Tim C. Kranemann

Bio: Tim C. Kranemann is an academic researcher from Ruhr University Bochum. The author has contributed to research in topics: Magnetic field & Estimator. The author has an hindex of 5, co-authored 16 publications receiving 52 citations.

On the Performance of Time Domain Displacement Estimators for Magnetomotive Ultrasound Imaging

Abstract: In magnetomotive (MM) ultrasound (US) imaging, magnetic nanoparticles (NPs) are excited by an external magnetic field and the tracked motion of the surrounding tissue serves as a surrogate parameter for the NP concentration. MMUS procedures exhibit weak displacement contrasts due to small forces on the NPs. Consequently, precise US-based displacement estimation is crucial in terms of a sufficiently high contrast-to-noise ratio (CNR) in MMUS imaging. Conventional MMUS detection of the NPs is based on samplewise evaluation of the phase of the in-phase and quadrature (IQ) data, where a low signal-to-noise ratio (SNR) in the data leads to strong phase noise and, thus, to an increased variance of the displacement estimate. This paper examines the performance of two time-domain displacement estimators (DEs) for MMUS imaging, which differ from conventional MMUS techniques by incorporating data from an axial segment. The normalized cross correlation (NCC) estimator and a recursive Bayesian estimator, incorporating spatial information from neighboring segments, weighted by the local SNR, are adapted for the small MMUS displacement magnitudes. Numerical simulations of MM-induced, time-harmonic bulk and Gaussian-shaped displacement profiles show that the two time-domain estimators yield a reduced estimation error compared to the phase-shift-based estimator. Phantom experiments, using our recently proposed magnetic excitation setup, show a 1.9-fold and 3.4-fold increase of the CNR in the MMUS images for the NCC and Bayes estimator compared to the conventional method, while the amount of required data is reduced by a factor of 100.

Modeling and Measurement of the Nonlinear Force on Nanoparticles in Magnetomotive Techniques

Abstract: Magnetomotive (MM) ultrasound (US) imaging is the identification of tissue in which magnetic nanoparticle tracers are present by detecting a magnetically induced motion. Although the nanoparticles have a magnetization that depends nonlinearly on the external magnetic field, this has often been neglected, and the presence of resulting higher harmonics in the detected motion has not been reported yet. Here, the magnetization of nanoparticles in gelatin was modeled according to the Langevin theory of superparamagnetism. This nonlinear relationship has a fundamental effect on the resulting force and motion. However, the magnetic field must contain regions with a strong magnetic gradient and a low absolute magnetic field to allow the significant generation of higher harmonics in the force. To validate the model, an MM setup that has a constant magnetic gradient on one axis superimposed by a homogeneous time-varying magnetic field was used. After the magnetic characterization of the nanoparticles and calculations of the expected displacement in the setup, experiments were conducted. A laser Doppler vibrometer was used to quantify the small displacements at higher harmonics. The experimental results followed theoretical predictions. Deviations between model and experiment were attributed to a simplified mechanical modeling and temperature rise during measurements. It is concluded that in MM techniques, the nonlinear magnetization of nanoparticles must generally be considered to reconstruct quantitative parameters, to achieve optimum matching of fields and particles, or to exploit nanoparticle magnetization for tissue characterization. In addition, with the presented experimental setup, the magnetization properties of nanoparticles can be determined by MM techniques alone.

Frequency response of soft tissue displacements induced by the force on magnetic nanoparticles

Abstract: In Magnetomotive Ultrasound (MMUS), tissue embedded magnetic nanoparticles (NP) are mechanically excited by a magnetic field and the resulting motion of the surrounding tissue is tracked to image the NP distribution. This mechanism of excitation is of particular interest for elastography applications and offers an alternative to the acoustic radiation force excitation in order to estimate the tissue's mechanical properties. Measuring the frequency-dependent, tissue properties is relevant for several elastography techniques but common MMUS setups utilize solenoid/yoke configurations with excitation frequencies limited up to a few Hz. In this work, we present an MMUS setup which allows to measure the entire frequency range, relevant in elastography, by decoupling the gradient field from the excitation field. A cylindrical inclusion containing 90% nanoparticles and 10% gelatin was embedded in phantoms of different gelatin concentrations (10%, 15% w/w). A static, magnetic field was generated by two permanent magnets arranged opposite each other, producing a constant gradient of 3.4 T/m along the displacement direction. A solenoid generated a superimposed, sinusoidal excitation field of 18 mT at frequencies between 50–1000 Hz. For validation, the on-axis displacement of the inclusion was tracked by a laser vibrometer at its surface due to reflections at a thin slice of glass powder. The frequency responses of the displacement measurements showed resonance frequencies of 310 Hz for the soft (10%) and 500 Hz for the stiff (15%) phantom, respectively. The proposed setup proved capable of tracking the mechanical tissue dynamics precisely beyond the frequency limit of standard MMUS setups.

