We conclude that the tapping stimulation method is recommended for use in the design of tactile communication devices based on Finger-Braille and fingerspelling methods reliant on finger tapping actions. The results did not demonstrate clear evidence for tactile subitising with passively experienced stimulation on the fingers.Herein, we present a 16.8 nW ultra-low-power (ULP) energy harvester integrated circuit (IC) for ingestible biomedical sensors. The energy harvester can be powered from the electro-galvanic operation inside a human body, which provides a sustainable and long-term energy source. The challenge of dealing with relatively high input impedance (∼kΩ) of the bio-galvanic energy source is addressed by introducing two design techniques. The first technique is an adaptive VMPP-controlled algorithm (AVCA) for a maximum power point tracking (MPPT) controller, and the second technique is a ULP delay-line-based zero current switching (ZCS) controller. Different from the conventional fractional open-circuit voltage (FOCV) method for MPPT, the proposed AVCA allows continuous source tracking without detachment of the harvester from the source. The ZCS operation is achieved using a delay-line controller without using either a comparator or an opamp. The proposed AVCA is realized using a 12.1 nW MPPT controller. Successful ZCS operation is achieved using a 2.1 nW delay controller. Overall power consumption of the IC is 16.8 nW. The converter has been fabricated in a 0.18 μm CMOS process with 2 μm thick top-metal option. The measured result shows that the converter achieves a peak efficiency of 72.1% to generate 507 nW output power. The ULP operation allows a significant reduction in electrode size down to the submillimeter scale (∼0.4 mm2), demonstrating the good potential of the proposed energy harvester IC.Analog DNA strand displacement circuits can be used to build artificial neural network due to the continuity of dynamic behavior. In this study, DNA implementations of novel catalysis, novel degradation and adjustment reaction modules are designed and used to build an analog DNA strand displacement reaction network. A novel adaptive linear neuron (ADALINE) is constructed by the ordinary differential equations of an ideal formal chemical reaction network, which is built by reaction modules. When reaction network approaches equilibrium, the weights of the ADALINE are updated without learning algorithm. Simulation results indicate that, ADALINE based on the analog DNA strand displacement circuit has ability to implement the learning function of the ADALINE based on the ideal formal chemical reaction networks, and fit a class of linear function.This paper introduces embComp, a novel approach for comparing two embeddings that capture the similarity between objects, such as word and document embeddings. We survey scenarios where comparing these embedding spaces is useful. From those scenarios, we derive common tasks, introduce visual analysis methods that support these tasks, and combine them into a comprehensive system. One of embComp’s central features are overview visualizations that are based on metrics for measuring differences in the local structure around objects. Summarizing these local metrics over the embeddings provides global overviews of similarities and differences. Detail views allow comparison of the local structure around selected objects and relating this local information to the global views. Integrating and connecting all of these components, embComp supports a range of analysis workflows that help understand similarities and differences between embedding spaces. We assess our approach by applying it in several use cases, including understanding corpora differences via word vector embeddings, and understanding algorithmic differences in generating embeddings.Deep neural networks have been successfully applied to many real-world applications. However, such successes rely heavily on large amounts of labeled data that is expensive to obtain. Recently, many methods for semi-supervised learning have been proposed and achieved excellent performance. In this study, we propose a new EnAET framework to further improve existing semi-supervised methods with self-supervised information. To our best knowledge, all current semi-supervised methods improve performance with prediction consistency and confidence ideas. We are the first to explore the role of self-supervised representations in semi-supervised learning under a rich family of transformations. Consequently, our framework can integrate the self-supervised information as a regularization term to further improve all current semi-supervised methods. In the experiments, we use MixMatch, which is the current state-of-the-art method on semi-supervised learning, as a baseline to test the proposed EnAET framework. Across different datasets, we adopt the same hyper-parameters, which greatly improves the generalization ability of the EnAET framework. Experiment results on different datasets demonstrate that the proposed EnAET framework greatly improves the performance of current semi-supervised algorithms. Moreover, this framework can also improve supervised learning by a large margin, including the extremely challenging scenarios with only 10 images per class. The code and experiment records are available in https//github.com/maple-research-lab/EnAET.This work presents a new method to analyze weak distributed nonlinear (NL) effects, with a focus on the generation of harmonics (H) and intermodulation products (IMD) in bulk acoustic wave (BAW) resonators and filters composed of them. The method consists of finding equivalent current sources [input-output equivalent sources (IOES)] at the H or IMD frequencies of interest that are applied to the boundary nodes of any layer that can contribute to the nonlinearities according to its local NL constitutive