Search

Malik A Althobiani,1,2 Bilal Khan,1 Amar J Shah,1,3 Yatharth Ranjan,4 Renata G Mendes,1,5 Amos Folarin,4 Swapna Mandal,1,3 Joanna C Porter,1,6 John R Hurst1,3 1UCL Respiratory, University College London, London, UK; 2Department of Respiratory Therapy, Faculty of Medical Rehabilitation Sciences, King Abdulaziz University, Jeddah, Saudi Arabia; 3Department of Respiratory Medicine, Royal Free London NHS Foundation Trust, London, UK; 4Department of Health Informatics and Biostatistics, Institute of Psychiatry, Psychology and Neuroscience, Kings College London, London, UK; 5Department of Physical Therapy, Cardiopulmonary Physiotherapy Laboratory, Federal University of São Carlos, São Paulo, Brazil; 6Department of Respiratory Medicine, University College London Hospital (UCLH), London, UKCorrespondence: Malik A Althobiani, UCL Respiratory, University College London, London, UK, Email malik.althobiani.20@ucl.ac.ukObjective: To investigate clinicians’ perspectives on the current use of wearable technology for detecting COPD exacerbations, and to identify potential facilitators and barriers to its adoption in clinical settings.Methods: A mixed-method survey was conducted through an online survey platform involving clinicians working with COPD patients. The questionnaires were developed by an expert panel specialising in respiratory medicine at UCL. The questionnaire evaluated clinicians’ perspectives on several aspects: the current extent of wearable technology utilisation, the perceived feasibility, and utility of these devices, as well as the potential facilitators and barriers that hinder its wider implementation.Results: Data from 118 clinicians were included in the analysis. Approximately 80% of clinicians did not currently use information from wearable devices in routine clinical care. A majority of clinicians did not have confidence in the effectiveness of wearables and their consequent impact on health outcomes. However, clinicians highlighted the potential value of wearables in helping deliver personalised care and more rapid assistance. Ease of use, technical support and accessibility of data were considered facilitating factors for wearable utilisation. Costs and lack of technical knowledge were the most frequently reported barriers to wearable utilisation.Conclusion: Clinicians’ perspectives of the use of wearable technology to detect and monitor COPD exacerbations are variable. While accessibility and technical support facilitate wearable implementation, cost, technical issues, and knowledge act as barriers. Our findings highlight the facilitators and barriers to using wearables in patients with COPD and emphasise the need to assess patients’ perspectives on wearable acceptability.Keywords: wearable technology, COPD exacerbations, clinicians’ perspectives

Many processes in nature such as conformal changes in biomolecules and clusters of interacting particles, genetic switches, noisy mechanical or electromechanical oscillators, and many others are modeled using stochastic differential equations with small white noise. The study of rare transitions between metastable states in such systems is of great interest and importance, but direct simulations are difficult due to long waiting times. Transition path theory is a mathematical framework for the quantitative description of rare events. Its direct implementation the key component of which is the solution of the committor problem, a boundary value problem for the backward Kolmogorov equation, is often challenging due to high dimensionality or other numerical issues. This work exploits the key fact that the optimal controller constructed from the committor leads to generation of transition trajectories exclusively.

Additional file 1. Tool for risk and needs assessment (RANA).

Maximal safe resection can improve patient outcomes for a variety of brain tumor types including low- and high-grade gliomas, pituitary adenomas, and other pathologies. Numerous intraoperative adjuncts exist to guide surgeons with maximizing extent of resection. Three distinct strategies exist including: 1) surgical navigation; 2) intraoperative imaging; and 3) tumor fluorescence. Surgical navigation involves registration of high-resolution three-dimensional imaging to the patient's cranial surface anatomy, allowing real-time localization of tumor and brain structures. Intraoperative imaging devices like intraoperative magnetic resonance imaging (iMRI), intraoperative computed tomography (iCT), 3-D fluoroscopy, and intraoperative ultrasonography (iUS) allow near real time visualization to assess the extent of resection. Intraoperative fluorescence via intravenous fluorescein or oral 5-aminolevulinic acid (5-ALA) causes brain tumors to "light up", which can be viewed through surgical optics using selective filters and specific wavelength light sources. A general overview, as well as implementation and utilization of some of these image guidance strategies at Washington University and by Siteman Cancer Center neurosurgeons at Barnes Jewish Hospital, is discussed in this review.

