Many-Core Architectures: Hardware-Software Optimization and Modeling Techniques

During the last few decades an unprecedented technological growth has been at the center of the embedded systems design paramount, with Moore’s Law being the leading factor of this trend. Today in fact an ever increasing number of cores can be integrated on the same die, marking the transition from state-of-the-art multi-core chips to the new many-core design paradigm. Despite the extraordinarily high computing power, the complexity of many-core chips opens the door to several challenges. As a result of the increased silicon density of modern Systems-on-a-Chip (SoC), the design space exploration needed to find the best design has exploded and hardware designers are in fact facing the problem of a huge design space. Virtual Platforms have always been used to enable hardware-software co-design, but today they are facing with the huge complexity of both hardware and software systems. In this thesis two different research works on Virtual Platforms are presented: the first one is intended for the hardware developer, to easily allow complex cycle accurate simulations of many-core SoCs. The second work exploits the parallel computing power of off-the-shelf General Purpose Graphics Processing Units (GPGPUs), with the goal of an increased simulation speed. The term Virtualization can be used in the context of many-core systems not only to refer to the aforementioned hardware emulation tools (Virtual Platforms), but also for two other main purposes: 1) to help the programmer to achieve the maximum possible performance of an application, by hiding the complexity of the underlying hardware. 2) to efficiently exploit the high parallel hardware of many-core chips in environments with multiple active Virtual Machines. This thesis is focused on virtualization techniques with the goal to mitigate, and overtake when possible, some of the challenges introduced by the many-core design paradigm.

Parahydrogen induced polarization on a clinical MRI system : polarization transfer of two spin order

Hyperpolarization techniques enhance the nuclear spin polarization and thus allow for new nuclear magnetic resonance applications like in vivo metabolic imaging. One of these techniques is Parahydrogen Induced Polarization (PHIP). It leads to a hyperpolarized 1H spin state which can be transferred to a heteronucleus like 13C by a radiofrequency (RF) pulse sequence. In this work, timing of such a sequence was analyzed and optimized for the molecule hydroxyethyl propionate. The pulse sequence was adapted for the work on a clinical magnetic resonance imaging (MRI) system which is usually equipped only with a single RF transmit channel. Optimal control theory optimizations were performed to achieve an optimized polarization transfer. A drawback of hyperpolarization is its limited lifetime due to relaxation processes. The lifetime can be increased by storing the hyperpolarization in a spin singlet state. The second part of this work therefore addresses the spin singlet state of the Cs-symmetric molecule dimethyl maleate which needs to be converted to the spin triplet state to be detectable. This conversion was realized on a clinical MRI system, both by field cycling and by two RF pulse sequences which were adapted and optimized for this purpose. Using multiple conversions enables the determination of the lifetime of the singlet state as well as the conversion efficiency of the RF pulse sequence. Both, the hyperpolarized 13C spin state and the converted singlet state were utilized for MR imaging. Careful choice of the echo time was shown to be crucial for both molecules.

Mappatura degli habitat e degli oggetti antropici della laguna di Venezia

L’utilizzo del Multibeam Echo sounder (MBES) in ambienti di transizione poco profondi, con condizioni ambientali complesse come la laguna di Venezia, è ancora in fase di studio e i dati biologici e sedimentologici inerenti ai canali della laguna di Venezia sono attualmente scarsi e datati in letteratura. Questo studio ha lo scopo di mappare gli habitat e gli oggetti antropici di un canale della laguna di Venezia in un intervallo di profondità tra 0.3 e 20 m (Canale San Felice) analizzando i dati batimetrici e di riflettività (backscatter) acquisiti da ISMAR-Venezia nell’ambito del progetto RITMARE. A tale scopo il fondale del canale San Felice (Venezia) è stato caratterizzato dal punto di vista geomorfologico, sedimentologico e biologico; descrivendo anche l’eventuale presenza di oggetti antropici. L’ecoscandaglio utilizzato è il Kongsberg EM2040 Dual-Compact Multibeam in grado di emettere 800 beam (400 per trasduttore) ad una frequenza massima di 400kHZ e ci ha consentito di ricavare ottimi risultati, nonostante le particolari caratteristiche degli ambienti lagunari. I dati acquisiti sono stati processati tramite il software CARIS Hydrographic information processing system (Hips) & Sips, attraverso cui è possibile applicare le correzioni di marea e velocità del suono e migliorare la qualità dei dati grezzi ricavati da MBES. I dati sono stati quindi convertiti in ESRI Grid, formato compatibile con il software ArcGIS 10.2.1 (2013) che abbiamo impiegato per le interpretazioni e per la produzione delle mappe. Tecniche di ground-truthing, basate su riprese video e prelievi di sedimento (benna Van Veen 7l), sono state utilizzate per validare il backscatter, dimostrandosi molto efficaci e soddisfacenti per poter descrivere i fondali dal punto di vista biologico e del substrato e quindi degli habitat del canale lagunare. Tutte le informazioni raccolte durante questo studio sono state organizzate all’interno di un geodatabase, realizzato per i dati relativi alla laguna di Venezia.

