Showing papers by "Philips published in 2015"

Definitions for a common standard for 2D speckle tracking echocardiography: consensus document of the EACVI/ASE/Industry Task Force to standardize deformation imaging

Abstract: Recognizing the critical need for standardization in strain imaging, in 2010, the European Association of Echocardiography (now the European Association of Cardiovascular Imaging, EACVI) and the American Society of Echocardiography (ASE) invited technical representatives from all interested vendors to participate in a concerted effort to reduce intervendor variability of strain measurement. As an initial product of the work of the EACVI/ASE/Industry initiative to standardize deformation imaging, we prepared this technical document which is intended to provide definitions, names, abbreviations, formulas, and procedures for calculation of physical quantities derived from speckle tracking echocardiography and thus create a common standard.

A revolution in lighting

Abstract: Key materials discoveries have prompted the rise of inorganic light-emitting diodes in the lighting industry. Remaining challenges are being addressed to further extend the impact of this technology in lighting, displays and other applications.

Kernel Interpolation for Scalable Structured Gaussian Processes (KISS-GP)

Abstract: We introduce a new structured kernel interpolation (SKI) framework, which generalises and unifies inducing point methods for scalable Gaussian processes (GPs). SKI methods produce kernel approximations for fast computations through kernel interpolation. The SKI framework clarifies how the quality of an inducing point approach depends on the number of inducing (aka interpolation) points, interpolation strategy, and GP covariance kernel. SKI also provides a mechanism to create new scalable kernel methods, through choosing different kernel interpolation strategies. Using SKI, with local cubic kernel interpolation, we introduce KISS-GP, which is 1) more scalable than inducing point alternatives, 2) naturally enables Kronecker and Toeplitz algebra for substantial additional gains in scalability, without requiring any grid data, and 3) can be used for fast and expressive kernel learning. KISS-GP costs O(n) time and storage for GP inference. We evaluate KISS-GP for kernel matrix approximation, kernel learning, and natural sound modelling.

Prevalence, Vascular Distribution, and Multiterritorial Extent of Subclinical Atherosclerosis in a Middle-Aged Cohort: The PESA (Progression of Early Subclinical Atherosclerosis) Study.

Abstract: Background—Data are limited on the presence, distribution, and extent of subclinical atherosclerosis in middle-aged populations. Methods and Results—The PESA (Progression of Early Subclinical Ather...

Exploiting Spatial Redundancy of Image Sensor for Motion Robust rPPG

Abstract: Remote photoplethysmography (rPPG) techniques can measure cardiac activity by detecting pulse-induced color variations on human skin using an RGB camera. State-of-the-art rPPG methods are sensitive to subject body motions (e.g., motion-induced color distortions). This study proposes a novel framework to improve the motion robustness of rPPG. The basic idea of this paper originates from the observation that a camera can simultaneously sample multiple skin regions in parallel, and each of them can be treated as an independent sensor for pulse measurement. The spatial redundancy of an image sensor can thus be exploited to distinguish the pulse signal from motion-induced noise. To this end, the pixel-based rPPG sensors are constructed to estimate a robust pulse signal using motion-compensated pixel-to-pixel pulse extraction, spatial pruning, and temporal filtering. The evaluation of this strategy is not based on a full clinical trial, but on 36 challenging benchmark videos consisting of subjects that differ in gender, skin types, and performed motion categories. Experimental results show that the proposed method improves the SNR of the state-of-the-art rPPG technique from 3.34 to 6.76 dB, and the agreement ( $\pm 1.96\sigma$ ) with instantaneous reference pulse rate from 55% to 80% correct. ANOVA with post hoc comparison shows that the improvement on motion robustness is significant. The rPPG method developed in this study has a performance that is very close to that of the contact-based sensor under realistic situations, while its computational efficiency allows real-time processing on an off-the-shelf computer.

