Search

International audience; When deploying Artificial Neural Networks (ANNs) onto multi-core embedded platforms, an intensive evaluation flow is necessaryto find implementations that optimize resource usage, timing andpower. ANNs require indeed significant amounts of computationaland memory resources to execute, while embedded execution plat-forms offer limited resources with strict power budget. Concurrentaccesses from processors to shared resources on multi-core plat-forms can lead to bottlenecks with impact on performance andpower. Existing approaches show limitations to deliver fast yetaccurate evaluation ahead of ANN deployment on the targetedhardware. In this paper, we present a modeling flow for timing andpower prediction in early design stage of fully-connected ANNs onmulti-core platforms. Our flow offers fast yet accurate predictionswith consideration of shared communication resources and scalabil-ity in regards of the number of cores used. The flow is evaluated onreal measurements for 42 mappings of 3 fully-connected ANNs exe-cuted on a clock-gated multi-core platform featuring two differentcommunication modes: polling or interrupt-based. Our modelingflow predicts timing with 97 % accuracy and power with 96 % accu-racy on the tested mappings for an average simulation time of 0.23 sfor 100 iterations. We then illustrate the application of our approachfor efficient design space exploration of ANN implementations.

National audience; Predicting the performance of Artificial NeuralNetworks (ANNs) on embedded multi-core platforms is tedious.Concurrent accesses to shared resources are hard to model dueto congestion effects on the shared communication medium,which affect the performance of the application. In this paperwe present a hybrid modeling environment to enable fast yetaccurate timing prediction for fully-connected ANNs deployedon multi-core platforms. The modeling flow is based on theintegration of an analytical computation time model with acommunication time model which are both calibrated throughmeasurement inside a system level simulation using SystemC. Theproposed flow enables the prediction of the end-to-end latencyfor different mappings of several fully-connected ANNs with anaverage of more than 99 % accuracy.

Evaluation of performance for complex applications such as Artificial Intelligence (AI) algorithms and more specifically neural networks on Multi-Processor Systems on a Chip (MPSoC) is tedious. Finding an optimized partitioning of the application while predicting accurately the latency induced by communication bus congestion, is hard using traditional analysis methods. This document presents a performance prediction workflow based on SystemC simulation models for timing prediction of neural networks on MPSoC.

Niemann-Pick type C (NPC) disease is an autosomal recessive lysosomal storage disorder caused by mutations in NPC1 or NPC2 genes. In 2009, the molecular characterization of 44 NPC Italian patients has been published. Here, we present an update of the genetic findings in 105 Italian NPC patients belonging to 83 unrelated families (77 NPC1 and 6 NPC2). NPC1 and NPC2 genes were studied following an algorithm recently published. Eighty-four different NPC1 and five NPC2 alleles were identified. Only two NPC1 alleles remained non detected. Sixty-two percent of NPC1 alleles were due to missense variants. The most frequent NPC1 mutation was the p.F284Lfs*26 (5.8% of the alleles). All NPC2 mutations were found in the homozygous state, and all but one was severe. Among newly diagnosed patients, 18 novel NPC1 mutations were identified. The pathogenic nature of 7/9 missense alleles and 3/4 intronic variants was confirmed by filipin staining and NPC1 protein analysis or mRNA expression in patient’s fibroblasts. Taken together, our previous published data and new results provide an overall picture of the molecular characteristics of NPC patients diagnosed so far in Italy.

