Temporal Parallelism Research Articles

Single breath-hold magnetic resonance cine imaging for fast assessment of global and regional left ventricular function in clinical routine

To evaluate a TGRAPPA (temporal parallel acquisition technique)-accelerated, single breath-hold multi-slice cine imaging approach for the assessment of left ventricular (LV) function. One hundred eleven patients were examined at 1.5T. Cine imaging was performed with single-slice breath-hold acquisitions in short-axis orientation using a SSFP (TR 2.63ms, TE 1.12ms, FA 72°) sequence and a TGRAPPA SSFP (TR 2.66ms, TE 1.11ms, FA 72°, AF 3) sequence, which covered the entire LV in multiple short-axis slices during a single breath-hold. End-diastolic (EDV), end-systolic (ESV), stroke volumes (SV), ejection fraction (EF), muscle mass (MM) and regional wall motion were assessed for both data sets. Single breath-hold imaging was feasible in 108 patients. Excellent correlations were observed for all volumetric parameters derived from both data sets (all r > 0.97). While EDV and ESV showed marginally lower values for single breath-hold imaging (EDV: -1.6 ± 7.9ml; ESV: -1.8 ± 6.0ml, p < 0.05), no differences were observed for SV, EF, MM and regional wall motion assessment. Single breath-hold imaging required significant shorter acquisition times (28 ± 6s vs. 335 ± 87s). TGRAPPA-accelerated multi-slice SSPF imaging allows for fast and accurate assessment of regional and global LV function within a single breath-hold.

Eocene Mollusca from the Tainohata Formation in Kôbe City, Southwest Japan

Abstract. The Tertiary Tainohata Formation of the Kobe Group has yielded 10 species of Gastropoda and 19 species of Bivalvia. The formation is of late Middle Eocene age on the basis of the dinoflagellate cyst assemblage and radiometric data. The occurrences of the genera Trinacria, Vicarya, and Sulcobuccinum strongly support an Eocene age for the Tainohata Formation. The molluscan fauna shows high endemism at the species level, whereas it is characterized by Tethyan—Indo-West Pacific genera. Two indigenous molluscan assemblages, Potamides-Cyclina and Corbicula, were discriminated. The composition of these assemblages compares with those of the Neogene intertidal assemblages in embayments. This fact indicates that “temporal parallelism” in the shallow embayment assemblages can be traced back to the late Middle Eocene age in the Japanese Islands. The following 10 new species are proposed herein: Potamides huzitai sp. nov., Cerithideopsilla hondai sp. nov., Crepidula tainohataensis sp. nov., Cantharus (s.l.)...

Guide-point modeling for the assessment of left ventricular function: comparison with the standard summation of slices method

Methods 52 consecutive patients were examined on a 1.5 T scanner with the standard SSFP(trueFISP, TR, 3.0 ms; TE, 1.5 ms; temporal resolution, 36 ms) and a highly accelerated, single breath-hold temporal parallel acquisition SSFPsequence (TR, 4.6 ms; TE, 1.1 ms; temporal resolution, 40 ms). The standard SSFP-sequence was post-processed both with the standard SoS-approach and the new GPMapproach, which relies on a 4-dimensional model of the LV and requires longand short-axis views for analysis. The highly accelerated sequence was solely evaluated with the GPM-approach. Thus, in each patient ejection fraction (EF), end-diastolic volume (EDV), and end-systolic volume (ESV) was calculated using three different approaches and results were compared by applying various statistical tests.

A Reconfigurable Architecture for Rotation Invariant Multi-View Face Detection Based on a Novel Two-Stage Boosting Method

We present a reconfigurable architecture model for rotation invariant multi-view face detection based on a novel two-stage boosting method. A tree-structured detector hierarchy is designed to organize multiple detector nodes identifying pose ranges of faces. We propose a boosting algorithm for training the detector nodes. The strong classifier in each detector node is composed of multiple novelly designed two-stage weak classifiers. With a shared output space of multicomponents vector, each detector node deals with the multidimensional binary classification problems. The design of the hardware architecture which fully exploits the spatial and temporal parallelism is introduced in detail. We also study the reconfiguration of the architecture for finding an appropriate tradeoff among the hardware implementation cost, the detection accuracy, and speed. Experiments on FPGA show that high accuracy and marvelous speed are achieved compared with previous related works. The execution time speedups range from 14.68 to 20.86 for images with size of 160×120 up to 800×600 when our FPGA design (98 MHz) is compared with software solution on PC (Pentium 4 2.8 GHz).