MRBrainS challenge: online evaluation framework for brain image segmentation in 3T MRI scans

Abstract: Many methods have been proposed for tissue segmentation in brain MRI scans. The multitude of methods proposed complicates the choice of one method above others. We have therefore established the MRBrainS online evaluation framework for evaluating (semi)automatic algorithms that segment gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF) on 3T brain MRI scans of elderly subjects (65-80 y). Participants apply their algorithms to the provided data, after which their results are evaluated and ranked. Full manual segmentations of GM, WM, and CSF are available for all scans and used as the reference standard. Five datasets are provided for training and fifteen for testing. The evaluated methods are ranked based on their overall performance to segment GM, WM, and CSF and evaluated using three evaluation metrics (Dice, H95, and AVD) and the results are published on the MRBrainS13 website. We present the results of eleven segmentation algorithms that participated in the MRBrainS13 challenge workshop at MICCAI, where the framework was launched, and three commonly used freeware packages: FreeSurfer, FSL, and SPM. The MRBrainS evaluation framework provides an objective and direct comparison of all evaluated algorithms and can aid in selecting the best performing method for the segmentation goal at hand.

Personalizing persuasive technologies

Abstract: This paper discusses how persuasive technologies can be made adaptive to users. We present persuasion profiling as a method to personalize the persuasive messages used by a system to influence its users. This type of personalization can be based on explicit measures of users? tendencies to comply to distinct persuasive strategies: measures based on standardized questionnaire scores of users. However, persuasion profiling can also be implemented using implicit, behavioral measures of user traits. We present three case studies involving the design, implementation, and field deployment of personalized persuasive technologies, and we detail four design requirements. In each case study we show how these design requirements are implemented. In the discussion we highlight avenues for future research in the field of adaptive persuasive technologies. Author-HighlightsPersuasive technologies can be more effective if they are personalized.We introduce persuasion profiles to personalize persuasive messages.Persuasion profiles can be effective using implicit or explicit measures.In three case studies we show the effects of personalized persuasion.

Metal Oxide Induced Charge Transfer Doping and Band Alignment of Graphene Electrodes for Efficient Organic Light Emitting Diodes

Abstract: The interface structure of graphene with thermally evaporated metal oxide layers, in particular molybdenum trioxide (MoO3), is studied combining photoemission spectroscopy, sheet resistance measurements and organic light emitting diode (OLED) characterization. Thin (<5 nm) MoO3 layers give rise to an 1.9 eV large interface dipole and a downwards bending of the MoO3 conduction band towards the Fermi level of graphene, leading to a near ideal alignment of the transport levels. The surface charge transfer manifests itself also as strong and stable p-type doping of the graphene layers, with the Fermi level downshifted by 0.25 eV and sheet resistance values consistently below 50 Ω/sq for few-layer graphene films. The combination of stable doping and highly efficient charge extraction/injection allows the demonstration of simplified graphene-based OLED device stacks with efficiencies exceeding those of standard ITO reference devices.

Magnetic Particle Imaging With Tailored Iron Oxide Nanoparticle Tracers

Abstract: Magnetic particle imaging (MPI) shows promise for medical imaging, particularly in angiography of patients with chronic kidney disease. As the first biomedical imaging technique that truly depends on nanoscale materials properties, MPI requires highly optimized magnetic nanoparticle tracers to generate quality images. Until now, researchers have relied on tracers optimized for MRI ${\rm T}2^ {\ast} $ -weighted imaging that are sub-optimal for MPI. Here, we describe new tracers tailored to MPI's unique physics, synthesized using an organic-phase process and functionalized to ensure biocompatibility and adequate in vivo circulation time. Tailored tracers showed up to 3 $\,\times$ greater signal-to-noise ratio and better spatial resolution than existing commercial tracers in MPI images of phantoms.

