A pilot study of resistance to aspirin in stroke patients

Aspirin resistance has been shown to be a significant risk factor for recurrent cardiovascular ischaemic events. However, there are a lack of data correlating aspirin resistance and risk of cerebrovascular ischaemic events. This pilot study aimed to determine the prevalence of aspirin resistance in an Australian stroke population and to correlate aspirin resistance with an increased risk of ischaemic stroke. Fifty patients treated with aspirin for 2 years were tested for aspirin resistance using the Ultegra Rapid Platelet Function Assay (Accumetrics, San Diego, CA, USA) on admission to Royal Melbourne Hospital for ischaemic stroke. The 2-year history of ischaemic stroke and transient ischaemic attack (TIA) were assessed. Prevalence of aspirin resistance among our patients was 30%. Univariate analysis suggested a non-significant trend towards increased rate of previous ischaemic stroke or TIA and aspirin resistance (odds ratio, OR=3.88; 95% confidence interval 0.54-29.87; p=0.18). This study shows that aspirin resistance is prevalent within the Australian ischaemic stroke population.     

Mediating role of the default mode network on parental acceptance/warmth and psychopathology in youth

Humans are reliant on their caregivers for an extended period of time, offering numerous opportunities for environmental factors, such as parental attitudes and behaviors, to impact brain development. The default mode network is a neural system encompassing the medial prefrontal cortex, posterior cingulate cortex, precuneus, and temporo-parietal junction, which is implicated in aspects of cognition and psychopathology. Delayed default mode network maturation in children and adolescents has been associated with greater general dimensional psychopathology, and positive parenting behaviors have been suggested to serve as protective mechanisms against atypical default mode network development. The current study aimed to extend the existing research by examining whether within- default mode network resting-state functional connectivity would mediate the relation between parental acceptance/warmth and youth psychopathology. Data from the Adolescent Brain and Cognitive Development study, which included a community sample of 9,366 children ages 8.9–10.9 years, were analyzed to test this prediction. Results demonstrated a significant mediation, where greater parental acceptance/warmth predicted greater within- default mode network resting-state functional connectivity, which in turn predicted lower externalizing, but not internalizing symptoms, at baseline and 1-year later. Our study provides preliminary support for the notion that positive parenting behaviors may reduce the risk for psychopathology in youth through their influence on the default mode network.

     

Catamenial epilepsy occurrence and patterns in a mixed population of women with epilepsy

We evaluated the occurrence and distribution of patterns of catamenial epilepsy in a heterogenous cohort of women with epilepsy on no hormonal therapies, enrolled in a prospective, observational study. The primary aim of the study was pregnancy rate in women with epilepsy with no prior reproductive problems. In this analysis, we included women who recorded one or more menstrual cycles with one or more seizures. We measured progesterone concentrations for one to three cycles. We defined catamenial patterns as twofold or greater average daily seizure frequency around menstruation (C1), ovulation (C2), and for anovulatory cycles, from midcycle through menstruation (C3). Twenty-three of the 89 enrolled women with epilepsy were eligible for this analysis; 12 of 23 met criteria for catamenial epilepsy; five of 23 demonstrated only a C1 pattern, two of 23 only a C2 pattern, five of 23 a combined C1/C2 pattern, and the one woman with anovulatory cycles did not demonstrate a C3 pattern. There were no differences in likelihood of demonstrating a catamenial pattern between those who reported a prior catamenial pattern and those who did not (p = .855). This analysis demonstrates the utility of app-based tracking to determine a catamenial pattern. Larger prospective studies could confirm these findings and inform potential therapeutic trial designs for catamenial epilepsy.     

Mental health in Syrian children with a focus on post-traumatic stress: a cross-sectional study from Syrian schools

Purpose

Studies show that conflict can negatively affect psychological health. The Syrian crisis is 8 years old and yet little is known about the impact of the conflict on the well-being of Syrians who remain. This gap was addressed by conducting an empirical study on the mental health burden of Syrian children in two areas of the country.