Rapid MR assessment of left ventricular systolic function after acute myocardial infarction using single breath-hold cine imaging with the temporal parallel acquisition technique (TPAT) and 4D guide-point modelling analysis of left ventricular function

We compared four-dimensional guide-point modelling left ventricular function analysis (4DVF) results of cine images in four short-axis and two long-axis slices acquired in a single breath-hold, obtained with the temporal parallel acquisition technique (TPAT), with standard left ventricular function (LVF) analysis results determined by the summation of discs method, in patients who had recently suffered myocardial infarction. Despite wall motion abnormalities, 4DVF yields results for left ventricular ejection fractions and end-diastolic and end-systolic volumes that are in excellent agreement with standard LVF analysis results in these patients. A shortened cardiac magnetic resonance (CMR) protocol using single breath-hold cine image acquisition could facilitate the assessment of left ventricular function soon after myocardial infarction in critically ill patients who are unable to comply with the multiple breath-holds required for standard LVF analysis.

Placing pipeline stages on a Grid: Single path and multipath pipeline execution

In a Grid computing environment, several applications such as scientific data analysis and visualization are naturally computation and communication intensive. These applications can be decomposed into a sequence of pipeline stages which can be placed on different Grid nodes for concurrent execution. Due to the aggregation of the computation and communication costs involved, finding the way to place such pipeline stages on a Grid in order to achieve the maximum application throughput becomes a challenging problem. This paper proposes a solution that considers both the pipeline placement and the data movement between stages. Specifically, we try to minimize the computation cost of the pipeline stages while preventing the communication overhead between successive stages from dominating the entire processing time. Our proposed solution consists of two novel methods. The first method is single path pipeline execution, which exploits only temporal parallelism, and the second method is multipath pipeline execution, which considers both temporal and spatial parallelism inherent in any pipeline applications. We evaluate our work in a simulated environment and also conduct a set of experiments in a real Grid computing system. When compared with the results from several traditional placement methods, our proposed methods give the highest throughput.

Assessment of left ventricular function with single breath-hold highly accelerated cine MRI combined with guide-point modeling

Purpose To prospectively assess the performance of highly accelerated cine MRI in multi-orientations combined with a new guide-point modeling post-processing technique (GPM approach) for assessment of left ventricular (LV) function compared to the standard summation of slices method based on a stack of short axis views (SoS approach). Materials and methods 33 consecutive patients were examined on a 1.5 T scanner with a standard steady state free precession (SSFP) sequence (TR, 3.0 ms; TE, 1.5 m; flip angle (FA), 60°; acceleration factor (AF), 2) analyzed with the SoS method and a highly accelerated, single breath-hold temporal parallel acquisition SSFP sequence (TR, 4.6 ms; TE, 1.1 ms; AF, 3) post-processed with the GPM method. LV function values were measured by two independent readers with different experience in cardiac MRI and compared by using the paired t-test and F-test. Inter- and intraobserver agreements were calculated using Bland–Altman-Plots. Results Mean acquisition and post-processing time was significantly shorter with the GPM approach (15 s/3 min versus 360 s/6 min). For all LV function parameters interobserver agreement between the experienced and non-experienced reader was significantly improved when the GPM approach was used. However, end-diastolic and end-systolic volumes were larger for the GPM technique when compared to the SoS method ( P < 0.001), whereas ejection fraction estimation yielded equivalent results ( P > 0.121). In both readers and for all parameters variances did not differ significantly ( P ≥ 0.409) and the two approaches showed an excellent linear correlation ( r > 0.951). Conclusion Due to its accurate, fast and reproducible assessment of LV function parameters highly accelerated MRI combined with the GPM technique may become the technique of first choice for assessment of LV function in clinical routine.