Native T1 in Discrimination of Acute and Convalescent Stages in Patients With Clinical Diagnosis of Myocarditis: A Proposed Diagnostic Algorithm Using CMR

Abstract: Objectives This study investigated whether T1 mapping by cardiac magnetic resonance (CMR) reflects the clinical evolution of disease in myocarditis and supports its diagnosis independently of the disease stages. Background Acute viral myocarditis is characterized by a range of intracellular changes due to viral replication and extracellular spill of debris within days of viral infection. Convalescence may be characterized by a chronic low-grade inflammation leading to ventricular remodelling, but also a complete resolution of myocardial changes. Methods Patients with clinical diagnosis of viral myocarditis (N = 165) underwent routine clinical CMR protocol (1.5- and 3.0-T) for assessment of cardiac function and structure, and tissue characterization with T2-weighted imaging and late gadolinium enhancement. T1 mapping was obtained in a mid-ventricular short-axis slice before and >20 min after administration of 0.2 mmol/kg of gadobutrol. Results Compared with control subjects (n = 40), T1 indexes were increased in patients with myocarditis. Patients with acute symptoms (n = 61) had higher values of T1 indexes compared with patients in clinical convalescence (n = 67). Native T1 is an independent discriminator between health and disease, as well as a discriminator between acute and convalescent stage of the disease. Native T1- was superior to T2-weighted imaging and late gadolinium enhancement with high diagnostic accuracy and positive and negative predictive values. Using pre-defined cutoff values for normal ranges, we demonstrated that acute myocarditis can be independently identified by native T1 of >5 SD above the mean of normal range, whereas convalescence is best defined by either abnormal native T1 (>2 SD) or presence of late gadolinium enhancement. We prospectively tested a new diagnostic algorithm in an independent dataset of patients with clinical diagnosis of myocarditis and achieved similar diagnostic performance. Conclusions The new diagnostic algorithm using native T1 can reliably discriminate between health and disease and determine the clinical disease stage in patients with a clinical diagnosis of myocarditis.

Dual-Modality Photoacoustic and Ultrasound Imaging System for Noninvasive Sentinel Lymph Node Detection in Patients with Breast Cancer

Abstract: The detection of regional lymph node metastases is important in cancer staging as it guides the prognosis of the patient and the strategy for treatment. Sentinel lymph node biopsy (SLNB) is an accurate, less invasive alternative to axillary lymph node dissection. The sentinel lymph node hypothesis states that the pathological status of the axilla can be accurately predicted by determining the status of the first lymph nodes that drain from the primary tumor. Physicians use radio-labeled sulfur colloid and/or methylene blue dye to identify the SLN, which is most likely to contain metastatic cancer cells. However, the surgical procedure causes morbidity and associated expenses. To overcome these limitations, we developed a dual-modality photoacoustic and ultrasonic imaging system to noninvasively detect SLNs based on the accumulation of methylene blue dye. Ultimately, we aim to guide percutaneous needle biopsies and provide a minimally invasive method for axillary staging of breast cancer.

First experimental evidence of the feasibility of multi-color magnetic particle imaging

Abstract: Magnetic particle imaging is a new approach to visualizing magnetic nanoparticles. It is capable of 3D real-time in vivo imaging of particles injected into the blood stream and is a candidate for medical imaging applications. To date, only one particle type has been imaged at a time, however, the ability to separate signals acquired simultaneously from different particle types or from particles in different environments would substantially increase the scope of the method. Different colors could be assigned to different signal sources to allow for visualization in a single image. Successful signal separation has been reported in spectroscopic experiments, but it was unclear how well separation would work in conjunction with spatial encoding in an imaging experiment. This work presents experimental evidence of the separability of signals from different particle types and aggregation states (fluid versus powder) using a 'multi-color' reconstruction approach. Several mechanisms are discussed that may form the basis for successful signal separation.