Methods

492 children between 8 and 15 years were randomly selected from schools in Damascus and Latakia. The incidence of psychological disorder symptoms was measured using self-report screening instruments, the Children’s Revised Impact of Event Scale (CRIES-8) and the Revised Children’s Anxiety and Depression Scale (RCADS-25). Simultaneously, sociodemographic and traumatic event information was collected. Binary logistic regression was used to identify factors that influence the development of post-traumatic stress disorder (PTSD) symptoms.

Results

In our sample, 50.2% of students were internally displaced and 32.1% reported a negative experience. 60.5% of those tested had at least one probable psychological disorder with PTSD the most common (35.1%), followed by depression (32.0%), and anxiety (29.5%). Binary logistic regression indicated that PTSD symptoms were predicted by: living in Damascus [odds ratio (OR) 2.36, 95% confidence interval (CI) 1.51–3.69], being female (1.54, 1.02–2.34), having depression and anxiety (2.55, 1.48–4.40), and the negative experiences; displacement and daily warzone exposure (1.84, 1.02–3.30 and 2.67, 1.08–6.60).

Conclusions

Syrian children are experiencing traumatic events and war-associated daily stresses that are hugely impacting psychological well-being. Our data offer guidance for mental health providers regarding risk factors and highlights the use of the school system to reach suffering children.

     

Impact of Implementing the 2013 ACC/AHA Cholesterol Guidelines on Vascular Events in a Statewide Community‐Based Practice Registry

Electronic health record data were analyzed to estimate the number of statin‐eligible adults with the 2013 American College of Cardiology/American Heart Association cholesterol guidelines not taking statin therapy and the impact of recommended statin therapy on 10‐year atherosclerotic cardiovascular disease (ASCVD₁₀) events. Adults aged 21 to 80 years in an outpatient network with ≥1 clinic visit(s) from January 2011 to June 2014 with data to calculate ASCVD₁₀ were eligible. Moderate‐intensity statin therapy was assumed to lower low‐density lipoprotein cholesterol by 30% and high‐intensity therapy was assumed to reduce low‐density lipoprotein cholesterol by 50%. ASCVD events were assumed to decline 22% for each 39 mg/dL decline in low‐density lipoprotein cholesterol. Among 411,768 adults, 260,434 (63.2%) were not taking statins and 103,478 (39.7%) were eligible for a statin, including 79,069 (76.4%) patients with hypertension. Estimated ASCVD₁₀ events were 18,781 without and 13,328 with statin therapy, a 29.0% relative and 5.3% absolute risk reduction with a number needed to treat of 19. The 2013 cholesterol guidelines are a relatively efficient approach to reducing ASCVD in untreated, statin‐eligible adults who often have concomitant hypertension.     

From the Cover: Convolutional networks for fast,energy-efficient neuromorphic computing