Ventricular prosthesis embolization during transapical aortic valve implantation: the role of transesophageal echocardiography in diagnosis and management

The authors present the case of an 81-year-old patient with severe aortic stenosis who experienced left ventricular embolization of an aortic bioprosthesis during transapical aortic valve implantation. The authors discuss reasons for prosthesis embolization and reinforce the attention to technical details and the widespread use of multimodality imaging techniques. In this context, transesophageal echocardiography appears indispensable in the detection and management of procedure-related complications.

Toward optimal display of physiologic status in critical care: I. Recreating bedside displays from archived physiologic data

BACKGROUND: Physiologic data display is essential to decision making in critical care. Current displays echo first-generation hemodynamic monitors dating to the 1970s and have not kept pace with new insights into physiology or the needs of clinicians who must make progressively more complex decisions about their patients. The effectiveness of any redesign must be tested before deployment. Tools that compare current displays with novel presentations of processed physiologic data are required. Regenerating conventional physiologic displays from archived physiologic data is an essential first step. OBJECTIVES: The purposes of the study were to (1) describe the SSSI (single sensor single indicator) paradigm that is currently used for physiologic signal displays, (2) identify and discuss possible extensions and enhancements of the SSSI paradigm, and (3) develop a general approach and a software prototype to construct such "extended SSSI displays" from raw data. RESULTS: We present Multi Wave Animator (MWA) framework-a set of open source MATLAB (MathWorks, Inc., Natick, MA, USA) scripts aimed to create dynamic visualizations (eg, video files in AVI format) of patient vital signs recorded from bedside (intensive care unit or operating room) monitors. Multi Wave Animator creates animations in which vital signs are displayed to mimic their appearance on current bedside monitors. The source code of MWA is freely available online together with a detailed tutorial and sample data sets.

Quantitative T2 mapping of knee cartilage: differentiation of healthy control cartilage and cartilage repair tissue in the knee with unloading - initial results

Purpose: To prospectively determine on T2 cartilage maps the effect of unloading during a clinical magnetic resonance (MR) examination in the postoperative follow-up of patients after matrix-associated autologous chondrocyte transplantation (MACT) of the knee joint. Materials and Methods: Ethical approval for this study was provided by the local ethics commission, and written informed consent was obtained. Thirty patients (mean age, 35.4 years +/- 10.5) with a mean postoperative follow-up period of 29.1 months +/- 24.4 were enrolled. A multiecho spin-echo T2-weighted sequence was performed at the beginning (early unloading) and end (late unloading) of the MR examination, with an interval of 45 minutes. Mean and zonal region of interest T2 measurements were obtained in control cartilage and cartilage repair tissue. Statistical analysis of variance was performed. Results: The change in T2 values of control cartilage (early unloading, 50.2 msec +/- 8.4; late unloading, 51.3 msec +/- 8.5) was less pronounced than the change in T2 values of cartilage repair tissue (early unloading, 51.8 msec +/- 11.7; late unloading, 56.1 msec +/- 14.4) (P = .024). The difference between control cartilage and cartilage repair tissue was not significant for early unloading (P = .314) but was significant for late unloading (P = .036). Zonal T2 measurements revealed a higher dependency on unloading for the superficial cartilage layer. Conclusion: Our results suggest that T2 relaxation can be used to assess early and late unloading values of articular cartilage in a clinical setting and that the time point of the quantitative T2 measurement affects the differentiation between native and abnormal articular cartilage. (c) RSNA, 2010.

