We evaluate our model for liver tumor segmentation using a PET-CT scans dataset, compare our method with the baseline techniques for multi-modal (multi-branches, multi-channels and cascaded networks) and then demonstrate that our method has a significantly higher accuracy (p less then 0.05) than the baseline models.We propose VFR-UFD, a new deep learning framework that performs vector field reconstruction (VFR) for unsteady flow data (UFD). Given integral flow lines (i.e., streamlines), we first generate low-quality UFD via diffusion. VFR- UFD then leverages a convolutional neural network to reconstruct spatiotemporally coherent, high-quality UFD. The core of VFR-UFD lies in recurrent residual blocks that iteratively refine and denoise the input vector fields at different scales, both locally and globally. We take consecutive time steps as input to capture temporal coherence and apply streamline-based optimization to preserve spatial coherence. To show the effectiveness of VFR-UFD, we experiment with several vector field data sets to report quantitative and qualitative results and compare VFR-UFD with two VFR methods and one compression algorithm.Non-Rigid Structure-from-Motion (NRSfM) reconstructs a deformable 3D object from the correspondences establishedbetween monocular 2D images. Current NRSfM methods lack statistical robustness, which is the ability to cope with correspondenceerrors. This prevents one to use automatically established correspondences, which are prone to errors, thereby strongly limiting thescope of NRSfM. We propose a three-step automatic pipeline to solve NRSfM robustly by exploiting isometry. Step(i)computes theoptical flow from correspondences, step(ii)reconstructs each 3D points normal vector using multiple reference images and integratesthem to form surfaces with the best reference and step(iii)rejects the 3D points that break isometry in their local neighborhood.Importantly, each step is designed to discard or flag erroneous correspondences. Our contributions include the robustification of opticalflow by warp estimation, new fast analytic solutions to local normal reconstruction and their robustification, and a newscale-independent measure of 3D local isometric coherence. Experimental results show that our robust NRSfM method consistentlyoutperforms existing methods on both synthetic and real datasets.It is appealing but challenging to achieve real-time deep neural network (DNN) inference on mobile devices, because even the powerful modern mobile devices are considered as '`resource-constrained'' when executing large-scale DNNs. It necessitates the sparse model inference via weight pruning, i.e., DNN weight sparsity, and it is desirable to design a new DNN weight sparsity scheme that can facilitate real-time inference on mobile devices while preserving a high sparse model accuracy. This paper designs a novel mobile inference acceleration framework GRIM that is General to both convolutional neural networks (CNNs) and recurrent neural networks (RNNs) and that achieves Real-time execution and high accuracy, leveraging fine-grained structured sparse model Inference and compiler optimizations for Mobiles. We start by proposing a new fine-grained structured sparsity scheme through the Block-based Column-Row (BCR) pruning. Based on this new fine-grained structured sparsity, our GRIM framework consists of two parts (a) the compiler optimization and code generation for real-time mobile inference; and (b) the BCR pruning optimizations for determining pruning hyperparameters and performing weight pruning. We compare GRIM with Alibaba MNN, TVM, TensorFlow-Lite, a sparse implementation based on CSR, PatDNN, and ESE (a representative FPGA inference acceleration framework for RNNs), and achieve up to 14.08x speedup. Muscle biomechanics is set by the spacing of repetitive striation patterns of individual sarcomeres within single muscle fibres of stacked myofibrils. Sarcomere lengths (SL) are rather unequally distributed than of equal distance. This non-uniformity may affect both, force production as well as passive-elastic deformation. However, online recording of SL during axially imposed strains is cumbersome due to a lack of compact technologies. To fuse SL pattern recognition with restoration force assessments during quasi-static axial stretch, we implemented live tracking of SL distributions simultaneous to voice-coil actuated stretch and restoration force recordings in our MyoRobot 2.0 automated biomechatronics platform. Both were obtained online during stretchrelaxation cycles of murine single muscle fibres. Under quasi-static stretch conditions (∼1 μm/s fibre length changes), almost no apparent hysteresis was detected in single fibres. SL showed a non-uniform distribution. While mean SL varied between 2.6 μm and 3.4 μm upon 140% stretch, two populations of fibres were noticed one showing a minor change in SL distribution with stretch, and one becoming more equally distributed upon stretch. A roughly 5% SL variability under rest either diminishes or remains almost unaltered upon elastic axial deformation. This may reflect differential impact of mostly extra-sarcomeric components to stretch in this stretch range. The augmented functionality of the MyoRobot 2.0 towards online sarcomere analyses within single fibres shall provide a valuable tool for the muscle community to study the contribution of serial elastic and force producing elements in health and disease models.The augmented functionality of the MyoRobot 2.0 towards online sarcomere analyses within single fibres shall provide a valuable tool for the muscle community to study the contribution of serial elastic and force producing elements in health and disease models.Visually guided behavior relies on the integration of sensory input and information held in working memory (WM). Yet it remains unclear how this is accomplished at the level of neural circuits. We studied the direct visual cortical inputs to neurons within a visuomotor area of prefrontal cortex in behaving monkeys. We show that the efficacy of visual input to prefrontal cortex is gated by information held in WM. Surprisingly, visual input to prefrontal neurons was found to target those with both visual and motor properties, rather than preferentially targeting other visual neurons. selleck products Furthermore, activity evoked from visual cortex was larger in magnitude, more synchronous, and more rapid, when monkeys remembered locations that matched the location of visual input. These results indicate that WM directly influences the circuitry that transforms visual input into visually guided behavior.

Incomplete neoendothelialization associated with still left atrial ap

Description

RAW Paste Data

Recent Pastes