Search

Introduction La lobomycose est une infection fongique chronique atteignant la peau et les tissus sous-cutanes. Rare (moins de 600 cas publies), elle est presente majoritairement en Amazonie. Le dauphin a longtemps ete considere comme reservoir mais la biologie moleculaire a recemment ecarte cette hypothese. Le mode de contamination precis n’est pas connu mais des expositions aquatiques sont frequemment mises en evidence. La forme clinique la plus commune est celle de nodules pseudo-cheloidiens d’evolution progressive, bien qu’une forme a type de plaques multicentriques soit decrite chez les Indiens Kaiabi au Bresil. L’agent fongique, Lacazia loboi, est identifiable sous forme de levures en chainettes a l’examen direct ou en histologie. Une augmentation des cas avait ete suggeree lors d’une precedente etude en Guyane (20 cas en 50 ans). Nous rapportons ici les cas de lobomycose diagnostiques dans ce territoire depuis 2010 en s’interessant aux caracteristiques epidemiologiques et cliniques. Materiel et methodes Il s’agit d’une etude retrospective conduite entre 2010 et 2020 au centre hospitalier de Cayenne. Le diagnostic etait evoque cliniquement et confirme par l’examen direct ou l’histologie. Resultats Dix patients etaient recenses, essentiellement des hommes (70 %) et originaires du Bresil (80 %), plus rarement creoles ou metropolitains (20 %). Une seule patiente rapportait un traumatisme a l’origine de la lesion (morsure de poisson). La majorite (80 %) des patients presentait un facteur d’exposition a l’eau ou a la foret, souvent professionnel (50 % d’orpailleurs). La contamination avait lieu en Guyane dans 60 % des cas. La duree d’evolution moyenne avant le diagnostic etait de 10 ans. La majorite (90 %) des patients avait des nodules pseudo-cheloidiens, localises sur les membres (70 %), le visage (30 %) ou le tronc (20 %). Aucune plaque multicentrique n’etait retrouvee. L’atteinte etait plurifocale dans 60 % des cas, plus rarement ganglionnaire (20 %). L’examen direct etait positif dans 100 % des cas. Neuf patients ont ete traites, 7 par traitement medicamenteux (itraconazole, terbinafine, dapsone, clofazimine), 3 par chirurgie. Seule une patiente a gueri apres exerese lesionnelle. La plupart des patients ont ete perdus de vue en quelques mois. Discussion L’absence de plaque multicentrique ou de cas chez des Amerindiens constitue une difference interessante par rapport aux donnees bresiliennes. L’essentiel des patients est originaire du Bresil mais semble s’etre contamine en Guyane. La chirurgie reste le traitement le plus efficace mais ne s’envisage qu’en cas de lesion limitee. Classee maladie tropicale negligee depuis 2017, cette mycose merite davantage de recherche therapeutique. L’incidence (1 cas/an) semble en augmentation par rapport a la precedente etude guyanaise.

Objectives: The Scottish Medicines Consortium (SMC) is the national HTA agency for medicines. Due to the ongoing pandemic, all multi-stakeholder meetings moved to a virtual setting. This study evaluated user satisfaction, captured any technical issues, and identified what worked well and areas for improvement in order to establish stakeholder confidence in the ongoing use of virtual meetings. Methods: A SmartSurvey consisting of 13 questions was distributed to participants of the three main committees of SMC: the New Drugs Committee (NDC);Patient and Clinician Engagement (PACE);and SMC committee and covered meetings held June-October 2020. Meetings were held using a combination of MS Teams and Zoom videoconferencing. The choice of platform reflected availability of wider functionality for example, break out rooms for private voting by ballot. In addition to SMC staff, meeting attendees included committee members (NDC, SMC), patient group representatives/advocates (PACE, SMC), pharmaceutical company representatives (SMC), and clinical experts (PACE). Results: 155 respondents completed the survey including attendees of NDC (n=47), PACE (n=15), and SMC meetings (n=93). 98% of respondents for NDC, 100% for PACE, and 89% for SMC rated their experience of attending virtually as “very good” or “good”. Most respondents found the technology easy to use (98%, 93%, and 95% respectively). Areas for improvement included: duration and comfort breaks;when to have cameras on;consistency of presentation slides;and discussion or social setting. Implementation of suggested improvements is ongoing. Reduced travel was one of the most valued aspects: “What … worked well is that all interested parties could attend, without geography/travel being a potential barrier” (Patient group representative/advocate). Conclusions: Across all committees, user satisfaction with virtual meetings was high and technical issues were minimal. Virtual meetings have important implications for the accessibility of HTA decision-making, beyond the pandemic.