Benchmark for Algorithms Segmenting the Left Atrium From 3D CT and MRI Datasets

Abstract: Knowledge of left atrial (LA) anatomy is important for atrial fibrillation ablation guidance, fibrosis quantification and biophysical modelling. Segmentation of the LA from Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) images is a complex problem. This manuscript presents a benchmark to evaluate algorithms that address LA segmentation. The datasets, ground truth and evaluation code have been made publicly available through the http://www.cardiacatlas.org website. This manuscript also reports the results of the Left Atrial Segmentation Challenge (LASC) carried out at the STACOM’13 workshop, in conjunction with MICCAI’13. Thirty CT and 30 MRI datasets were provided to participants for segmentation. Each participant segmented the LA including a short part of the LA appendage trunk and proximal sections of the pulmonary veins (PVs). We present results for nine algorithms for CT and eight algorithms for MRI. Results showed that methodologies combining statistical models with region growing approaches were the most appropriate to handle the proposed task. The ground truth and automatic segmentations were standardised to reduce the influence of inconsistently defined regions (e.g., mitral plane, PVs end points, LA appendage). This standardisation framework, which is a contribution of this work, can be used to label and further analyse anatomical regions of the LA. By performing the standardisation directly on the left atrial surface, we can process multiple input data, including meshes exported from different electroanatomical mapping systems.

Performance of T1 and T2 Mapping Cardiovascular Magnetic Resonance to Detect Active Myocarditis in Patients With Recent-Onset Heart Failure

Abstract: Background— This study evaluated the performance of novel quantitative T1 and T2 mapping cardiovascular magnetic resonance (CMR) techniques to identify active myocarditis in patients with recent-onset heart failure. Methods and Results— Thirty-one consecutive patients with recent-onset heart failure, reduced left ventricular function and clinically suspected myocarditis underwent endomyocardial biopsy and CMR at 1.5 Tesla. The CMR protocol included standard Lake-Louise parameters as well as T1 mapping using a modified Look-Locker inversion recovery sequence and T2 mapping using a hybrid gradient and spin-echo sequence. Short-axis maps were generated using an OsiriX plug-in to calculate global myocardial T1, T2, and extracellular volume fraction. Active myocarditis was defined by ongoing inflammation on endomyocardial biopsy. Endomyocardial biopsy revealed active myocarditis in 16 (52%) of 31 patients. Neither clinical characteristics, standard Lake-Louise CMR parameters, global myocardial T1 nor extracellular volume fraction differed significantly between patients with and without active myocarditis. However, median global myocardial T2 was significantly higher in patients with active myocarditis (65 ms [Q1–Q3, 61–70 ms]) than in patients without active myocarditis (59 ms [Q1–Q3, 55–64 ms]; P <0.01). A cutoff value for global myocardial T2 of ≥60 ms provided a sensitivity, specificity, accuracy, negative and positive predictive value of 94% (70%–100%), 60% (32%–84%), 77% (60%–89%), 90% (56%–100%), and 71% (48%–89%) for active myocarditis, respectively. Conclusions— T2 mapping seems to be superior when compared with standard CMR parameters, global myocardial T1, and extracellular volume fraction values for assessing the activity of myocarditis in patients with recent-onset heart failure and reduced left ventricular function.

Hexagonal silicon realized

Abstract: Silicon, arguably the most important technological semiconductor, is predicted to exhibit a range of new and interesting properties when grown in the hexagonal crystal structure. To obtain pure hexagonal silicon is a great challenge because it naturally crystallizes in the cubic structure. Here, we demonstrate the fabrication of pure and stable hexagonal silicon evidenced by structural characterization. In our approach, we transfer the hexagonal crystal structure from a template hexagonal gallium phosphide nanowire to an epitaxially grown silicon shell, such that hexagonal silicon is formed. The typical ABABAB... stacking of the hexagonal structure is shown by aberration-corrected imaging in transmission electron microscopy. In addition, X-ray diffraction measurements show the high crystalline purity of the material. We show that this material is stable up to 9 GPa pressure. With this development, we open the way for exploring its optical, electrical, superconducting, and mechanical properties.

Sleep stage classification with ECG and respiratory effort.