Real-Time Magnetomotive Ultrasound Imaging Using a Recursive Estimator

Abstract: In magnetomotive ultrasound, real-time capability is useful for the alignment of magnetic coils and ultrasonic transducers and it is considered as necessary for clinical applications. Since usually a harmonic motion is induced, the magnitude of the displacement is often estimated via Fourier transform of the phase of IQ-data. However, a standard block-by-block processing of IQ-data data can limit estimator performance. Here, hardware-and estimator-parameters are decoupled by implementing a recursive algorithm. Therefore, a sliding windowed infinite Fourier transform was used. The different characteristics of the block-by-block and the sliding processing are demonstrated. A higher performance of the recursive algorithm was shown with simulated data at low signal-to-noise ratios. The implementation on an ultrasound research platform confirmed real-time capability. The presented algorithm is expected to be easily implemented on any ultrasound system. Future work needs to address the optimal parametrization in real-world scenarios.

Towards the Integration of an MPI Compatible Ultrasound Transducer

Abstract: Magnetic particle imaging (MPI) is a tracer based modality and thus lacks anatomical information. Medical ultrasound (US) imaging provides the desired morphological data in real-time that could complement functional MPI images. However, most customary US devices will certainly be damaged by the strong alternating magnetic fields inside the MPI bore. Moreover, US equipment might degrade MPI signal quality. In this work, US components that are prone to eddy current heating were pointed out by presenting the major components of a medical US transducer. A theoretical model was utilized to derive maximum applicable sizes for conducting parts like electrode surfaces and cables. Heating experiments inside the MPI bore showed that heating can be managed by dispensing extensive electric shields and keeping conducting structures reasonably small. Further, transducer dummies that were placed inside the MPI scanner bore and actively driven with US signals were used to assess the interferences between both modalities. The MPI signals showed a minor increase in the noise level when transducer dummies were present, while serious interferences were recognized when a dummy comprising a tuning inductor was electrically driven. The US signals showed strong disturbances in the frequency range of the MPI drive fields during MPI acquisition, which was decades lower than the relevant US frequency range. It is concluded that the combination of an MPI scanner and adapted US hardware is feasible. Future work needs to address the suppression of interferences and their impact on image quality of both modalities.

Medical Imaging of Microrobots: Toward In Vivo Applications.

Abstract: Medical microrobots (MRs) have been demonstrated for a variety of non-invasive biomedical applications, such as tissue engineering, drug delivery, and assisted fertilization, among others. However, most of these demonstrations have been carried out in in vitro settings and under optical microscopy, being significantly different from the clinical practice. Thus, medical imaging techniques are required for localizing and tracking such tiny therapeutic machines when used in medical-relevant applications. This review aims at analyzing the state of the art of microrobots imaging by critically discussing the potentialities and limitations of the techniques employed in this field. Moreover, the physics and the working principle behind each analyzed imaging strategy, the spatiotemporal resolution, and the penetration depth are thoroughly discussed. The paper deals with the suitability of each imaging technique for tracking single or swarms of MRs and discusses the scenarios where contrast or imaging agent's inclusion is required, either to absorb, emit, or reflect a determined physical signal detected by an external system. Finally, the review highlights the existing challenges and perspective solutions which could be promising for future in vivo applications.

Displacement Estimation in Ultrasound Elastography Using Pyramidal Convolutional Neural Network

Abstract: In this article, two novel deep learning methods are proposed for displacement estimation in ultrasound elastography (USE). Although convolutional neural networks (CNNs) have been very successful for displacement estimation in computer vision, they have been rarely used for USE. One of the main limitations is that the radio frequency (RF) ultrasound data, which is crucial for precise displacement estimation, has vastly different frequency characteristics compared with images in computer vision. Top-rank CNN methods used in computer vision applications are mostly based on a multilevel strategy, which estimates finer resolution based on coarser ones. This strategy does not work well for RF data due to its large high-frequency content. To mitigate the problem, we propose modified pyramid warping and cost volume network (MPWC-Net) and RFMPWC-Net, both based on PWC-Net, to exploit information in RF data by employing two different strategies. We obtained promising results using networks trained only on computer vision images. In the next step, we constructed a large ultrasound simulation database and proposed a new loss function to fine-tune the network to improve its performance. The proposed networks and well-known optical flow networks as well as state-of-the-art elastography methods are evaluated using simulation, phantom, and in vivo data. Our two proposed networks substantially outperform current deep learning methods in terms of contrast-to-noise ratio (CNR) and strain ratio (SR). Also, the proposed methods perform similar to the state-of-the-art elastography methods in terms of CNR and have better SR by substantially reducing the underestimation bias.

Magnetic particle imaging in vascular medicine.