Many studies have highlighted concerns about the completeness and quality of information found online and how this may affect patients' education about their medical problems. One aspect of internet usage that has received less attention in the literature, however, is patient perception of the information that is gathered online, and how patients use it related to their musculoskeletal care.The objective of the study is to utilize a cross-sectional study design to describe internet usage and patient perceptions of orthopedic online information and to identify differences in usage patterns.One thousand two hundred ninety-six questionnaires were distributed to consecutive patients at orthopedic outpatient clinics which consisted of questions pertaining to patients' internet use. Basic demographic data were collected, and subgroup analyses were performed to examine the effect of three variables (age, gender, and clinic type) on various outcomes.84.9% of patients reported access to the internet. Of patients with internet access, 64.7% reported using the internet for obtaining orthopedic information. 43.1% of the respondents who searched for orthopedic information rated it as "very useful," 56.3% found it "somewhat useful," and 0.6% found it "not at all useful". Younger patients were more likely to have used the internet for health and orthopedic information and to have found this information either very or somewhat useful. Males were more likely to have found the internet information very useful. Overall, only 33.7% of patients who researched their current orthopedic complaint accessed the institutional website for information.A large proportion of patients use the internet to research orthopedic information and most patients, especially younger males, find the information useful.

Dispersal and local environmental factors are major determinants of invasive species distribution. We examined how both dispersal-related geospatial characteristics and environmental factors influence an ongoing invasion of wetlands in a south-central Florida ranchland by non-native apple snails (Pomacea maculata, Ampullariidae). We found P. maculata in 73 (43%) of a random set of 171 wetlands in 2014. We used model selection to evaluate multiple hypotheses of predictors of P. maculata occurrence in 95 wetlands with standing water, including spatially-explicit distances in ditches from wetlands to the presumed entry point, Euclidean (overland) distances, presence/absence of ditches in wetlands, and environmental variables (e.g. pH). We also performed a 5-month field experiment in 20 wetlands to evaluate if snail absence was associated with conditions that limit survival and growth (i.e. unfavorable habitats). Snail occurrence was primarily associated with presence of ditches in wetlands and more neutral wetland pH. These variables more plausibly explained snail occurrence than did Euclidean (overland) distance and minimum ditch travel (rectilinear) distance from propagule sources (a major waterway). Wetland pH best explained survival and growth under the experimental conditions. We found no evidence that prior occupancy by conspecifics affected survival and growth, suggesting that dispersal limitation may contribute to lack of occupancy of wetlands, despite suitable pH. Our study supports man-made conduits as facilitators of dispersal by non-native species, where environmental characteristics (here pH) then also affect colonization within habitats. An understanding of both dispersal mechanisms and local environmental factors is necessary to better predict invasive species distribution.

Resume Introduction En France, 26 equipes regionales ressources de soins palliatifs pediatriques (ERRSPP) ont ete creees entre 2008 et 2012. Nous avons conduit la premiere etude prospective francaise visant a decrire les caracteristiques du recours initial a une ERRSPP et analyser les reponses apportees par cette equipe. Materiel et methodes Toutes les demandes d’interventions formulees a l’ERRSPP entre septembre 2013 et septembre 2014 ont ete recueillies de facon prospective a l’aide d’un questionnaire comportant : les elements demographiques et cliniques du patient (âge, sexe, pathologie), les caracteristiques de la demande (origine du demandeur, specialite du service, motif de la demande) et de la reponse apportee par l’ERRSPP (intervalle diagnostic initial–demande, duree de la prise en charge, deces eventuel, modifications de la demande initiale). Les pathologies ont ete classees dans l’un des 6 groupes de maladies de soins palliatifs pediatriques (SPP), selon les normes edictees par l’Association canadienne de soins palliatifs. Resultats Nous avons recense 67 demandes, concernant 61 patients (37 garcons). L’âge median etait de 49,8 mois [2,3–145]. Les motifs de demande ont ete multiples : reunion ethique pluridisciplinaire (42 %), coordination (34 %), aide a la prise en charge symptomatique (21 %), aide au domicile (19 %), soutien psychologique (15 %), enseignement (9 %) et discussion autour d’un cas (6 %). La demande d’une reunion ethique a ete predominante pour la periode neonatale (61 %) alors que la coordination de la prise en charge l’a ete pour les enfants et adolescents (78 % au-dela de 4 ans). Le delai median « diagnostic–demande » rapporte aux groupes de maladies a ete court pour le groupe 5 (neonatalogie : 0,36 mois) et long pour le groupe 4 (maladies irreversibles et non evolutives : 54,6 mois) (p Conclusion Cette etude montre l’heterogeneite du premier recours a une equipe de SPP, validant ainsi les differentes missions devolues aux ERRSPP et la necessite d’une equipe pluridisciplinaire. La reponse de l’ERRSPP a ete elargie dans la moitie des cas, demontrant une modulation des besoins dans le temps.