Additional file 2. Prevention Care and Support Card (PCS).

Optimisation is integral to all sorts of processes in science, economics and arguably underpins the fruition of human intelligence through millions of years of optimisation, or $\textit{evolution}$. Scarce resources make it crucial to maximise their efficient usage. In this thesis, we consider the task of maximising unknown functions which we are able to query point-wise. The function is deemed to be $\textit{costly}$ to evaluate e.g. larger run time or financial expense, requiring a judicious querying strategy given previous observations. We adopt a probabilistic framework for modelling the unknown function and Bayesian non-parametric modelling. In particular, we focus on the $\textit{Gaussian process}$ (GP), a popular non-parametric Bayesian prior on functions. We motivate these choices and give an overview of the Gaussian process in the introduction, and its application to $\textit{Bayesian optimisation}$. A GP's behaviour is intimately controlled by the choice of $\textit{kernel}$ or covariance function, typically chosen to be a parametric function. In chapter 2 we instead place a non-parametric Bayesian prior, known as an Inverse Wishart process prior, over a GP kernel function, and show that this may be marginalised analytically leading to a $\textit{Student-$t$ process}$ (TP). Furthermore we explore a larger class of $\textit{elliptical processes}$, and show that the TP is the most general for which analytic calculation is possible, and apply it successfully for Bayesian optimisation. The remainder of the thesis focusses on various Bayesian optimisation settings. In chapter 3, we consider a setting where we are able to evaluate a function at multiple locations in parallel. Our approach is to consider a measure of information, $\textit{entropy}$, to decide which batch of points to evaluate a function at next. We similarly apply information gain for $\textit{multi-objective}$ Bayesian optimisation in chapter 4. Here, one wishes to find a $\textit{Pareto frontier}$ of efficient settings with respect to several different objectives through sequential evaluation. Finally, in chapter 5 we exploit the idea that in a multi-objective setting, functions are $\textit{correlated}$, incorporating this belief in our choice of prior distribution over the multiple objectives.

Context: A subpopulation of hypothalamic neurons colocalize three neuropeptides, namely kisspeptin, neurokinin B (NKB), and dynorphin, collectively termed KNDy neurons. Animal studies suggest they interact to affect pulsatile GnRH release (KNDy hypothesis); kisspeptin stimulates, NKB modulates, and dynorphin (an opioid) inhibits. Objective: To investigate the KNDy hypothesis in humans, we assessed for the first time the effects of the coadministration of kisspeptin-54, NKB, and an opioid receptor antagonist, naltrexone, on LH pulsatility (surrogate marker for GnRH pulsatility) and gonadotropin release. Design, Setting, and Participants: This was an ethically approved prospective, single-blinded, placebo-controlled study. Healthy male volunteers (n = 5/group) attended our research facility for eight study visits. Intervention and Main Outcome Measure: After 1 hour of baseline blood sampling, participants received a different intervention at each visit: oral 50 mg naltrexone, 8-hour iv infusions of vehicle, 2.56 nmol/kg · h NKB, 0.1 nmol/kg · h kissspeptin-54 (KP) alone and in combination. Frequent blood sampling to measure plasma gonadotropins and sex steroids was conducted and LH pulsatility was determined using blinded deconvolution analysis. Results: All kisspeptin and naltrexone containing groups potently increased LH and LH pulsatility (P < .001 vs vehicle). NKB alone did not affect gonadotropins. NKB+KP had significantly lower increases in gonadotropins compared with kisspeptin alone (P < .01). Naltrexone+KP was the only group to significantly increase LH pulse amplitude (P < .001 vs vehicle). Conclusions: Our results suggest significant interactions between the KNDy neuropeptides on LH pulsatility and gonadotropin release in humans. This has important implications for improving our understanding of GnRH pulse generation in humans., Co-administration of kisspeptin-54, NKB and an opioid antagonist naltrexone result in significant interactions to modulate LH pulsatility and gonadotrophin release in humans.