Al cor del desafiament de l'Agenda Digital Europea SmarCities és la construcció d'un nou tipus de bé comú per a produir la inclusió i millorar la nostra vida quotidiana, la 6 creació de valor social i econòmic no només a través de l'ús de tecnologies i aplicacions de nova generació. L'objectiu d'aquest estudi és aclarir què s'entén per SmartCity, i analitzar si els instruments existents detecten la creació de valor públic. S'ha investigat la importància del paradigma acadèmic, la recerca d'una àmplia gamma de contribucions teòriques, però que no tenen un fil conductor que permet una visió objectiva del grau de smartness assolit. Cada enfocament ha sigut acompanyat per la seva pròpia escala d'indicadors. Les valoracions actual smartness Urbana són indicadors explícits que detecten el valor públic. Hem indagat algunes de les iniciatives Smartness establertes per les administracions públiques: la decisió pública sovint es va trobar davant d'una creixent oferta de solucions tecnològiques ‘clau en mà’. L'AP sovint s'ha centrat en l'efecte o d'emergència solucions sense preguntarse sobre la seva funció principal: la creació de valor públic. Per tant, es va trobar que el primer pas cap al valor públic és per tant a nivell de govern: és essencial per a una 'visió' al llarg termini a través de la participació activa dels ciutadans dels objectius de millora de la vida quotidiana i la identificació de les necessitats reals dels ciutadans. La creació de valor implica també la protecció dels llocs públics i el dret a la bellesa de la seva ciutat, recuperant el valor inherent als conceptes no inclosos en els grups d'indicadors intelligents: la vitalitat de la ciutat, la llegibilitat del teixit urbà, la millora dels espais públics, la recuperació dels suburbis i centres urbans com a llocs d'importància per a l'individu i la comunitat. La Vision Política, per planejar futures SmartCities, requereix una valorització del 'Bé Públic'., En el corazón del desafío de la Agenda Digital Europea encontramos la construcción de un nuevo tipo de bien común para producir la inclusión, mejorar nuestra vida cotidiana y crear valor social mediante el uso de las nuevas tecnologías: la SmartCity. El objetivo de este estudio es aclarar qué se entiende por SmartCity y analizar si los instrumentos existentes detectan la creación de Valor Público. Se ha investigado el significado del paradigma a nivel académico, encontrando una vasta gama de contribuciones teóricas, pero sin un hilo conductor que permita una visión objetiva del grado de smartness producido: cada enfoque ha seguido su propia escala de indicadores. En las actuales valoraciones de Inteligencia Urbana faltan indicadores explícitos que manifiesten el Valor producido. Hemos averiguado algunas iniciativas de smartness llevadas a cabo por las Administraciones Públicas: a menudo la Administración se ha encontrado de frente a una creciente oferta de soluciones tecnológicas listas para ser usadas. Las Administraciones Públicas, a menudo, han apostado por soluciones efectistas o de emergencia, sin preguntarse por su propia razón de ser: la creación de Valor Público. Encontramos que el primer paso hacia acciones explícitas está a nivel ejecutivo: la Administracion Pública necesita una visión a largo plazo para implicarse activamente con la ciudadanía sobre objetivos de mejora de la vida cotidiana y la individualización de las necesidades reales de los ciudadanos. La creación de Valor supone también la protección de los lugares públicos y el derecho a tener una ciudad bonita, recobrando el valor inherente a los conceptos no incluidos en los grupos de indicadores inteligentes: la vitalidad de la ciudad, la legibilidad del tejido urbano, la mejora de los espacios públicos, la recuperación de los suburbios y centros urbanos como lugares significativos para el individuo y la comunidad. La Vision Política, para planear futuras SmartCities, requiere una valorización del „Bien Público‟., To the center of the challenge launched by the European Digital Agenda, there is the construction of a new kind of ‘common good’ that produces inclusion and improves our daily way of living, creating social value through the use of technology. The aim of this study is clarifing what SmartCity meands and analyzing if the actual tools notice the creation of public value. We investigated the significance of academic paradigm, finding a wide range of theoretical contributions, but without a common thread that allows an objective view of the degree of smartness product. Each approach has 7 followed its own scale of indicators. In the current feedback of Urban Intelligence are explicit indicators that detect the public value. We checked out a few of smartness initiatives implemented by Public Administrations: often the public decision-maker is faced with a growing range of technological solutions ‘turnkey’. Public administrations often have focused on the effect or emergency solutions without wondering about its primary function: the creation of public value. We found that the first step towards the Public Value is therefore a Governance level: it is essential to a 'vision' in the long term through the active involvement of citizens of improvement goals of daily life and the identification of the real needs of citizens. The creation of value also involves the protection of public places and the right to the beauty of their city, recovering the value inherent in concepts not yet included in the Smart indicator sets: vitality of the city, readability of the urban fabric, enhancement of public spaces, recovery of suburbs and urban centers as places of significance for the individual and the community. To design the future SmartCity, it requires a political vision that values the ‘Public Good’.