The XXL survey currently covers two 25 sq. deg. patches with XMM observations of ~10ks. We summarise the scientific results associated with the first release of the XXL data set, that occurred mid 2016. We review several arguments for increasing the survey depth to 40 ks during the next decade of XMM operations. X-ray (z 1 cluster density. It will eventually constitute a reference study and an ideal calibration field for the upcoming eROSITA and Euclid missions.

In this paper, a novel BCD multiplier approach is proposed. The main highlight of the proposed architecture is the generation of the partial products and parallel binary operations based on 2-digit columns. 1 × 1-digit multipliers used for the partial product generation are implemented directly by 4-bit binary multipliers without any code conversion. The binary results of the 1 × 1-digit multiplications are organized according to their two-digit positions to generate the 2-digit column-based partial products. A binary-decimal compressor structure is developed and used for partial product reduction. These reduced partial products are added in optimized 6-LUT BCD adders. The parallel binary operations and the improved BCD addition result in improved performance and reduced resource usage. The proposed approach was implemented on Xilinx Virtex-5 and Virtex-6 FPGAs with emphasis on the critical path delay reduction. Pipelined BCD multipliers were implemented for 4 × 4, 8 × 8, and 16 × 16-digit multipliers. Our realizations achieve an increase in speed by up to 22% and a reduction of LUT count by up to 14% over previously reported results.

Convolutional Neural Networks (CNNs) have achieved significant success in image classification. One of the main reasons that CNNs achieve state-of-the-art accuracy is using many multi-scale learnable windowed feature detectors called kernels. Fetching of kernel feature weights from memory and performing the associated multiply and accumulate computations consume massive amount of energy. This hinders the widespread usage of CNNs, especially in embedded devices. In comparison with CNNs, decision forests are computationally efficient since they are composed of decision trees, which are binary classifiers by nature and can be implemented using AND-OR gates instead of costly multiply and accumulate units. In this paper, we investigate the migration of CNNs to decision forests as one of the promising approaches for reducing both execution time and power consumption while achieving acceptable accuracy. We introduce POLYCiNN, an architecture composed of a stack of decision forests. Each decision forest classifies one of the overlapped sub-images of the original image. Then, all decision forest classifications are fused together to classify the input image. In POLYCiNN, each decision tree is implemented in a single 6-input Look-Up Table and requires no memory access. Therefore, POLYCiNN can be efficiently mapped to simple and densely parallel hardware designs. We validate the performance of POLYCiNN on the benchmark image classification tasks of the MNIST, CIFAR-10 and SVHN datasets.

The gold standard for the detection of malingered psychosis involves expert clinical assessment augmented by standardized psychometric testing. The evaluation of malingered auditory verbal hallucinations is complicated, however, by increasing evidence that voice-hearing is a broadly heterogeneous experience that does not always reflect psychopathology, with atypical features nearly as common as typical characteristics. The detection of malingered auditory verbal hallucinations in clinical settings may be particularly vulnerable to false positives and false negatives due to low suspicion on the part of clinicians, low utilization and poor specificity of psychometric testing, and "iatrogenic malingering" that is less likely to include cartoonish claims and more likely to involve voice-hearing as a sole presenting symptom (i.e., monosymptomatic auditory verbal hallucinations). In both clinical and forensic settings, the detection of malingered auditory verbal hallucinations requires detailed exploration of phenomenologic features along with mediating factors that influence the risk of associated violence or suicide.

Nous nous proposons de répondre à la question suivante : « Y eut-il sous Charles le Téméraire une seule ligne de conduite vis-à-vis de la principauté de Liège ? » Notre réponse à cette question distingue deux politiques bourguignonnes. Des documents clés du début du règne, des deux paix datées respectivement du 28 décembre 1465 et du 18 novembre 1467, nous dégageons les composantes de la politique bourguignonne au début du règne. Nous les comparons au langage des faits de la fin du règne. Nou...

Introduction Attention deficit hyperactivity disorder (ADHD) is a frequent neurodevelopmental mental disorder. It can persist in adulthood and be expressed as a cognitive complaint. Methods We conducted a descriptive study in a French memory center concerning patients seen over a period of two years. All patients for whom the final diagnosis was ADHD were included. All patients benefited from standard neuropsychological tests and a psychiatric specific consultation. Results Thirteen patients were included with an average age of 50.2 ± 19 years. Main complaints related to memory, attention, focusing and organizational functioning. These difficulties had negative social, professional and academic consequences. ADHD history in descendants was noted in 46% of patients. More than 20% of subjects had motor, verbal or mental restlessness. Neuropsychological assessment highlighted impaired performances in executive functions (38%), sustained attention (67%), divided attention (45%), working memory (46%) and information processing speed (75%). A psychiatric history or comorbidities were present in 85% of patients, mostly of the anxio-depressive type. The more prevalent presentations of ADHD were the combined (38%) and inattentive (38%) types. Discussion Adult ADHD can masquerade as a cognitive impairment, including a stable cognitive complaint from infancy to old age. Inattentive, hyperactive and impulsive symptoms change with time and become more internalized (such as concentration difficulties or mental restlessness). No neuropsychological pattern has been reported but fluctuating deficits in sustained, divided attention, working memory and information processing speed are frequently observed in adult ADHD. A specific psychiatric expertise is essential in diagnosis and care for ADHD and its commonly associated psychiatric comorbidities.