Abstract: Automatic sleep stage classification with cardiorespiratory signals has attracted increasing attention In contrast to the traditional manual scoring based on polysomnography, these signals can be measured using advanced unobtrusive techniques that are currently available, promising the application for personal and continuous home sleep monitoring This paper describes a methodology for classifying wake, rapid-eye-movement (REM) sleep, and non-REM (NREM) light and deep sleep on a 30 s epoch basis A total of 142 features were extracted from electrocardiogram and thoracic respiratory effort measured with respiratory inductance plethysmography To improve the quality of these features, subject-specific Z-score normalization and spline smoothing were used to reduce between-subject and within-subject variability A modified sequential forward selection feature selector procedure was applied, yielding 80 features while preventing the introduction of bias in the estimation of cross-validation performance PSG data from 48 healthy adults were used to validate our methods Using a linear discriminant classifier and a ten-fold cross-validation, we achieved a Cohen's kappa coefficient of 049 and an accuracy of 69% in the classification of wake, REM, light, and deep sleep These values increased to kappa = 056 and accuracy = 80% when the classification problem was reduced to three classes, wake, REM sleep, and NREM sleep

Magnetic resonance imaging in Alzheimer's Disease Neuroimaging Initiative 2

Abstract: Introduction Alzheimer's Disease Neuroimaging Initiative (ADNI) is now in its 10th year. The primary objective of the magnetic resonance imaging (MRI) core of ADNI has been to improve methods for clinical trials in Alzheimer's disease (AD) and related disorders. Methods We review the contributions of the MRI core from present and past cycles of ADNI (ADNI-1, -Grand Opportunity and -2). We also review plans for the future-ADNI-3. Results Contributions of the MRI core include creating standardized acquisition protocols and quality control methods; examining the effect of technical features of image acquisition and analysis on outcome metrics; deriving sample size estimates for future trials based on those outcomes; and piloting the potential utility of MR perfusion, diffusion, and functional connectivity measures in multicenter clinical trials. Discussion Over the past decade the MRI core of ADNI has fulfilled its mandate of improving methods for clinical trials in AD and will continue to do so in the future.

Discovery and Development of Pyridine-bis(imine) and Related Catalysts for Olefin Polymerization and Oligomerization.

Abstract: For over 40 years following the polyolefin catalyst discoveries of Hogan and Banks (Phillips) and Ziegler (Max Planck Institute), chemists traversed the periodic table searching for new transition metal and lanthanide-based olefin polymerization systems. Remarkably, none of these "hits" employed iron, that is, until three groups independently reported iron catalysts for olefin polymerization in the late 1990's. The history surrounding the discovery of these catalysts was only the beginning of their uniqueness, as the ensuing years have proven these systems remarkable in several regards. Of primary importance are the pyridine-bis(imine) ligands (herein referred to as PDI), which produced iron catalysts that are among the world's most active for ethylene polymerization, demonstrated "staying power" despite over 15 years of ligand improvement efforts, and generated highly active polymerization systems with cobalt, chromium, and vanadium. Although many ligands have been employed in iron-catalyzed polymerization, the PDI family has thus far provided the most information about iron's capabilities and tendencies. For example, iron systems tend to be highly selective for ethylene over higher olefins, making them strong candidates for producing highly crystalline polyethylene, or highly linear α-olefins. Iron PDI polymerizes propylene with 2,1-regiochemistry via a predominantly isotactic, chain end control mechanism. Because the first insertion proceeds via 1,2-regiochemistry, iron (and cobalt) PDI systems can be tailored to make highly linear dimers of α-olefins by "head-to-head" coupling, resulting from a switch in regiochemistry after the first insertion. Finally, PDI ligands, while not being surpassed in activity, have inspired the development of related ligand families and complexes, such as pendant donor diimines (PDD), which are also highly efficient at producing linear α-olefins. This Account will detail a variety of oligomerization and polymerization results achieved with PDI and PDD catalysts. Our studies on ligand modification are discussed, but numerous ligands have been synthesized by others. Computational approaches, identification of catalyst active sites, noninnocent ligand studies, commercialization efforts, and other outstanding research are only briefly mentioned, at most. The reader is directed to review articles where appropriate, in order to address the cursory treatment of these areas.