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware’s underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer.The human brain is capable of remarkable acts of perception while consuming very little energy. The dream of brain-inspired computing is to build machines that do the same, requiring high-accuracy algorithms and efficient hardware to run those algorithms. On the algorithm front, building on classic work on backpropagation (1), the neocognitron (2), and convolutional networks (3), deep learning has made great strides in achieving human-level performance on a wide range of recognition tasks (4). On the hardware front, building on foundational work on silicon neural systems (5), neuromorphic computing, using novel architectural primitives, has recently demonstrated hardware capable of running 1 million neurons and 256 million synapses for extremely low power (just 70 mW at real-time operation) (6). Bringing these approaches together holds the promise of a new generation of embedded, real-time systems, but first requires reconciling key differences in the structure and operation between contemporary algorithms and hardware. Here, we introduce and demonstrate an approach we call Eedn, energy-efficient deep neuromorphic networks, which creates convolutional networks whose connections, neurons, and weights have been adapted to run inference tasks on neuromorphic hardware.For structure, typical convolutional networks place no constraints on filter sizes, whereas neuromorphic systems can take advantage of blockwise connectivity that limits filter sizes, thereby saving energy because weights can now be stored in local on-chip memory within dedicated neural cores. Here, we present a convolutional network structure that naturally maps to the efficient connection primitives used in contemporary neuromorphic systems. We enforce this connectivity constraint by partitioning filters into multiple groups and yet maintain network integration by interspersing layers whose filter support region is able to cover incoming features from many groups by using a small topographic size (7).For operation, contemporary convolutional networks typically use high precision ( ≥ 32-bit) neurons and synapses to provide continuous derivatives and support small incremental changes to network state, both formally required for backpropagation-based gradient learning. In comparison, neuromorphic designs can use one-bit spikes to provide event-based computation and communication (consuming energy only when necessary) and can use low-precision synapses to colocate memory with computation (keeping data movement local and avoiding off-chip memory bottlenecks). Here, we demonstrate that by introducing two constraints into the learning rule—binary-valued neurons with approximate derivatives and trinary-valued ({−1,0,1}) synapses—it is possible to adapt backpropagation to create networks directly implementable using energy efficient neuromorphic dynamics. This approach draws inspiration from the spiking neurons and low-precision synapses of the brain (8) and builds on work showing that deep learning can create networks with constrained connectivity (9), low-precision synapses (10, 11), low-precision neurons (12–14), or both low-precision synapses and neurons (15, 16). For input data, we use a first layer to transform multivalued, multichannel input into binary channels using convolution filters that are learned via backpropagation (12, 16) and whose output can be sent on chip in the form of spikes. These binary channels, intuitively akin to independent components (17) learned with supervision, provide a parallel distributed representation to carry out high-fidelity computation without the need for high-precision representation.Critically, we demonstrate that bringing the above innovations together allows us to create networks that approach state-of-the-art accuracy performing inference on eight standard datasets, running on a neuromorphic chip at between 1,200 and 2,600 frames/s (FPS), using between 25 and 275 mW. We further explore how our approach scales by simulating multichip configurations. Ease-of-use is achieved using training tools built from existing, optimized deep learning frameworks (18), with learned parameters mapped to hardware using a high-level deployment language (19). Although we choose the IBM TrueNorth chip (6) for our example deployment platform, the essence of our constructions can apply to other emerging neuromorphic approaches (20–23) and may lead to new architectures that incorporate deep learning and efficient hardware primitives from the ground up.     

Parallel analysis of genetic selections using whole genome oligonucleotide arrays

Thousands of genes have recently been sequenced in organisms ranging from Escherichia coli to human. For the majority of these genes, however, available sequence does not define a biological role. Efficient functional characterization of these genes requires strategies for scaling genetic analyses to the whole genome level. Plasmid-based library selections are an established approach to the functional analysis of uncharacterized genes and can help elucidate biological function by identifying, for example, physical interactors for a gene and genetic enhancers and suppressors of mutant phenotypes. The application of these selections to every gene in a eukaryotic genome, however, is generally limited by the need to manipulate and sequence hundreds of DNA plasmids. We present an alternative approach in which identification of nucleic acids is accomplished by direct hybridization to high-density oligonucleotide arrays. Based on the complete sequence of Saccharomyces cerevisiae, high-density arrays containing oligonucleotides complementary to every gene in the yeast genome have been designed and synthesized. Two-hybrid protein–protein interaction screens were carried out for S. cerevisiae genes implicated in mRNA splicing and microtubule assembly. Hybridization of labeled DNA derived from positive clones is sufficient to characterize the results of a screen in a single experiment, allowing rapid determination of both established and previously unknown biological interactions. These results demonstrate the use of oligonucleotide arrays for the analysis of two-hybrid screens. This approach should be generally applicable to the analysis of a range of genetic selections.