The aim of this study was to report our experience after 10 years of practice of feminizing genitoplasty in prepubertal and adolescent patients with disorders of sex development (DSD) assigned females as females in a developing country.This was a cross-sectional, descriptive and retrospective study over a period of 9 years. All pre-pubertal (8-12 years) and adolescent patients female sex assigned with DSD who had willfully consented to the surgery with their guardians and underwent feminizing genital surgery were enrolled in the study. Data collection included: age at presentation, precise diagnosis, surgical procedures, complications, cosmetic result and duration of follow-up. Each patient had a precise diagnosis and the surgery was planned after discussion with the multidisciplinary team. Cosmetic results were assessed based on: appearance of the clitoris and separation of the vaginal and urethral openings.Nine patients raised as females with a median age of 8 years (IR: 10.75) were recorded. Surgery was performed at a median age of 11 years (IR: 9.5). In this series, 6 had a 46, XY karyotype with varying diagnoses: partial androgen insensitivity syndrome (n=2); 5-alphareductase insufficiency (n=2); 17-ketoreductase insufficiency (n=2); gonadal dysgenesis with a mutation in the NR5A1 gene (n=2), 2 had ovostesticular DSD, (karyotypes 46, XX), and 1 had mixed gonadal dysgenesis (karyotype 45, X/46, XY). Partial or total gonad(s) removal in accordance with assigned gender was the most common associated procedure. It was bilateral in 7 cases and unilateral in 2 cases. Follow-up ranged from 3 months to 4.5 years (median: 26 months, IR:18.25). One patient had acute urinary retention in the early follow-up. No other complication such as incision bleeding was recorded. The cosmetic appearance of the external genitalia was satisfactory in all patients.Feminizing genital surgery in Cameroon remains a major challenge and should seldom be realized without a precise diagnosis. Late age at presentation is peculiar to our setting; however, it gives room for the patients' participation and input to decisions that will have a life-long personal impact on their lives in terms of psychosocial development and fertility.3.

In both clinical and forensic psychiatry, it can often be difficult to distinguish delusions from normal beliefs. The categorical approach of the Diagnostic and Statistical Manual of Mental Disorders (DSM) leaves few options to describe intermediate delusion-like beliefs (DLBs). Neurocognitive models offer an alternative view of DLBs as existing on a continuum that can be quantified based on dimensions of severity as well as underlying cognitive biases. The Internet provides broadened access to putative evidence for diverse beliefs, with filter bubbles and echo chambers that can amplify confirmation bias and strengthen conviction. It is therefore much easier now for fringe beliefs to be shared and much less clear when they should be considered delusional. To place DLBs into a forensically relevant framework, psychiatric expert witnesses should adopt a broad biopsychosocial understanding of belief formation and maintenance that integrates clinical expertise with knowledge about dimensional aspects of delusions, cognitive biases, and the processing of online misinformation. The unavoidable conclusion that normal thinking is replete with cognitive biases and misbeliefs challenges the legal concept of

Deep Neural Networks (DNNs) have gained significant popularity in several classification and regression applications. The massive computation and memory requirements of DNNs pose special challenges for FPGA implementation. Moreover, programming FPGAs requires hardware-specific knowledge that many machine-learning researchers do not possess. To make the power and versatility of FPGAs available to a wider DNN user community and to improve DNN design efficiency, we introduce a Single hidden layer Neural Network (SNN) multiplication-free overlay architecture with fully connected DNN-level performance. This FPGA inference overlay can be used for applications that are normally solved with fully connected DNNs. The overlay avoids the time needed to synthesize, place, route and regenerate a new bitstream when the application changes. The SNN overlay inputs and activations are quantized to power-of-two values, which allows utilizing shift units instead of multipliers. Since the overlay is a SNN, we fill the FPGA chip with the maximum possible number of neurons that can work in parallel in the hidden layer. On a ZYNQ-7000 ZC706 FPGA, it is thus possible to implement 2450 neurons in the hidden layer and 30 neurons in the output layer. We evaluate the proposed architecture on typical benchmark datasets and demonstrate higher throughput with respect to the state-of-the-art while achieving the same accuracy.