Circulation dynamics and its influence on European and Mediterranean January-April climate over the past half millennium: results and insights from instrumental data, documentary evidence and coupled climate models

We use long instrumental temperature series together with available field reconstructions of sea-level pressure (SLP) and three-dimensional climate model simulations to analyze relations between temperature anomalies and atmospheric circulation patterns over much of Europe and the Mediterranean for the late winter/early spring (January–April, JFMA) season. A Canonical Correlation Analysis (CCA) investigates interannual to interdecadal covariability between a new gridded SLP field reconstruction and seven long instrumental temperature series covering the past 250 years. We then present and discuss prominent atmospheric circulation patterns related to anomalous warm and cold JFMA conditions within different European areas spanning the period 1760–2007. Next, using a data assimilation technique, we link gridded SLP data with a climate model (EC-Bilt-Clio) for a better dynamical understanding of the relationship between large scale circulation and European climate. We thus present an alternative approach to reconstruct climate for the pre-instrumental period based on the assimilated model simulations. Furthermore, we present an independent method to extend the dynamic circulation analysis for anomalously cold European JFMA conditions back to the sixteenth century. To this end, we use documentary records that are spatially representative for the long instrumental records and derive, through modern analogs, large-scale SLP, surface temperature and precipitation fields. The skill of the analog method is tested in the virtual world of two three-dimensional climate simulations (ECHO-G and HadCM3). This endeavor offers new possibilities to both constrain climate model into a reconstruction mode (through the assimilation approach) and to better asses documentary data in a quantitative way.

Skeletal muscle (1) H MRSI before and after prolonged exercise. II. visibility of free carnitine

Carnitine (Car) buffers excess acetyl-CoA through the formation of acetylCar (AcCar). AcCar's acetyl group (AG-AcCar) gives rise to a peak at 2.13 ppm in ¹H MR spectra of skeletal muscle, whereas the trimethylammonium (TMA) groups of both, AcCar and Car, are thought to contribute to the TMA peak at 3.23 ppm. Surprisingly, in previous studies both resonances, AG-AcCar and TMA, increased after exercise. The aim of this study was to assess if the exercise-related TMA increase correlated with AcCar production. Magnetic resonance spectroscopic imaging (pulse repetition time/echo time = 1200/35 ms) was performed before and after prolonged exercise in the lower leg and thigh of eight runners and eight cyclists, respectively. TMA and AG-AcCar increased after exercise (P < 0.001). TMA's increase correlated with the AG-AcCar increase (R² = 0.73, P < 0.001, lower leg; R² = 0.28, P < 0.001, thigh). The correlation of ΔTMA with ΔAG-AcCar suggests that the TMA increase is due to AcCar formation. As total Car (Car + AcCar) remains unchanged with exercise, these findings suggest that the contribution of free Car to the TMA peak is limited and, therefore, is partly invisible in muscle ¹H MR spectra. This indicates that the biochemically relevant cytosolic content of free Car is considerably lower than the overall concentration determined by radioisotopic assays, a potentially important result with respect to regulation of substrate oxidation.

Gadolinium diethylenetriaminepentaacetate enhancement kinetics in the menisci of asymptomatic subjects: a first step towards a dedicated dGEMRIC (delayed gadolinium-enhanced MRI of cartilage)-like protocol for biochemical imaging of the menisci

