International Journal on Magnetic Particle Imaging (IJMPI)
Not a member yet
555 research outputs found
Sort by
Evaluating Magnetic Staging Following Sentinel Lymph Node Biopsy: The LowMag Trial
No full textBackground: Accurate identification of metastatic lymph nodes (LNs) is essential for cancer prognosis and treatment planning. In the LowMag clinical trial [1], a magnetic sentinel lymph node biopsy (SLNB) procedure was introduced for patients with invasive breast cancer. Difference between metastatic and non-metastatic LNs was assessed using amount of iron captured in individual LNs and their respective AC susceptibility (ACS).
Methods: Two magnetic devices—the superparamagnetic quantifier (SPaQ) [1,2] and the differential magnetometer handheld (DMH) probe [1,3] —were used to measure iron content and ACS in individual LNs. Additionally, ex vivo LN imaging was performed using a low-field MRI system [1,4], followed by detailed histopathological analysis. A total of 33 LNs from 18 consecutive patients, including four metastatic nodes, were examined.
Results: The low-field MRI findings and histopathological analysis of iron levels were consistent with measurements obtained from SPaQ and DMH probe. A significant difference in iron content was observed between metastatic LNs (DMH: 203.12 ± 87.67 µg; SPaQ: 131.28 ± 53.38 µg) and non-metastatic LNs (DMH: 92.47 ± 89.8 µg; SPaQ: 42.45 ± 46.9 µg). Furthermore, combining two features extracted from the ACS curve—Full Width at Half Maximum (FWHM) and peak values—resulted in a sensitivity of 75.0%, a specificity of 96.6%, and an overall accuracy of 93.9%.
Conclusion: This study suggests that the DiffMag method may offer precise detection of non-metastatic LNs. Additionally, SPaQ in both DiffMag and ACS mode, exhibits higher specificity and greater accuracy than the DMH probe due to its unique design
Single-Sided MPI without Selection Fields: Proof-of-Concept Results
No full textIn a standard single-sided magnetic nanoparticles (MNPs) systems, all coils including the selection field coils are positioned on one side of the object to be imaged. In this work, we propose a single-sided MPI system that combines a single-sided magnetic particle spectrometer (MPS) setup and linear movement to resolve the magnetic nanoparticle (MNP) distribution along the axis of the system, without the need for selection fields
SMART RHESINs for magnetic particle imaging: impact of viscosity-independent relaxation on image reconstruction
No full textMagnetic particle imaging (MPI) is a promising medical imaging modality that leverages the magnetic properties of nanoparticles. Traditionally, MPI tracers have been limited to commercially available nanoparticles, but specialized tracers could improve signal detection and expand applications. Here, we introduce SMART RHESINs, an innovative tracer design for MPI. In SMART RHESINs, synomag®-D nanoparticles are encapsulated in hollow nanospheres, shielding them from external influences like viscosity changes and enabling signal quantification independent of the surrounding medium. We demonstrated that the phase angle obtained through MPS is a rapid, predictive metric for assessing tracer suitability for MPI applications. Unlike the system matrix from non-encapsulated synomag®-D, the system matrix from aqueous synomag®-D-encapsulated SMART RHESINs allowed reconstructing immobilized SMART RHESINs, underscoring design robustness. SMART RHESINs hold potential for quantitative measurements across diverse environments, broadening the scope of MPS and MPI as a versatile tracer platform for quantitative imaging
Optimized Single-Channel Head Coil for Maximizing Drive Field Amplitude Within Safety Limits
No full textThis study investigates the design of an optimized single-channel head coil to maximize drive field (DF) amplitudes while minimizing the risk of peripheral nerve stimulation (PNS) in the human head. Using an optimization algorithm, we select the optimal winding positions on a fixed-length coil and achieve up to 25% increase in DF amplitude for superior portions of the head. Future work will analyze coil parameters and the choice of optimization constraints to increase DF amplitudes within the safety limits, across the entire human head
Phase-sensitive signal processing in DiffMag handheld probe
No full textSentinel Lymph Node Biopsy (SLNB) is a surgical procedure that employs a tracer and a handheld detection device. Superparamagnetic iron oxide nanoparticles (SPIONs) show promise as tracers, while handheld magnetometers are utilized for detection. However, conventional magnetometers face interference from tissue and surgical instruments, complicating procedures. The DiffMag processing technique, assessing solely SPIONs, is prototyped as the DiffMag handheld (DMH) probe. However, movement of the DMH-probe near surgical instrumentation may generate inconvenient signal artefacts. This study provides an extension of the DiffMag sensing principle with phase-sensitive signal processing. In this study we evaluated the efficacy of phase-sensitive detection through assessment of system phase shift factors originating from the introduced phase from electrical components and we examined SPION phase lag within