Antipsychotics are some of the most frequently prescribed medications not only for psychotic disorders and symptoms but also for a wide range of on-label and off-label indications. Because second-generation antipsychotics have largely replaced first-generation antipsychotics as first-line options due to their substantially decreased risk of extrapyramidal side effects, attention has shifted to other clinically concerning adverse events associated with antipsychotic therapy. The focus of this article is to update the nonextrapyramidal side effects associated with second-generation antipsychotics. Issues surrounding diagnosis and monitoring as well as clinical management are addressed.

OBJECTIVE To demonstrate the effectiveness of a Diabetes Prevention Program-inspired 12-month behavioral intervention for patients with severe mental illness (SMI) and medication-associated obesity. METHOD This randomized, controlled, parallel, superiority study screened 225 volunteers from November 2005 to August 2008 at the VA Greater Los Angeles Healthcare System. 122 outpatients with DSM-IV-diagnosed SMI taking antipsychotic medications who had ≥ 7% weight gain or body mass index (BMI) > 25 were randomized by computer-generated number to Lifestyle Balance treatment intervention (n = 60) or usual care control (n = 62) groups. Clinical raters were masked to randomization. Treatment intervention included weekly classes and individual counseling for 8 weeks, food and exercise diaries, rewards, caregiver consultations, and monthly booster classes and counseling for 1 year. Controls received self-help materials and visited at equivalent intervals without formal classes or counseling. Outcomes were changes in anthropometric measurements, psychiatric symptoms, health knowledge, and glucose, hemoglobin A1c, and lipid levels. RESULTS Our intention-to-treat analysis found significant differences in predicted trajectory of mean weight change between the groups over 12 months (P < .01), with treatment participants expected to lose an average 4.6 kg, while control participants would gain an average 0.6 kg. BMI and body fat percentage followed the same pattern. Both groups demonstrated statistically significant improvements in health knowledge quiz scores over time (P = .006), without significant difference between groups. CONCLUSIONS Treatment was more effective than usual care control in treating medication-associated obesity, independent of SMI diagnosis, antipsychotic medication, and knowledge gained, suggesting that behavioral interventions are effective in SMI patients. TRIAL REGISTRATION ClinicalTrials.gov Identifier: NCT00344500.

SUMMARY Next generation sequencing (NGS) technologies, which allow high speed automated DNA sequencing, DNA sequence analysis and comparisons using big data algorithms are being used more and more in medical diagnosis, to provide prognosis information and for choosing a treatment that best fits the patient. The French “Academie nationale de medecine ” (ANM) and “ Academie des technologies “ (NATF) jointly stress the scientific and medical importance of these technologies and call for bringing together strengths of the national computer science industry and the medical community in a public-private consortium in order to build a Demonstrator and Center network filling up the growing gap between France and the most advanced countries in this field with great potential. They underline also common pitfalls which need to be addressed in technic, legal, economic, education and ethics issues, in order to make this complex project a success.

Application-specific customisation of micro-processor architectures has been widely accepted as an effective way to improve the efficiency of processor-based designs. In this work, the authors propose a new processor customisation method based on fixed-point word-length optimisation. Accuracy-aware word-length optimisation (WLO) of fixed-point circuits is an active research area with a large body of literature. For the first time, this work introduces a method to combine the WLO with the processor customisation. The data type word-lengths, the size of register-files and the architecture of the functional units are the main target objectives to be optimised. Accuracy requirements, defined as the worst-case error bound, is the key consideration that must be met by any solution. A custom processor design environment, called PolyCuSP, is used to realise the processor architecture based on the solution found in the proposed optimisation algorithm. The results achieved by evaluating five benchmark show that this method can reduce the number of necessary LUTs and flip-flops by an average of 11.9% and 5.1%, respectively. The latency is also improved by an average of 33.4%. Moreover, the method was further examined through a case study on a JPEG decoder. The results suggest 16.2% and 56.2% reduction in area consumption and latency, respectively.

Deep Neural Networks (DNNs) and Convolutional Neural Networks (CNNs) have gained significant popularity in several classification and regression applications. The massive computation and memory requirements of deep NN architectures pose particular challenges for their FPGA implementation. Moreover, programming FPGAs requires hardware-specific knowledge that many machine-learning researchers do not possess. To make the power and versatility of FPGAs available to a wider deep learning user community and to improve DNN design efficiency, we introduce POLYBiNN, a scalable and efficient combinatorial inference engine for DNNs and CNNs. POLYBiNN is composed of a stack of decision trees, which are binary classifiers by nature, and utilizes AND-OR gates instead of multipliers and accumulators. POLYBiNN drastically cuts the hardware consumption down while maintaining high accuracy, and it is a memory free inference engine. We also propose a tool that generates automatically a low-level hardware description of the trained POLYBiNN for a given application. We evaluate POLYBiNN for the MNIST dataset when implemented in a ZYNQ-7000 ZC706 FPGA platform. The system achieves a throughput of up to 100 million image classifications per second with 90 ns latency on the MNIST dataset with 97.18% accuracy. The power consumption of PLOYBiNN is less than 1.2 W. We also show how POLYBiNN can be used in the fully connected layers of a CNN and apply this approach to the CIFAR-10 dataset.

