With traditional beamforming methods, ultrasound B-mode images contain speckle noise caused by the random interference of subresolution scatterers. In this paper, we present a framework for using neural networks to beamform channel signals into speckle-reduced images. We introduce log-domain normalization-independent loss functions that are appropriate imaging. A fully convolutional network was trained with simulated were coregistered spatially ground-truth maps echogenicity. Networks...

Phase aberration is widely considered a major source of image degradation in medical pulse-echo ultrasound. Traditionally, near-field phase correction techniques are unable to account for distributed aberrations due spatially varying speed sound the medium, while most require use point-like sources and impractical clinical applications where diffuse scattering dominant. Here, we present two that utilize estimates from tomographic estimator builds on our previous work with layered media. We...

We report on the performance of a vector apodizing phase plate coronagraph that operates over wavelength range $2-5 μ$m and is installed in MagAO/Clio2 at 6.5 m Magellan Clay telescope Las Campanas Observatory, Chile. The manipulates pupil to produce three beams yielding two coronagraphic point-spread functions (PSFs) one faint leakage PSF. pattern imposed through inherently achromatic geometric phase, enabled by liquid crystal technology polarization techniques. optic manufactured using...

Diffuse reverberation is ultrasound image noise caused by multiple reflections of the transmitted pulse before returning to transducer, which degrades quality and impedes estimation displacement or flow in techniques such as elastography Doppler imaging. appears spatially incoherent channel signals, where it also performance adaptive beamforming methods, sound speed estimation, methods that require measurements from signals. In this paper, we propose a custom 3D fully convolutional neural...

Ultrasound molecular imaging (UMI) is enabled by targeted microbubbles (MBs), which are highly reflective ultrasound contrast agents that bind to specific biomarkers. Distinguishing between adherent MBs and background signals can be challenging in vivo. The preferred preclinical technique differential enhancement (DTE), wherein a strong acoustic pulse used destroy verify their locations. However, DTE intrinsically cannot for real-time may cause undesirable bioeffects. In this work, we...

Fiber optic temperature sensors were fabricated by depositing vanadium oxide thin films on the tips of optical fibers, and incorporating materials into core fibers. It was found that properties initially amorphous can be controllably converted to those crystalline VOx compounds via plasma arc a fiber fusion splicer. These then over-coated with SiO2, subsequently fused another form an in-line sensor. well defined absorption edge formed when (VOx) annealed using splicer, suggesting formation...

This research represents a pioneering application of automated pose estimation from drone data to study elephant behavior in the wild, utilizing video footage captured Samburu National Reserve, Kenya. The evaluates two workflows: DeepLabCut, known for its laboratory settings and emerging wildlife fieldwork, YOLO-NAS-Pose, newly released model not previously applied behavioral studies. These models are trained analyze herd behavior, focusing on low-resolution ($\sim$50 pixels) subjects detect...

Reverberation clutter is image noise resulting from multiple reflections between tissue layers and leads to degradation. We propose a 3D fully convolutional neural network (FCNN) selectively remove reverberation ultrasound channel data. A training set was generated using Field II Pro simulate full synthetic aperture approximated by adding bandpass filtered the noise-free simulated This data, with without noise, used train FCNN custom architecture thermal noise. The evaluated on validation an...

ABSTRACT While deep learning has revolutionized ecological data analysis, existing strategies often rely on supervised learning, which is subject to limitations real-world applicability. In this paper, we apply self-supervised methods bioacoustic enable unsupervised detection of event boundaries. We propose a convolutional neural network that operates the raw waveform directly and trained in accordance with Noise Contrastive Estimation principle, enables system detect spectral changes input...

We present the technical design of art installation Rainbow Station, that projects a 40-m diameter true-color rainbow.The core technology is comprised patterned polarization grating produces rainbow with correct shape and color order.We achieve an effective period as small 1.55 µm, obtain high diffraction efficiency over entire visible spectral range thanks to multi-layer liquid-crystal implementation.The -1 order suppressed by circular filtering.

Multi-Twist Retarders (MTRs) are a class of complex retarders created by stacking multiple layers various polymerizable nematic liquid crystal mixtures, where chiral dopant may be added to each layer control its twisting power. MTRs have set twist angles and thicknesses for that selected use merit function. In this work, we utilize these degrees freedom create single retarder film can generate variety highly-chromatic retardation spectra, selectable the relative orientation MTR input...

Imaging spectrometers are frequently used in remote sensing for their increased target discrimination capabilities over conventional imaging. Increasing the spectral resolution of these sensors further enables system's ability to discriminate certain targets and adds potential monitoring narrow-line features. We describe a high (Δλ=1.1 nm full-width at half maximum) snapshot imaging spectrometer capable distinguishing two narrowly separated bands red-visible spectrum. A theoretical model is...

We show how highly chromatic Multi-Twist Retarder (MTR) films can be used to create a single-film color filter wherein the may selected only by MTR orientation angle. By this approach, we multi- images with just an between polarizers. study design method and limits of available gamut possibilities in experimentally demonstrate several designs continuous discrete patterns. This technique useful art, displays, microscopy, remote sensing.

Targeted microbubbles (MBs)enable ultrasound molecular imaging (USMI)by binding to specific biomarkers and producing strong reflections ultrasound. However, current USMI techniques are not easily translatable for clinical use. In particular, preclinical studies often utilize destruction-subtraction imaging, wherein a destructive pulse is used destroy MBs confirm their locations. This approach potentially unsafe, intrinsically real-time. this work, novel neural-network-based beamformer...

Polarization spatial heterodyne interferometry (PSHI) allows for the development of compact, vibration insensitive, high spectral resolution sensors. Introducing imaging qualities a lenslet array extends advantages PSHI to interferometers. The use Savart plates enables birefringent interferometer that obtains higher with fewer optical aberrations when compared alternative designs. In this paper, we describe design, construction, calibration and validation narrowband emission line...

Deep neural networks (DNNs) have recently emerged as powerful function approximators that can learn to transform given inputs into desired outputs when provided with enough training samples. Ultrasound B-mode images are traditionally formed using the delay-and-sum beamformer display echogenicity (i.e., backscattering strength) of medium. Advanced beamformers often designed improve indirect metrics image quality such lesion detectability and point target resolution. Here, we describe how DNNs...

Isolating the mainlobe and sidelobe contribution to ultrasound image can improve imaging contrast by removing off-axis clutter. Previous work achieves this separation of contributions based on covariance received signals. However, formation a matrix at each point be computationally burdensome memory intensive for real-time applications. Our demonstrates that isolated receive aperture spectrum, greatly reducing computational requirements.

Deep learning has gained tremendous popularity as a tool for ultrasound beamforming and image reconstruction. In previous work, we trained deep neural networks (DNNs) to estimate the echogenicity of medium, improve acoustical electronic signal-to-noise ratio (SNR) in channel data, detect targeted microbubbles nondestructively real-time molecular imaging. Here, present several advancements each application. First, compare speckle- noise-reducing performance DNNs with simple linear Field II...