Abstract: Magnetic particle imaging (MPI) is a new medical imaging technique that enables three-dimensional real-time imaging of a magnetic tracer material. Although it is not yet in clinical use, it is highly promising, especially for vascular and interventional imaging. The advantages of MPI are that no ionizing radiation is necessary, its high sensitivity enables the detection of very small amounts of the tracer material, and its high temporal resolution enables real-time imaging, which makes MPI suitable as an interventional imaging technique. As MPI is a tracer-based imaging technique, functional imaging is possible by attaching specific molecules to the tracer material. In the first part of this article, the basic principle of MPI will be explained and a short overview of the principles of the generation and spatial encoding of the tracer signal will be given. After this, the used tracer materials as well as their behavior in MPI will be introduced. A subsequent presentation of selected scanner topologies will show the current state of research and the limitations researchers are facing on the way from preclinical toward human-sized scanners. Furthermore, it will be briefly shown how to reconstruct an image from the tracer materials' signal. In the last part, a variety of possible future clinical applications will be presented with an emphasis on vascular imaging, such as the use of MPI during cardiovascular interventions by visualizing the instruments. Investigations will be discussed, which show the feasibility to quantify the degree of stenosis and diagnose strokes and traumatic brain injuries as well as cerebral or gastrointestinal bleeding with MPI. As MPI is not only suitable for vascular medicine but also offers a broad range of other possible applications, a selection of those will be briefly presented at the end of the article.

Hybrid MPI-MRI System for Dual-Modal In Situ Cardiovascular Assessments of Real-Time 3D Blood Flow Quantification—A Pre-Clinical In Vivo Feasibility Investigation

Abstract: Non-invasive quantification of functional parameters of the cardiovascular system, in particular the heart, remains very challenging with current imaging techniques. This aspect is mainly due to the fact, that the spatio-temporal resolution of current imaging methods, such as Magnetic Resonance Imaging (MRI) or Positron Emission Tomography (PET), does not offer the desired data repetition rates in the context of real-time data acquisition and thus, can cause artifacts and misinterpretations in accelerated data acquisition approaches. We present a fast non-invasive and quantitative dual-modal in situ cardiovascular assessment using a hybrid imaging system which combines the new imaging modality Magnetic Particle Imaging (MPI) and MRI. This pre-clinical hybrid imaging system provides either a 0.5 T homogeneous ${B}_{{0}}$ field for MRI or a 2.2 T/m gradient field featuring a Field-Free-Point for MPI. A comprehensive coil system allows in both imaging modes for spatial encoding, signal excitation and reception. In this work, 3-dimensional anatomical information acquired with MRI is combined with in situ sequentially acquired time-resolved 3D (i.e. 3D + t) MPI bolus tracking of superparamagnetic iron oxide nanoparticles. MPI data were acquired during a 21 $\mu \text{l}$ (40 $\mu $ mol(Fe)/kgBW) bolus tail vein injection under free-breathing with an ungated and non-triggered MPI scan with a repetition rate of 46 volumes per seconds. We successfully determined quantitative hemodynamics as 3D + t velocity vector estimations of a beating rat’s heart by analyzing 3 seconds of 3D + t MPI image data. The used hybrid system allows for MR-based MPI Field-of-View planning and cardiac cross-sectional anatomy analysis, precise co-registration of dual-modal datasets, as well as for MPI-based hemodynamic functional analysis using an optical flow technique. We present the first in-vivo results of a new methodology, allowing for fast, non-invasive, quantitative and in situ hybrid cardiovascular assessment, showing its potential for future clinical applications.

Magnetomotive Ultrasound Imaging Systems: Basic Principles and First Applications

Abstract: This review discusses magnetomotive ultrasound, which is an emerging technique that uses superparamagnetic iron oxide nanoparticles as a contrast agent. The key advantage of using nanoparticle-based contrast agents is their ability to reach extravascular targets, whereas commercial contrast agents for ultrasound comprise microbubbles confined to the blood stream. This also extends possibilities for molecular imaging, where the contrast agent is labeled with specific targeting molecules (e.g., antibodies) so that pathologic tissue may be visualized directly. The principle of action is that an external time-varying magnetic field acts to displace the nanoparticles lodged in tissue and thereby their immediate surrounding. This movement is then detected with ultrasound using frequency- or time-domain analysis of echo data. As a contrast agent already approved for magnetic resonance imaging (MRI) by the US Food and Drug Administration, there is a shorter path to clinical translation, although safety studies of magnetomotion are necessary, especially if particle design is altered to affect biodistribution or signal strength. The external modulated magnetic field may be generated by electromagnets, permanent magnets, or a combination of the two. The induced nanoparticle motion may also reveal mechanical material properties of tissue, healthy or diseased, one of several interesting potential future aspects of the technique.