A tree-ring δ18O chronology of Linzhi spruce, spanning from AD 1781 to 2005, was developed in Bomi, Southeast Tibetan Plateau (TP). During the period with instrumental data (AD 1961–2005), this record is strongly correlated with regional CRU (Climate Research Unit) summer cloud data, which is supported by a precipitation δ18O simulation conducted with the isotope-enabled atmospheric general circulation model LMDZiso. A reconstruction of a regional summer cloud index, based upon the empirical relationship between cloud and diurnal temperature range, was therefore achieved. This index reflects regional moisture variability in the past 225 yr. The climate appears drier and more stable in the 20th century than previously. The drying trend in late 19th century of our reconstruction is consistent with a decrease in the TP glacier accumulation recorded in ice cores. An exceptional dry decade is documented in the 1810s, possibly related to the impact of repeated volcanic eruptions on monsoon flow.

Convolutional neural networks (CNNs) are responsible for many recent successes in the computer vision field and are now the dominant approach for image classification. However, CNN-based methods perform many convolution operations and have high power consumption which makes them difficult to deploy on mobile devices. In this paper, we propose a new method to reduce CNN power consumption by simplifying computations before max-pooling layers. The proposed method estimates the output of the max-pooling layer by approximating the preceding convolutional layer with a preliminary partial computation. Then, the method performs a complementary computation to generate an exact convolution output only for the selected feature. We also present an analysis of the approximation parameters. Simulation results show that the proposed method reduces the power consumption by 21% and the silicon area by 19% with negligible degradation in classification accuracy for the CIFAR−10 dataset.

International audience; Cloud Radio Access Network is foreseen as one of the key features of the future 5G mobile communication standard. In this context, all the baseband processing is intended to be performed on CPUs in order to keep a high level of flexibility. The challenge is then to propose efficient software implementations of baseband processing algorithms that guarantee a sufficient throughput, while limiting the energy consumption. In this paper, as an alternative to general purpose processors, we propose an implementation of an Application Specific Instruction set Processor customized for the Successive Cancellation decoding of polar codes. The resulting software decoder achieves throughputs similar to state-of-the-art ARM processor implementations, while reducing the energy consumption by a factor 10.

The emergence of 5G networks and real-time applications across networks has a strong impact on the performance requirements of IP lookup engines. These engines must support not only high-bandwidth but also low-latency lookup operations. This paper presents the hardware architecture of a low-latency IPv6 lookup engine capable of supporting the bandwidth of current Ethernet links. The engine implements the SHIP lookup algorithm, which exploits prefix characteristics to build a compact and scalable data structure. The proposed hardware architecture leverages the characteristics of the data structure to support low-latency lookup operations, while making efficient use of memory. The architecture is described in C++, synthesized with a high-level synthesis tool, then implemented on a Virtex-7 FPGA. Compared to other well-known approaches, the proposed IPvThe emergence of 5G networks and real-time applications across networks has a strong impact on the performance requirements of IP lookup engines. These engines must support not only high-bandwidth but also low-latency lookup operations. This paper presents the hardware architecture of a low-latency IPv6 lookup engine capable of supporting the bandwidth of current Ethernet links. The engine implements the SHIP lookup algorithm, which exploits prefix characteristics to build a compact and scalable data structure. The proposed hardware architecture leverages the characteristics of the data structure to support lowlatency lookup operations, while making efficient use of memory. The architecture is described in C++, synthesized with a highlevel synthesis tool, then implemented on a Virtex-7 FPGA. Compared to the proposed IPv6 lookup architecture, other wellknown approaches use at least 87% more memory per prefix, while increasing the lookup latency by a factor of 2.3×.6 lookup architecture reduces lookup latency by a factor of 2.3x and uses as much as 46% less memory per prefix for a synthetic prefix table holding 580 k entries.

To perform accurate and smooth behaviors in dynamic environments with moving objects, robotic visuomotor control should include the ability to process spatio-temporal information. We propose a system that uses a spatio-temporal deep neural network (DNN), with video camera pixels as the only input, to handle all the visual perception and visuomotor control functions needed to perform robotic behaviors such as leader following. Our approach combines: (1) end-to-end deep learning for inferring motion control outputs from visual inputs, (2) multi-task learning for simultaneously producing multiple control outputs with the same DNN, and (3) spatio-temporal deep learning for perceiving motion across multiple video frames. We use driving simulations to quantitatively show that spatio-temporal DNNs increase driving accuracy and driving smoothness by improving machine perception of scene kinematics. Experiments conducted with mobile robots in a laboratory test track show real-time embedded systems performance comparable to human reaction times to visual stimuli, and indicate that a spatio-temporal deep learning robot is able to follow a leader for long periods of time, while keeping within lanes and avoiding obstacles.