It was our aim to investigate the gadolinium diethylenetriaminepentaacetate (Gd-DTPA(2-) ) enhancement kinetics in the menisci of the knee joint over a prolonged period of time. Six asymptomatic volunteers (four men and two women; mean age, 25 ± 2.4 years) were enrolled. Sagittal, T(1) -weighted, spin-echo MR sequences of the right knee joint were obtained at 3 T. Imaging was performed before (baseline), 1 h after and in half-hour intervals up to 9 h after the intravenous administration of 0.2 mmol/kg of Gd-DTPA(2-) . To measure the rates of contrast enhancement relative to the baseline, regions of interest that covered the anterior and posterior horns of the medial and lateral meniscus were defined on each of two adjacent sections, and enhancement curves were constructed. An enhancement peak between 2.5 and 4.5 h after Gd-DTPA(2-) administration was observed, and analysis of variance also revealed no significant difference (p=0.94), in terms of enhancement, within this time interval. Pair-wise, post hoc testing also revealed no significant differences between 2.5 and 3, 3 and 3.5, 3.5 and 4, and 4 and 4.5 h post Gd-DTPA(2-) application. Our preliminary data therefore suggest that the time window suitable for a dGEMRIC (delayed gadolinium-enhanced MRI of cartilage)-like T(1) mapping of the menisci is relatively short, and lies between 2.5 and 4.5 h after Gd-DTPA(2-) injection.

Quantitative mapping of T2 using partial spoiling

Fast quantitative MRI has become an important tool for biochemical characterization of tissue beyond conventional T1, T2, and T2*-weighted imaging. As a result, steady-state free precession (SSFP) techniques have attracted increased interest, and several methods have been developed for rapid quantification of relaxation times using steady-state free precession. In this work, a new and fast approach for T2 mapping is introduced based on partial RF spoiling of nonbalanced steady-state free precession. The new T2 mapping technique is evaluated and optimized from simulations, and in vivo results are presented for human brain at 1.5 T and for human articular cartilage at 3.0 T. The range of T2 for gray and white matter was from 60 msec (for the corpus callosum) to 100 msec (for cortical gray matter). For cartilage, spatial variation in T2 was observed between deep (34 msec) and superficial (48 msec) layers, as well as between tibial (33 msec), femoral, (54 msec) and patellar (43 msec) cartilage. Excellent correspondence between T2 values derived from partially spoiled SSFP scans and the ones found with a reference multicontrast spin-echo technique is observed, corroborating the accuracy of the new method for proper T2 mapping. Finally, the feasibility of a fast high-resolution quantitative partially spoiled SSFP T2 scan is demonstrated at 7.0 T for human patellar cartilage.

Does quantification of USPIO uptake-related signal loss allow differentiation of benign and malignant normal-sized pelvic lymph nodes?

Ultrasmall superparamagnetic iron oxide (USPIO) particles are promising contrast media, especially for molecular and cellular imaging besides lymph node staging owing to their superior NMR efficacy, macrophage uptake and lymphotropic properties. The goal of the present prospective clinical work was to validate quantification of signal decrease on high-resolution T(2)-weighted MR sequences before and 24-36 h after USPIO administration for accurate differentiation between benign and malignant normal-sized pelvic lymph nodes. Fifty-eight patients with bladder or prostate cancer were examined on a 3 T MR unit and their respective lymph node signal intensities (SI), signal-to-noise (SNR) and contrast-to-noise (CNR) were determined on pre- and post-contrast 3D T(2)-weighted turbo spin echo (TSE) images. Based on histology and/or localization, USPIO-uptake-related SI/SNR decrease of benign vs malignant and pelvic vs inguinal lymph nodes was compared. Out of 2182 resected lymph nodes 366 were selected for MRI post-processing. Benign pelvic lymph nodes showed a significantly higher SI/SNR decrease compared with malignant nodes (p < 0.0001). Inguinal lymph nodes in comparison to pelvic lymph nodes presented a reduced SI/SNR decrease (p < 0.0001). CNR did not differ significantly between benign and malignant lymph nodes. The receiver operating curve analysis yielded an area under the curve of 0.96, and the point with optimal accuracy was found at a threshold value of 13.5% SNR decrease. Overlap of SI and SNR changes between benign and malignant lymph nodes were attributed to partial voluming, lipomatosis, histiocytosis or focal lymphoreticular hyperplasia. USPIO-enhanced MRI improves the diagnostic ability of lymph node staging in normal-sized lymph nodes, although some overlap of SI/SNR-changes remained. Quantification of USPIO-dependent SNR decrease will enable the validation of this promising technique with the final goal of improving and individualizing patient care.