Search

The study objective was to investigate the performance of a dedicated convolutional neural network (CNN) optimized for wrist cartilage segmentation from 2D MR images. CNN utilized a planar architecture and patch-based (PB) training approach that ensured optimal performance in the presence of a limited amount of training data. The CNN was trained and validated in twenty multi-slice MRI datasets acquired with two different coils in eleven subjects (healthy volunteers and patients). The validation included a comparison with the alternative state-of-the-art CNN methods for the segmentation of joints from MR images and the ground-truth manual segmentation. When trained on the limited training data, the CNN outperformed significantly image-based and patch-based U-Net networks. Our PB-CNN also demonstrated a good agreement with manual segmentation (Sorensen-Dice similarity coefficient (DSC) = 0.81) in the representative (central coronal) slices with large amount of cartilage tissue. Reduced performance of the network for slices with a very limited amount of cartilage tissue suggests the need for fully 3D convolutional networks to provide uniform performance across the joint. The study also assessed inter- and intra-observer variability of the manual wrist cartilage segmentation (DSC=0.78-0.88 and 0.9, respectively). The proposed deep-learning-based segmentation of the wrist cartilage from MRI could facilitate research of novel imaging markers of wrist osteoarthritis to characterize its progression and response to therapy.

Metasurfaces represent a new paradigm in artificial subwavelength structures due to their potential to overcome many challenges typically associated with bulk metamaterials. The ability making very thin structures and change their properties dynamically make metasurfaces an exceptional meta-optics platform for engineering advanced electromagnetic and photonic metadevices. Here, we suggest and demonstrate experimentally a novel tunable metasurface capable to enhance significantly the local image quality in magnetic resonance imaging (MRI). We present a design of the hybrid metasurface based on electromagnetically-coupled dielectric and metallic elements. We demonstrate how to tailor the spectral characteristics of the metasurface eigenmodes by changing dynamically the effective permittivity of the structure. By maximizing a coupling between metasurface eigenmodes and transmitted and received fields in the MRI system, we enhance the device sensitivity that results in a substantial improvement of the image quality., Comment: 4 figures

In this contribution we present numerical and experimental results of a parametric quantitative study of radiative dipole antennas in a phased array configuration for efficient body magnetic resonance imaging at 7T via parallel transmission. For magnetic resonance imaging (MRI) at ultrahigh fields (7T and higher) dipole antennas are commonly used in phased arrays, particularly for body imaging targets. This study reveals the effects of dipole positioning in the array (elevation of dipoles above the subject and inter-dipole spacing) on their mutual coupling, $B_1^{+}$ per $P_{acc}$ and $B_1^{+}$ per maximum local SAR efficiencies as well as the RF-shimming capability. The numerical and experimental results are obtained and compared for a homogeneous phantom as well as for a real human models confirmed by in-vivo experiments.

Metasurfaces are artificial electromagnetic boundaries or interfaces usually implemented as two-dimensional periodic structures with subwavelength periodicity and engineered properties of constituent unit cells. The electromagnetic bandgap (EBG) effect in metasurfaces prevents all surface modes from propagating in a certain frequency band. While metasurfaces provide a number of important applications in microwave antennas and antenna arrays, their features are also highly suitable for MRI applications. In this work we manufacture and experimentally study finite samples based on mushroom-type EBG metasurfaces and employ them for suppression of inter-element coupling in dipole transmit coil arrays for body imaging at 7T. We show experimentally that employment of the samples EBG leads to reduction of coupling between adjacent closely-spaced dipole antenna elements of a 7T transmit/receive body array, which reduces scattering losses in neighboring channels and thereby improves the B1+ efficiency. The setup consists of two fractionated dipole antennas previously designed by the authors for body imaging at 7 Tesla. These are placed on top of a body-mimicking phantom and equipped with the manufactured finite-size sample of the metasurface tuned to have EBG properties at the Larmor frequency of 298 MHz. To improve the detection range of the B1+ field distribution of the top elements, four additional elements were positioned along the bottom side of the phantom. Scattering matrix measurements show that coupling between the two top elements is indeed reduced while the measurements performed on a 7T MRI machine confirm the array's B1+ efficiency improvement due to reduced scattering losses. This study provides a tool for the decoupling of dipole antennas in ultrahigh field transmit arrays, possibly resulting in denser element placement and/or larger subject-element spacing.