equations. The new methodology is compared with the harmonic balance (HB) analysis, by means of a commercial tool, of a discretized NL Mason model, which is the most used model for NL BAW resonators. While the computation time is drastically reduced, the results are fully identical. For the simulation of a seventh-order filter, the IOES method is around 700 times faster than the HB simulations.This article presents a motion compensation procedure that significantly improves the accuracy of synthetic aperture tensor velocity estimates for row-column arrays. The proposed motion compensation scheme reduces motion effects by moving the image coordinates with the velocity field during summation of low-resolution volumes. The velocity field is estimated using a transverse oscillation cross-correlation estimator, and each image coordinate’s local tensor velocity is determined by upsampling the field using spline interpolation. The motion compensation procedure is validated using Field II simulations and flow measurements acquired using a 3-MHz row-column addressed probe and the research scanner SARUS. For a peak velocity of 25 cm/s, a pulse repetition frequency of 2 kHz, and a beam-to-flow angle of 60°, the proposed motion compensation procedure was able to reduce the relative bias from -27.0% to -9.4% and the standard deviation from 8.6% to 8.1%. In simulations performed with a pulse repetition frequency of 10 kHz, the proposed method reduces the bias in all cases with beam-to-flow angles of 60° and 75° and peak velocities between 10 and 150 cm/s.This paper addresses digital staining and classification of the unstained white blood cell images obtained with a differential contrast microscope. We have data coming from multiple domains that are partially labeled and partially matching across the domains. Using unstained images removes time-consuming staining procedures and could facilitate and automatize comprehensive diagnostics. To this aim, we propose a method that translates unstained images to realistically looking stained images preserving the inter-cellular structures, crucial for the medical experts to perform classification. We achieve better structure preservation by adding auxiliary tasks of segmentation and direct reconstruction. Segmentation enforces that the network learns to generate correct nucleus and cytoplasm shape, while direct reconstruction enforces reliable translation between the matching images across domains. Besides, we build a robust domain agnostic latent space by injecting the target domain label directly to the generator, i.e., bypassing the encoder. It allows the encoder to extract features independently of the target domain and enables an automated domain invariant classification of the white blood cells. We validated our method on a large dataset composed of leukocytes of 24 patients, achieving state-of-the-art performance on both digital staining and classification tasks.Wall shear stress (WSS) has been suggested as a potential biomarker in various cardiovascular diseases and it can be estimated from phase-contrast Magnetic Resonance Imaging (PC-MRI) velocity measurements. We present a parametric sequential method for MRI-based WSS quantification consisting of a geometry identification and a subsequent approximation of the velocity field. This work focuses on its validation, investigating well controlled high-resolution in vitro measurements of turbulent stationary flows and physiological pulsatile flows in phantoms. Initial tests for in vivo 2D PC-MRI data of the ascending aorta of three volunteers demonstrate basic applicability of the method to in vivo.We propose weakly supervised training schemes to train end-to-end cell segmentation networks that only require a single point annotation per cell as the training label and generate a high-quality segmentation mask close to those fully supervised methods using mask annotation on cells. Three training schemes are investigated to train cell segmentation networks, using the point annotation. First, self-training is performed to learn additional information near the annotated points. Next, cotraining is applied to learn more cell regions using multiple networks that supervise each other. Finally, a hybrid-training scheme is proposed to leverage the advantages of both self-training and co-training. During the training process, we propose a divergence loss to avoid the overfitting and a consistency loss to enforce the consensus among multiple co-trained networks. Furthermore, we propose weakly supervised learning with human in the loop, aiming at achieving high segmentation accuracy and annotation efficiency simultaneously. Evaluated on two benchmark datasets, our proposal achieves high-quality cell segmentation results comparable to the fully supervised methods, but with much less amount of human annotation effort.

Spinal fusion surgeries require accurate placement of pedicle screws in anatomic corridors without breaching bone boundaries. We are developing a combined ultrasound and photoacoustic image guidance system to avoid pedicle screw misplacement and accidental bone breaches, which can lead to nerve damage.

Pedicle cannulation was performed on a human cadaver, with co-registered photoacoustic and ultrasound images acquired at various time points during the procedure. Bony landmarks obtained from coherence-based ultrasound images of lumbar vertebrae were registered to post-operative CT images. Registration methods were additionally tested on an ex vivo caprine vertebra.

Locally weighted short-lag spatial coherence (LW-SLSC) ultrasound imaging enhanced the visualization of bony structures with generalized contrast-to-noise ratios (gCNRs) of 0.99 and 0.98-1.00 in the caprine and human vertebrae, respectively. Short-lag spatial coherence (SLSC) and amplitude-based delay-and-sum (DAS) ultrasound imaging generally produced lower gCNRs of 0.