We develop parallel predictive entropy search (PPES), a novel algorithm for Bayesian optimization of expensive black-box objective functions. At each iteration, PPES aims to select a batch of points which will maximize the information gain about the global maximizer of the objective. Well known strategies exist for suggesting a single evaluation point based on previous observations, while far fewer are known for selecting batches of points to evaluate in parallel. The few batch selection schemes that have been studied all resort to greedy methods to compute an optimal batch. To the best of our knowledge, PPES is the first non-greedy batch Bayesian optimization strategy. We demonstrate the benefit of this approach in optimization performance on both synthetic and real world applications, including problems in machine learning, rocket science and robotics., 12 pages in Neural Information Processing Systems 2015

Recurrent neural networks (RNNs) are notoriously difficult to train. When the eigenvalues of the hidden to hidden weight matrix deviate from absolute value 1, optimization becomes difficult due to the well studied issue of vanishing and exploding gradients, especially when trying to learn long-term dependencies. To circumvent this problem, we propose a new architecture that learns a unitary weight matrix, with eigenvalues of absolute value exactly 1. The challenge we address is that of parametrizing unitary matrices in a way that does not require expensive computations (such as eigendecomposition) after each weight update. We construct an expressive unitary weight matrix by composing several structured matrices that act as building blocks with parameters to be learned. Optimization with this parameterization becomes feasible only when considering hidden states in the complex domain. We demonstrate the potential of this architecture by achieving state of the art results in several hard tasks involving very long-term dependencies.

We present PESMO, a Bayesian method for identifying the Pareto set of multi-objective optimization problems, when the functions are expensive to evaluate. The central idea of PESMO is to choose evaluation points so as to maximally reduce the entropy of the posterior distribution over the Pareto set. Critically, the PESMO multi-objective acquisition function can be decomposed as a sum of objective-specific acquisition functions, which enables the algorithm to be used in \emph{decoupled} scenarios in which the objectives can be evaluated separately and perhaps with different costs. This decoupling capability also makes it possible to identify difficult objectives that require more evaluations. PESMO also offers gains in efficiency, as its cost scales linearly with the number of objectives, in comparison to the exponential cost of other methods. We compare PESMO with other related methods for multi-objective Bayesian optimization on synthetic and real-world problems. The results show that PESMO produces better recommendations with a smaller number of evaluations of the objectives, and that a decoupled evaluation can lead to improvements in performance, particularly when the number of objectives is large.

Semi-supervised clustering is the task of clustering data points into clusters where only a fraction of the points are labelled. The true number of clusters in the data is often unknown and most models require this parameter as an input. Dirichlet process mixture models are appealing as they can infer the number of clusters from the data. However, these models do not deal with high dimensional data well and can encounter difficulties in inference. We present a novel nonparameteric Bayesian kernel based method to cluster data points without the need to prespecify the number of clusters or to model complicated densities from which data points are assumed to be generated from. The key insight is to use determinants of submatrices of a kernel matrix as a measure of how close together a set of points are. We explore some theoretical properties of the model and derive a natural Gibbs based algorithm with MCMC hyperparameter learning. The model is implemented on a variety of synthetic and real world data sets., Appears in Proceedings of the Twenty-Ninth Conference on Uncertainty in Artificial Intelligence (UAI2013)

Laparoscopic repair of Morgagni hernia has been described in adults and children. In the published reports, the crux of the repair consists of suturing the posterior part of the diaphragmatic defect to the undersurface of the sternum or the posterior rectus sheath. The tissue on the undersurface of the sternum is variable is in its nature and may be inadequate for suturing, hence compromising the strength of the repair. A technique that circumvents this problem and offers a strong anatomical repair is described. A Morgagni hernia was diagnosed in a 2-year-old girl with trisomy 21, who presented with recurrent chest infections. She underwent laparoscopic repair of the hernia using three ports. The tissue on the undersurface of the sternum was inadequate for a conventional repair. The procedure was modified as follows: a small transverse incision was made over the lower end of the sternum. Three nonabsorbable mattress sutures were inserted through the sternum, the anterior edge of the diaphragmatic defect, and back through the sternum and tied with extracorporeal knots. The child was discharged home on the second postoperative day. At 6-month follow up, the child was asymptomatic, and had been infection free. A chest radiograph was normal. This is a simple, novel, noninvasive method, which offers a secure anatomical repair and it is not dependent on the adequacy of the tissue on the undersurface of the sternum.