Rapid MR assessment of left ventricular systolic function early after acute myocardial infarction using single breath-hold cine imaging with temporal parallel acquisition technique (TPAT) and 4D guide-point modeling

Rapid MR assessment of left ventricular systolic function early after acute myocardial infarction using single breath-hold cine imaging with temporal parallel acquisition technique (TPAT) and 4D guide-point modeling

Control of population transfer in a multilevelLi2molecule by stimulated hyper-Raman nonadiabatic passage with chirped laser pulses

We have theoretically studied the population transfer in a five-level ${\mathrm{Li}}_{2}$ dimer by $(2+2)$-photon stimulated hyper-Raman nonadiabatic passage (STIHRNAP) from the initial ground level ${v}_{g}=0$, ${J}_{g}=0$ to the final rovibrational levels ${v}_{f}=1$ (2), ${J}_{f}=0,2$, of the ground electronic state $X\phantom{\rule{0.2em}{0ex}}^{1}\ensuremath{\Sigma}_{g}^{+}$ via the resonant intermediate levels ${v}_{i}=1$, ${J}_{i}=0,2$, of the excited electronic state $2\phantom{\rule{0.2em}{0ex}}^{1}\ensuremath{\Sigma}_{g}^{+}$. Linearly chirped pump and Stokes laser pulses with different chirp rates and without any initial detuning are applied simultaneously. Both the pulses are taken to have the same temporal shape, pulse width, and linear parallel polarizations. The density matrix method has been used to compute the populations of the levels. We have investigated in detail the population transfer for laser wavelengths (at time $t=0$) in the range of $992\char21{}1028\phantom{\rule{0.3em}{0ex}}\mathrm{nm}$ and (peak) intensities in the range of $1.5\ifmmode\times\else\texttimes\fi{}{10}^{10}\char21{}3.0\ifmmode\times\else\texttimes\fi{}{10}^{11}\phantom{\rule{0.3em}{0ex}}\mathrm{W}∕{\mathrm{cm}}^{2}$. The required pulse widths are of the order of $35\char21{}275\phantom{\rule{0.3em}{0ex}}\mathrm{ps}$ for maximum population inversion. We have controlled rotational branching in population transfer to the final rotational levels ${J}_{f}=0$ ($Q$ branch) and ${J}_{f}=2$ ($S$ branch) of the fundamental $({v}_{f}=1)$ and first overtone $({v}_{f}=2)$ transitions by judicious choice of laser parameters. We have applied a $(2+2)$-photon STIHRNAP process to a model multilevel molecular system $({\mathrm{Li}}_{2})$ and achieved almost complete population transfer from the initial ground to the final target rovibrational levels with chirped ir laser pulses.

Resource management and task partitioning and scheduling on a run-time reconfigurable embedded system

There are many design challenges in the hardware-software co-design approach for performance improvement of data-intensive streaming applications with a general-purpose microprocessor and a hardware accelerator. These design challenges are mainly to prevent hardware area fragmentation to increase resource utilization, to reduce hardware reconfiguration cost and to partition and schedule the tasks between the microprocessor and the hardware accelerator efficiently for performance improvement and power savings of the applications. In this paper a modular and block based hardware configuration architecture named memory-aware run-time reconfigurable embedded system (MARTRES) is proposed for efficient resource management and performance improvement of streaming applications. Subsequently we design a task placement algorithm named hierarchical best fit ascending (HBFA) algorithm to prove that MARTRES configuration architecture is very efficient in increased resource utilization and flexible in task mapping and power savings. The time complexity of HBFA algorithm is reduced to O( n) compared to traditional Best Fit (BF) algorithm’s time complexity of O( n 2), when the quality of the placement solution by HBFA is better than that of BF algorithm. Finally we design an efficient task partitioning and scheduling algorithm named balanced partitioned and placement-aware partitioning and scheduling algorithm (BPASA). In BPASA we exploit the temporal parallelism in streaming applications to reduce reconfiguration cost of the hardware, while keeping in mind the required throughput of the output data. We balance the exploitation of spatial parallelism and temporal parallelism in streaming applications by considering the reconfiguration cost vs. the data transfer cost. The scheduler refers to the HBFA placement algorithm to check whether contiguous area on FPGA is available before scheduling the task for HW or for SW.

