author(s) :
Thomas März (Hochschule Darmstadt)
Abstract : In Magnetic Particle Imaging (MPI) images are usually reconstructed using a system matrix obtained via
a time-consuming calibration procedure.
Our approach employs a mathematical model based on the MPI signal encoding and its analytical properties.
We present our two-stage algorithm:
First stage: we estimate components of the MPI Core Operator by using a variational formulation.
Second stage: the image is reconstructed by regularized deconvolution while fitting
the results of the first stage.
We demonstrate the performance of our algorithm with simulated data.
[04806] Reconstruction of Dynamic Concentrations with Sequential Subspace Optimization
Format : Talk at Waseda University
Author(s) :
Marius Nitzsche (University of Stuttgart)
Abstract : Magnetic particle imaging faces challenges in dealing with dynamics, which often results in motion artifacts and lower quality reconstructions due to limited averaging. Standard algorithms cannot produce high-quality images under these circumstances. To address these issues, we utilize the Regularized Sequential Subspace Optimization (Resesop) algorithm, which can account for model imperfections caused by motion without requiring strong prior information. We demonstrate the effectiveness of Resesop on both simulated and real dynamic MPI data.
[04730] Joint motion estimation and image reconstruction for dynamic MPI
Format : Talk at Waseda University
Author(s) :
Christina Brandt (Universität Hamburg)
Lena Zdun (Universität Hamburg)
Abstract : Potential applications of MPI include highly dynamic tasks as blood flow imaging and instrument tracking during
inteventions. In this talk, we propose to tackle the additional challenges caused by the dynamics by a joint image
reconstruction and motion estimation approach. We combine a multi-scale motion estimation algorithm with a
stochastic primal-dual algorithm for image reconstruction. Convincing numerical results are achieved on in-vitro and
in-vivo data using motion models matching the specific application.