We present an experimental demonstration as well as a theoretical model of an integrated circuit designed for the manipulation of a microwave field down to the single-photon level. The device is made of a superconducting resonator coupled to a transmission line via a second frequency-tunable resonator. The tunable resonator can be used as a tunable coupler between the fixed resonator and the transmission line. Moreover, the manipulation of the microwave field between the two resonators is possible. In particular, we demonstrate the swapping of the field from one resonator to the other by pulsing the frequency detuning between the two resonators. The behavior of the system, which determines how the device can be operated, is analyzed as a function of one key parameter of the system, the damping ratio of the coupled resonators. We show a good agreement between experiments and simulations, realized by solving a set of coupled differential equations., 10 pages, 6 figures

Robotic following has many practical applications and involves several cognitive sub-tasks including object identification, localization, tracking, path planning, and motion control. We propose a system that uses a single deep neural network (DNN), with video camera pixels as the only input, to handle all the cognitive perception and visuomotor control functions needed to perform robotic following behaviors. Our approach combines 1) end-to-end deep learning for inferring motion control outputs from visual inputs, 2) multi-task learning for simultaneously producing multiple outputs with the same DNN, and 3) spatio-temporal deep learning for perceiving motion across multiple video frames. We use simulations of truck platooning to quantitatively show that spatio-temporal DNNs increase driving accuracy and improve machine perception of scene kinematics. Experiments conducted with mobile robots in a laboratory test track show real-time embedded systems performance and indicate that an end-to-end deep learning robot is able to follow its leader for long periods of time, while keeping within lanes and avoiding obstacles.

Packet parsing is a key step in SDN-aware devices. Packet parsers in SDN networks need to be both reconfigurable and fast, to support the evolving network protocols and the increasing multi-gigabit data rates. The combination of packet processing languages with FPGAs seems to be the perfect match for these requirements. In this work, we develop an open-source FPGA-based configurable architecture for arbitrary packet parsing to be used in SDN networks. We generate low latency and high-speed streaming packet parsers directly from a packet processing program. Our architecture is pipelined and entirely modeled using templated C++ classes. The pipeline layout is derived from a parser graph that corresponds a P4 code after a series of graph transformation rounds. The RTL code is generated from the C++ description using Xilinx Vivado HLS and synthesized with Xilinx Vivado. Our architecture achieves 100 Gb/s data rate in a Xilinx Virtex-7 FPGA while reducing the latency by 45% and the LUT usage by 40% compared to the state-of-the-art., Accepted for publication at the 26th ACM/SIGDA International Symposium on Field-Programmable Gate Arrays February 25 - 27, 2018 Monterey Marriott Hotel, Monterey, California, 7 pages, 7 figures, 1 table

Galaxy clusters trace the highest density peaks in the large-scale structure of the Universe. Their clustering provides a powerful probe that can be exploited in combination with cluster mass measurements to strengthen the cosmological constraints provided by cluster number counts. We investigate the spatial properties of a homogeneous sample of X-ray selected galaxy clusters from the XXL survey, the largest programme carried out by the XMM-Newton satellite. The measurements are compared to $\Lambda$-cold dark matter predictions, and used in combination with self-calibrated mass scaling relations to constrain the effective bias of the sample, $b_{eff}$, and the matter density contrast, $\Omega_{\rm M}$. We measured the angle-averaged two-point correlation function of the XXL cluster sample. The analysed catalogue consists of $182$ X-ray selected clusters from the XXL second data release, with median redshift $\langle z \rangle=0.317$ and median mass $\langle M_{500} \rangle\simeq1.3\cdot10^{14} M_\odot$. A Markov chain Monte Carlo analysis is performed to extract cosmological constraints using a likelihood function constructed to be independent of the cluster selection function. Modelling the redshift-space clustering in the scale range $10, Comment: 10 pages, 10 figures. Accepted for publication in Astronomy & Astrophysics