Design and baseline characteristics of the Food4Me study: a web-based randomised controlled trial of personalised nutrition in seven European countries

Abstract: Improving lifestyle behaviours has considerable potential for reducing the global burden of non-communicable diseases, promoting better health across the life-course and increasing well-being. However, realising this potential will require the development, testing and implementation of much more effective behaviour change interventions than are used conventionally. Therefore, the aim of this study was to conduct a multi-centre, web-based, proof-of-principle study of personalised nutrition (PN) to determine whether providing more personalised dietary advice leads to greater improvements in eating patterns and health outcomes compared to conventional population-based advice. A total of 5,562 volunteers were screened across seven European countries; the first 1,607 participants who fulfilled the inclusion criteria were recruited into the trial. Participants were randomly assigned to one of the following intervention groups for a 6-month period: Level 0-control group-receiving conventional, non-PN advice; Level 1-receiving PN advice based on dietary intake data alone; Level 2-receiving PN advice based on dietary intake and phenotypic data; and Level 3-receiving PN advice based on dietary intake, phenotypic and genotypic data. A total of 1,607 participants had a mean age of 39.8 years (ranging from 18 to 79 years). Of these participants, 60.9 % were women and 96.7 % were from white-European background. The mean BMI for all randomised participants was 25.5 kg m(-2), and 44.8 % of the participants had a BMI ≥ 25.0 kg m(-2). Food4Me is the first large multi-centre RCT of web-based PN. The main outcomes from the Food4Me study will be submitted for publication during 2015.

How I do it: Cone-beam CT during transarterial chemoembolization for liver cancer.

Abstract: In light of the substantial clinical evidence available in the literature, we recommend that cone-beam CT should be accepted as standard of care for imaging in intraarterial therapy of hepatocellular carcinoma.

Eu2+ luminescence in strontium aluminates

Abstract: The luminescence properties of Eu2+ doped strontium aluminates are reported and reviewed for a variety of aluminates, viz. SrAl12O19, SrAl4O7, Sr4Al14O25, SrAl2O4 and Sr3Al2O6. The aim of the research is to investigate the role of local coordination and covalency of the aluminate host lattice, related to the Sr/Al ratio, on the optical properties of the Eu2+ ion. The UV and VUV excited luminescence spectra as well as luminescence decay curves were recorded to characterize the luminescence properties of the investigated aluminates. The emission of Eu2+ ions varies over a wide spectral range, from ultraviolet (UV) to red, for the series of aluminates. The variation in emission color can be related to the crystal-field splitting of the 5d levels and the covalent interaction with the surrounding oxygen anions. In the least covalent material, viz. SrAl12O19: Eu2+, narrow line emission due to the P-6(7/2)-S-8(7/2) transition occurs at 4 K, indicating that the 4f(6)5d excited state is situated above the P-6(7/2)(4f(7)) excited state around 360 nm. The most alkaline material, viz. Sr3Al2O6: Eu2+ is the most covalent host and exhibits several d-f emission bands in the yellow to red spectral range due to the Eu2+ ions located on different crystallographic Sr2+ sites. The Eu2+ emission spectra in the other aluminates confirm the trend that with increasing Sr/Al ratio the Eu2+ emission shifts to longer wavelengths. Interesting differences are observed for the Eu2+ from different crystallographic sites which cannot always be related with apparent differences in the first oxygen coordination sphere. The discussion gives insight into how in a similar class of materials, strontium aluminates, the emission color of Eu2+ can be tuned over a wide spectral region.