A space–time-ensemble parallel nudged elastic band algorithm for molecular kinetics simulation

A scalable parallel algorithm has been designed to study long-time dynamics of many-atom systems based on the nudged elastic band method, which performs mutually constrained molecular dynamics simulations for a sequence of atomic configurations (or states) to obtain a minimum energy path between initial and final local minimum-energy states. A directionally heated nudged elastic band method is introduced to search for thermally activated events without the knowledge of final states, which is then applied to an ensemble of bands in a path ensemble method for long-time simulation in the framework of the transition state theory. The resulting molecular kinetics (MK) simulation method is parallelized with a space–time-ensemble parallel nudged elastic band (STEP-NEB) algorithm, which employs spatial decomposition within each state, while temporal parallelism across the states within each band and band-ensemble parallelism are implemented using a hierarchy of communicator constructs in the Message Passing Interface library. The STEP-NEB algorithm exhibits good scalability with respect to spatial, temporal and ensemble decompositions on massively parallel computers. The MK simulation method is used to study low strain-rate deformation of amorphous silica.

Parallel Simulation Study of a Laser Surface Alloying Process

ABSTRACT Numerical simulation of transient phase-change processes is massively time consuming, and can benefit from parallel processing to reduce the computational costs. In the present work, an implicit finite-volume-based three-dimensional, transient, turbulent model is developed to describe the transport phenomena for a typical laser surface alloying process, and the code is subsequently parallelized for effective computational economy. Parallelization of the code is done following the pipelining method, which uses temporal parallelism in a distributed computing environment and runs on a cluster of workstations (COW) connected with local area networking (LAN). Equal partitioning of the spatial domain into subdomains along the longer geometric dimension is performed, in order to decompose the computational domain. The Message Passing Interface (MPI) utility is used for communications among the subdomains on each processor. To demonstrate the parallel performance, a systematic analysis is carried out, with up to 16 nodes and various grid sizes. It is found that the overall parallel performance improves up to a certain number of processors, and then degrades with an increase in number of processors, due to communication delays in the LAN and pipeline hazards.

A High Speed Fuzzy Inference Processor with Dynamic Analysis and Scheduling Capabilities

The most obvious architectural solution for high-speed fuzzy inference is to exploit the temporal parallelism and spatial parallelism inherited in a fuzzy inference execution. However, the active rules in a fuzzy inference execution are often only a small part of the total rules. In this paper, we present a new architecture, which uses less hardware resources by discarding non-active rules in the earlier pipeline stage. Implementation data demonstrates that the proposed architecture achieves very good results in terms of the inference speed and the chip area.

An efficient VLSI implementation of IDEA encryption algorithm using VHDL

Data security is an important issue in computer networks and cryptographic algorithms are essential parts in network security. So far, International Data Encryption Algorithm (IDEA) is very secure. In this paper, we present a VLSI implementation of the IDEA block cipher using VHDL using AMI 0.5 process technology standard cells. We have optimized the modulus multiplier and exploited the temporal parallelism available in the IDEA algorithm. In our implementation, the subkeys are generated internally once the original key is fetched. This key is retained unless a new key is used for encryption. This implementation does not employ an additional RAM to store the subkeys. Our chip contains the same eight units, and each unit can execute one round of the algorithm. Using pipelined design, eight rounds of the algorithm are executed in parallel in a chip. Our implementation operating at 10 MHz achieves a throughput of greater than 700 Mbps, which is several times higher than previous implementations.

Rank order decoding of temporal parallel fibre input patterns in a complex Purkinje cell model

The processing speed of many neuronal systems requires temporal coding. Recently, a temporal rank order code has been suggested that uses the temporal order of spikes, disregarding their precise timing. A rank order-coded spike pattern can be decoded by an array of synaptic weights and a postsynaptic desensitization process. We show that a multi-compartmental model of a cerebellar Purkinje cell can implement rank order decoding of temporal parallel fibre input patterns. Basis of the temporal decoding is the activation of K Ca channels in the Purkinje cell dendrites. The model responds preferentially to spatio-temporal patterns which are ordered according to increasing synaptic strengths.