Large thrombus burden (LTB) lesions in the context of primary percutaneous coronary intervention (p-PCI) have been related to unsuccessful angiographic reperfusion and unfavorable clinical outcomes. However, the hazard of LTB treatment on myocardial damage has not been evaluated. We investigated the impact of LTB on myocardial damage using contrast-enhanced cardiac magnetic resonance (CE-CMR) in the setting of p-PCI. In 327 patients, who underwent p-PCI without thrombus aspiration within 12 hours from symptom onset, we prospectively assessed the impact of LTB on infarct size and microvascular damage using CE-CMR. LTB was defined by the presence of Thrombolysis In Myocardial Infarction thrombus score ≥3 in patent infarct-related artery (IRA); or by "cut-off" occlusion pattern and/or large reference vessel diameter (≥3.5 mm) in occluded IRA. One hundred ninety-seven patients (60.2%) showed LTB and 130 (39.8%) did not. Distal embolization occurred in 18.8% patients with versus 6.9% without LTB (p = 0.003). At CE-CMR, patients with LTB had larger infarct size index (27.5 ± 11.1 vs 22.1 ± 17.5, p = 0.009) and more often transmural necrosis (70.5% vs 55.4%, p = 0.008) compared with patients without LTB. Excluding patients with distal embolization, patients with LTB still had larger necrosis. At multivariate analysis, occluded (IRA) at baseline, anterior infarction, and presence of LTB predicted transmural necrosis. In conclusion, LTB in the setting of p-PCI is related to larger myocardial damage as detected by CE-CMR, regardless of angiographic detectable distal embolization.

This paper presents the design of a novel single square millimeter three-axial accelerometer for head injury detection of racing car drivers. The main requirements of this application are miniaturization and medium/high-G measurement range. We propose a new miniature accelerometer to be incorporated into an earpiece. Nanowires as nanoscale piezoresistive devices have been chosen as sensing element, due to their high sensitivity and miniaturization achievable. By exploiting the electromechanical features of nanowires as nanoscale piezoresistors, the nominal sensor sensitivity is overall boosted by more than 30 times. This approach allows significant higher accuracy and resolution with smaller sensing element in comparison with conventional devices without the need of signal amplification. This achievement opens up new developments in the area of implanted devices where the high-level of miniaturization and sensitivity is essential.

The purpose of this study was to examine the relationship between students' academic competitiveness and their involvement behaviors (i.e., student motives for communicating with their instructors, out-of-class communication with their instructors, and classroom participation). Considering the desire inherent among academically competitive students to outperform other students, it was predicted that students' academic competitiveness would be related positively to their involvement behaviors. Participants (N=246) completed a series of self-report measures. The results of Pearson correlation analyses largely provided support for the hypothesized association among students' academic competitiveness and their involvement in the learning process.

This study aims to develop and validate a new angiographic risk score to predict the risk of distal embolization (DE) during primary percutaneous coronary intervention (p-PCI) for ST-elevation myocardial infarction. Study included data from 1,200 patients who underwent p-PCI. The cohort was randomly split into a derivation cohort (n = 814) and a validation cohort (n = 386). Logistic regression was used to examine the relation between risk factors and the occurrence of DE. To each covariate in the model was assigned an integer score based on the regression coefficients. Variables included in the risk score, according to multivariable analysis, were occlusion pattern of infarct-related artery, Thrombolysis In Myocardial Infarction Thrombus Score 2 to 4, reference vessel diameter ≥3.5 mm, and lesion length >20 mm. To each variable was assigned a 0- to +2-point score according to the strength of the statistical association. Rates of DE in low-, intermediate-, and high-risk groups were 5.6%, 15.8%, and 40% in the derivation cohort (p for trend

