Deep learning-based fully automatic segmentation of wrist cartilage in MR images.

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 20 multi-slice MRI datasets acquired with two different coils in 11 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 PB-U-Net networks. Our PB-CNN also demonstrated a good agreement with manual segmentation (Sørensen-Dice similarity coefficient [DSC] = 0.81) in the representative (central coronal) slices with a 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.

Volumetric wireless coil based on periodically coupled split-loop resonators for clinical wrist imaging.

PURPOSE: Design and characterization of a new inductively driven wireless coil (WLC) for wrist imaging at 1.5 T with high homogeneity operating due to focusing the B1 field of a birdcage body coil. METHODS: The WLC design has been proposed based on a volumetric self-resonant periodic structure of inductively coupled split-loop resonators with structural capacitance. The WLC was optimized and studied regarding radiofrequency fields and interaction to the birdcage coil (BC) by electromagnetic simulations. The manufactured WLC was characterized by on-bench measurements and in vivo and phantom study in comparison to a standard cable-connected receive-only coil. RESULTS: The WLC placed into BC gave the measured B1+ increase of the latter by 8.6 times for the same accepted power. The phantom and in vivo wrist imaging showed that the BC in receiving with the WLC inside reached equal or higher signal-to-noise ratio than the conventional clinical setup comprising the transmit-only BC and a commercial receive-only flex-coil and created no artifacts. Simulations and on-bench measurements proved safety in terms of specific absorption rate and reflected transmit power. CONCLUSIONS: The results showed that the proposed WLC could be an alternative to standard cable-connected receive coils in clinical magnetic resonance imaging. As an example, with no cable connection, the WLC allowed wrist imaging on a 1.5 T clinical machine using a full-body BC for transmitting and receive with the desired signal-to-noise ratio, image quality, and safety.