clinically relevant parameter ranges, such as low SPION concentrations and increased environmental viscosity for assessment of technique limitations. The DMH-probe was used for acquisition of phase data of various SPIONs (Magtrace®, Resotran®, Resovist®, Ferrotrace®) and compared to a baseline acquisition in the absence of SPIONs. System phase shift was observed to be a wide-sense stationarity signal with consistent values for consecutive days, highlighting the stability of the system. Probe heating is observed results in a relatively small non-linear increase in system phase shift. All SPIONs display a distinct phase difference from the baseline acquisition in both low concentrations, and relatively high environmental viscosities. In conclusion, the phase-sensitive signal processing utilised in the DMH-probe demonstrates strong potential for identifying signal artefacts
Denoising the system matrix with deep neural networks for better MPI reconstructions
No full textMagnetic Particle Imaging commonly relies on the system matrix (SM) to reconstruct particle distributions, but noise during acquisition limits both its resolution and image quality. Traditionally, noise reduction requires averaging multiple measurements, which increases acquisition time. This paper presents a deep neural network trained on simulated SMs and measured background noise, which effectively generalizes to real-world data. The model recovers higher frequency components of the SM and serves as a general pre-processing step, enhancing image reconstruction quality while reducing the need for extensive averaging, thus accelerating SM acquisition
Data rebinning based decoupling of pFOV overlap scanning in magnetic particle imaging
No full textMagnetic particle imaging(MPI) is rapidly developing as a novel biomedical imaging technique. Due to the safety constraints of bioelectromagnetic stimulation, the partial Field of View (pFOV) scanning approach has been widely used in MPI devices towards small animals and human scales. pFOV scanning has variable degrees of overlap between neighbouring pixels, which may lead to signal aliasing and blurring on harmonic imaging by Fourier transformation. In this paper we propose a data rebinning method for decoupling overlap scanning under a continuous pFOV scanning process. The results show that data rebinning as a digital signal preprocessing technique has the potential to improve the spatial resolution of imaging
30-fold acceleration using sinogram-based system calibration for field free line MPI
No full textWe propose a field-based calibration directly in the sinogram domain for a 3D field free line (FFL) dynamic sinusoidal trajectory. This approach involves performing calibration measurements with alternating magnetic field offsets for each offset and z-position in sinogram domain (covering a total of Nxy × Nz positions on a 2D grid), rather than calibration measurements for each MPI image voxel in 3D space (e.g., for Nxy × Nxy × Nz voxels). Using the sinogram-based calibration data, we synthesize the 3D system matrix (SM) in the image domain by exploiting the shift and rotation invariance of MPI systems. The proposed 2D sinogram-based calibration method for the 3D FFL trajectory achieved a 30-fold reduction in system calibration time due to reduced dimensionality, with less than 0.5% normalized RMS-error compared to 3D image-based calibration
Software Implementation of Greedy Optimized MPS Sequences
No full textMulti-dimensional magnetic particle spectrometers provide fast and high quality hybrid system matrix measurements, using different DC offsets to emulate the same saturation that magnetic nanoparticles are exposed to in a real scanner. The electrical implementation of these offsets poses signal generation challenges that impact runtime and power dissipation. In this work, we address the challenges of minimizing sign changes of DC sources with H-bridges and efficient offset sorting for arbitrary numbers of offset channels
MPI-based estimation of absolute temperature using 1D system matrices
No full textMagnetic Particle Imaging (MPI) thermometry enables temperature mapping by exploiting temperature-dependent response of magnetic nanoparticles. These changes, influenced by particle physics (thermal agitation, relaxation dynamics), allow MPI to provide time-resolved temperature monitoring, making it a promising feed-back parameter for applications such as hyperthermia. However, current system matrix-based or ‘multi-color’ methods require time-consuming, three-dimensional calibration measurements that limit practical application and primarily yield relative temperature values only. Here, we present a novel system matrix thermometry method to enhance the efficiency of local MPI-based temperature estimation comprising MPI signals acquired at various temperatures and spatial positions along a defined line. Using MPI signals with 1D excitation and 100 averages, we demonstrate accurate estimation of absolute temperature (std <2 K) based on calibration with 11 individual matrices (~1 min each) in 2 K increments (Tmin=25°C, Tmax=45°C). This approach enables estimation of absolute temperatures within ~1 s and significantly reduces calibration time to ~20 minutes