Efficient water reduction with gallium phosphide nanowires

Abstract: Photoelectrochemical hydrogen production from solar energy and water offers a clean and sustainable fuel option for the future. Planar III/V material systems have shown the highest efficiencies, but are expensive. By moving to the nanowire regime the demand on material quantity is reduced, and new materials can be uncovered, such as wurtzite gallium phosphide, featuring a direct bandgap. This is one of the few materials combining large solar light absorption and (close to) ideal band-edge positions for full water splitting. Here we report the photoelectrochemical reduction of water, on a p-type wurtzite gallium phosphide nanowire photocathode. By modifying geometry to reduce electrical resistance and enhance optical absorption, and modifying the surface with a multistep platinum deposition, high current densities and open circuit potentials were achieved. Our results demonstrate the capabilities of this material, even when used in such low quantities, as in nanowires.

A multicentre comparison of quantitative 90Y PET/CT for dosimetric purposes after radioembolization with resin microspheres: the QUEST Phantom Study

Abstract: Purpose To investigate and compare the quantitative accuracy of 90Y imaging across different generation PET/CT scanners, for the purpose of dosimetry after radioembolization with resin microspheres.

Motion Robust Remote-PPG in Infrared

Abstract: Current state-of-the-art remote photoplethysmography (rPPG) algorithms are capable of extracting a clean pulse signal in ambient light conditions using a regular color camera, even when subjects move significantly. In this study, we investigate the feasibility of rPPG in the (near)-infrared spectrum, which broadens the scope of applications for rPPG. Two camera setups are investigated: one setup consisting of three monochrome cameras with different optical filters, and one setup consisting of a single RGB camera with a visible light blocking filter. Simulation results predict the monochrome setup to be more motion robust, but this simulation neglects parallax. To verify this, a challenging benchmark dataset consisting of 30 videos is created with various motion scenarios and skin tones. Experiments show that both camera setups are capable of accurate pulse extraction in all motion scenarios, with an average SNR of +6.45 and +7.26 dB, respectively. The single camera setup proves to be superior in scenarios involving scaling, likely due to parallax of the multicamera setup. To further improve motion robustness of the RGB camera, dedicated LED illumination with two distinct wavelengths is proposed and verified. This paper demonstrates that accurate rPPG measurements in infrared are feasible, even with severe subject motion.

Asymmetric magnetic bubble expansion under in-plane field in Pt/Co/Pt: Effect of interface engineering

Abstract: We analyze the impact of growth conditions on the asymmetric magnetic bubble expansion under an in-plane field in ultrathin Pt/Co/Pt films. Specifically, using sputter deposition, we vary the Ar pressure during the growth of the top Pt layer. This induces a large change in the interfacial structure as evidenced by a factor three change in the effective perpendicular magnetic anisotropy. Strikingly, a discrepancy between the current theory for domain-wall propagation based on a simple domain-wall energy density and our experimental results is found. This calls for further theoretical development of domain-wall creep under in-plane fields and varying structural asymmetry.

Kernel Interpolation for Scalable Structured Gaussian Processes (KISS-GP)

Abstract: We introduce a new structured kernel interpolation (SKI) framework, which generalises and unifies inducing point methods for scalable Gaussian processes (GPs). SKI methods produce kernel approximations for fast computations through kernel interpolation. The SKI framework clarifies how the quality of an inducing point approach depends on the number of inducing (aka interpolation) points, interpolation strategy, and GP covariance kernel. SKI also provides a mechanism to create new scalable kernel methods, through choosing different kernel interpolation strategies. Using SKI, with local cubic kernel interpolation, we introduce KISSGP, which is 1) more scalable than inducing point alternatives, 2) naturally enables Kronecker and Toeplitz algebra for substantial additional gains in scalability, without requiring any grid data, and 3) can be used for fast and expressive kernel learning. KISS-GP costs O(n) time and storage for GP inference. We evaluate KISS-GP for kernel matrix approximation, kernel learning, and natural sound modelling.