Mechanisms of conjoint failure of the somatotropic and gonadal axes in ageing men.

Endogenous growth hormone (GH) production falls by 50% every 7 years and bioavailable testosterone concentrations decline concomitantly by 12-15% every decade in ageing men. Despite this temporal parallelism, the neuroendocrine bases of the somatopause and gonadopause are not known. This knowledge deficit contrasts with the recent unfolding of new insights into the nature of oestrogen-dependent control of the GH-insulin-like growth factor (IGF)1 axis in pre- and postmenopausal women. The present overview examines the postulate that the pathophysiology of somatopause and gonadopause in ageing men is bidirectionally linked. According to this broader thesis, hyposomatotropism accentuates Leydig cell steroidogenic failure and, conversely, progressive androgen deficiency exacerbates the decline in GH-IGF1 output in ageing.

Spatial and temporal data parallelization of the H.261 video coding algorithm

The parallelization of the H.261 video coding algorithm on the IBM SP2(R) multiprocessor system is described. The effect of parallelizing computations and communications in the spatial, temporal, and both spatial-temporal domains are considered through the study of frame rate, speedup, and implementation efficiency, which are modeled and measured with respect to the number of nodes (n) and parallel methods used. Four parallel algorithms were developed, of which the first two exploited the spatial parallelism in each frame, and the last two exploited both the temporal and spatial parallelism over a sequence of frames. The two spatial algorithms differ in that one utilizes a single communication master, while the other attempts to distribute communications across three masters. On the other hand, the spatial-temporal algorithms use a pipeline structure for exploiting the temporal parallelism together with either a single master or multiple masters. The best median speedup (frame rate) achieved was close to 15 [15 frames per second (fps)] for 352/spl times/240 video on 24 nodes, and 13 (37 fps) for QCIF video, by the spatial algorithm with distributed communications. For n<10, the single-master spatial algorithm performs better with efficiency up to 90%, while the multiple-master spatial algorithm is superior for n>10, with efficiency up to 70%. The spatial-temporal algorithms achieved average speedup performance, but are most scalable for large n.

Protein kinase C-like immunoreactive cells in embryo and adult chicken retinas

Morphological evidence of a temporal parallelism between the appearance of the α isoform of protein kinase C (PKC) and some processes such as synaptogenesis in the plexiform layers of the chicken retina is offered. Immunostaining experiments were performed throughout embryonic, young and adult chicken life. The results help to understand the development of rod bipolar cells.

Music, rhythm, and development: Chairs’ introduction

This special session is sponsored by the Caroline B. Monahan Fund and by the Center for Cognitive Science at Ohio State University. Caroline Monahan’s contributions to the field of music cognition focus on auditory perception and memory, including topics of pitch and temporal parallels in music and speech perception, recognition memory for rhythmic and pitch patterns, and the role of timbral and loudness variables in discrimination of auditory patterns. Dr. Monahan’s work has been concerned with applications of these topics in speech as well as in music, including language acquisition, hearing impairment, and second-language learning. Dr. Monahan received her Ph.D. at UCLA in 1984 in Psychology, and served as a Research Scientist at the Central Institute for the Deaf. She also served as a Research Associate in the Department of Communication Disorders at the University of Oklahoma. We acknowledge her contributions to and generous support of the field of music cognition.

Adaptive leaky integrator models of cerebellar Purkinje cells can learn the clustering of temporal patterns

We have shown previously that the metabotropic glutamate receptor signalling network in a cerebellar Purkinje cell can implement adaptive postsynaptic delays. Here we present a leaky integrator version of the Purkinje cell model which uses a simple synaptic delay learning rule. We show that a single leaky integrator can learn a radial basis function-like response to temporal parallel fibre patterns, and that different leaky integrators in a group are able to discover different clusters in a temporal parallel fibre input space. The clustering performance of the model can be improved by desensitization of the input currents.