BackgroundRecent studies have described neo-atherosclerosis, developing inside the stent, as cause of very late stent thrombosis.Case reportA 59-year-old man, with family history of coronary artery disease, presented to our Department because of anterior ST-segment elevation myocardial infarction. Two years before he had underwent percutaneous coronary intervention with multiple drug-eluting stents (DES) implantation on proximal-mid left anterior descending artery (LAD), and mid-right coronary artery (RCA), respectively. The angiogram revealed stent thrombosis with total occlusion of proximal LAD. Multiple passages with manual thrombus-aspiration catheter were successfully performed with improvement in TIMI flow. Optical Coherence Tomography (OCT) imaging revealed fully expanded stents without areas of inappropriate apposition to vessel wall; and mild to moderate intimal hyperplasia throughout the stented segment, with full covered stent struts; areas of ulcerated and ruptured plaque within the proximal struts of stented segment was depicted with intraluminal protruding material. Thus, an additional bare metal stent (BMS) was deployed inside and overlapping the previous in order to seal this plaque. OCT post procedure revealed optimal stent expansion and apposition, without residual protruding material. At 9-month follow-up patient was alive and free from symptoms. Coronary angiogram revealed patency of implanted stents without significant restenosis.ConclusionsNeo-atherosclerosis with thrombosis on top of ruptured necrotic plaque core may play a role in the pathophysiology of very late stent thrombosis in both BMS and DES. Our report highlights the role OCT to assess the mechanism of VLST.

Objectives The aim of this in vitro study is to evaluate the marginal and internal fit of zirconia core crowns manufactured following different digital and traditional workflows. Methods A 6° taper shoulder prepared abutment tooth was used to produce 20 zirconia core crowns using four different scanning techniques: scanned directly with the extraoral lab scanner, scanned with intraoral scanner, dental impressions using individual dental tray and polyether, dental casts from a polyether impressions. Marginal and internal fits were evaluated with digital photography and the silicone replica method. Results Medium marginal gaps were 76,00 μm ± 28.9 for extraoral lab scanner, 80.50 μm ± 36,2 for intraoral scanner, 88.10 μm ± 34,8 for dental impression scan and 112,4 μm ± 37,2 for dental cast scan. Medium internal gaps were 23.20 μm ± 10,3 for extraoral lab scanner, 16.20 μm ± 8.3 for intraoral scanner, 27.20 μm ± 16.7 for dental impression scan and 30.20 μm ± 12.7 for dental cast scan. Conclusion Internal gap were extensively lower than 70 μm described in literature. Marginal fit was higher than ideal values for all the techniques but within the limit of clinical success. Intraoral scanners obtained the best results for internal gap.

The LHC dispersion suppressors and matching sections will be equipped with individually powered superconducting quadrupoles with an aperture of 56 mm. In order to optimise the parameters and cost of the magnets and of their powering, the quadrupole has been designed on the basis of an 8.2 mm wide Rutherford-type cable for a nominal current of 5300 A, corresponding to a gradient of 200 T/m at 1.9 K. In order to validate the design two 1-m single-aperture quadrupoles and one twin-aperture quadrupole have been built and tested. In this report we describe the construction features of the magnets and present the results of the magnet tests. (4 refs).

The LHC dispersion suppressors and matching sections will be equipped with individually powered superconducting quadrupoles with an aperture of 56 mm. In order to minimise the cost of the powering circuits, the quadrupole has been designed on the basis of an 8 mm wide NbTi Rutherford-type cable for a nominal current of 5300 A, corresponding to a gradient of 200 T/m at 1.9 K. In order to validate the design options a model magnet program has been launched. In this report we describe the construction features of the first 1-m long magnet, and present its training performance and the results of protection studies.