Introduction et but de l’etude Le boost de curietherapie de bas debit de dose par iode 125 augmente la probabilite de survie biochimique au prix d’une toxicite urinaire accrue par rapport a une radiotherapie externe moderne. Nous nous sommes propose d’etudier la toxicite urinaire et digestive du boost de curietherapie prostatique de haut debit de dose dans une cohorte monocentrique. Materiel et methodes Les 263 premiers patients traites de 2010 a 2015 au CHU de Poitiers par irradiation conformationnelle tridimensionnelle de 46 Gy avec tomographie conique quotidienne et boost curietherapie de haut debit de dose totale de la prostate par iridium 192, avec escalade de dose focale sur lesion index par fusion de l’IRM et de l’ultrasonographie en temps reel ont ete inclus. Dix-huit patients ont ete exclus, pour second cancer synchrone, atteinte ganglionnaire/metastatique, et radiotherapie externe en dehors du CHU. Il s’agit d’une etude retrospective de la toxicite aigue (moins de 6mois) et chronique (selon les Common Toxicity Criteria for Adverse Events, CTCAEv4.0), initialement a 1 mois, puis annuellement. Resultats et analyse statistique Les donnees de 245patients ont ete analysees, avec un suivi median de 4,15 ans [0,6 ;7,5 ans]. Selon la classification du National Comprehensive Cancer Network (NCCN), 49 % des cancers etaient de haut risque, 43 % de risque intermediaire, 9 % de risque faible avec depassement capsulaire a l’IRM. Il y a eu 78 % d’hormonotherapies prescrites, dont 81,2 % de longue duree, 7 % de radiotherapies externes ganglionnaires pelviennes. Le volume prostatique moyen etait de 35,4 cm3 [11,1 ;88,1 cm3], le nombre moyen d’aiguilles de 20 [14 ;24]. La dose de curietherapie de haut debit de dose prescrite dans le volume cible anatomoclinique etait chronologiquement de 2 × 7,5 Gy (n = 18), 2 × 9 Gy (n = 12), 2 × 10Gy (n = 124), 14 Gy (n = 1) et 15 Gy (n = 88). Le taux de toxicite urinaire de grade 2 aigue etait de 86 %, tardive 62 %, de grade 3 aigue 0 %, tardive 4,5 % (n = 11) en incidence cumulee (sept uretrotomies internes, une resection prostatique tranduretrale, deux dilatations, une derivation extra-anatomique de type Bricker), 1,6 % (n = 4) persistaient aux dernieres nouvelles (une sonde a demeure pour incontinence secondaire a une uretrotomie, deux pollakiuries horaires, une chirurgie de type Bricker). Le taux de toxicite digestive de grade 2 aigue etait de 20 %, tardive de 16 %, de grade 3 aigue 0 %, tardive 0,4 % (n = 1) en incidence cumulee (rectorragies resolutives sous caisson hyperbare). La probabilite de survie sans recidive biochimique etait de 94,7 %. Conclusion Pour le cancer de prostate localement evolue, notre etude a retrouve une toxicite urinaire et digestive cumulee de grade 3 faible en faveur de l’utilisation du boost par curietherapie de haut debit de dose par rapport au bas debit de dose par iode 125. Ce resultat doit etre confirme en situation prospective et multicentrique.

Introduction La polykystose renale est une maladie genetique associee au developpement de kystes hepato-renaux. Les patients polykystiques peuvent necessiter une reduction volumique renale en pre-transplantation. Le traitement standard est la nephrectomie exposant au risque d’allo-immunisation et, en cas d’epuration extra-renale par dialyse peritoneale (DP), a un arret temporaire de la technique voire au transfert definitif en hemodialyse. L’embolisation arterielle renale (EAR) unilaterale peut etre utilisee en alternative a la nephrectomie. Methodes Nous avons realise une etude retrospective multicentrique dans 12 centres. L’objectif de cette etude etait d’evaluer la survie technique de la DP apres EAR unilaterale en comparaison a la nephrectomie unilaterale chez les patients polykystiques. Les objectifs secondaires visaient a comparer la frequence et la duree d’hemodialyse transitoire entre les 2 groupes et l’impact de l’intervention sur les parametres d’adequation de la DP. 37 patients ont ete inclus, 21 patients embolises et 16 patients nephrectomises. Resultats obtenus ou attendus Un transfert permanent en hemodialyse a ete observe chez 6 patients embolises (28,6 %) versus 11 patients nephrectomises (68,8 %) (p = 0,0001). L’EAR etait associe a une meilleure survie technique : SHR 0,29 [0,12–0,75] (p = 0,01). En analyse multivariee, la reduction volumique renale par EAR et le sexe masculin diminuait le risque de transfert definitif en hemodialyse. Apres EAR unilaterale, une reduction mediane du volume de 48,4 % etait observee (p = 0,0001). Malgre une diminution des parametres d’adequation, seulement 6 patients ont ete transferes de facon permanente en hemodialyse dont 3 d’entre eux apres nephrectomie secondaire en raison de l’inefficacite premiere de l’embolisation. Aucun patient embolise n’a eu besoin de seances d’hemodialyses transitoires. La duree de l’hospitalisation etait reduite a 5 jours [4,0–6,0] versus 8,5 jours [6,0–11,0] chez les patients nephrectomises. Conclusion L’EAR unilaterale permet une meilleure survie technique chez les patients polykystiques en DP et doit etre utilisee preferentiellement dans cette population de patients.