Search

The success of deep neural networks suggests that cognition may emerge from indecipherable patterns of distributed neural activity. Yet these networks are pattern-matching black boxes that cannot simulate higher cognitive functions and lack numerous neurobiological features. Accordingly, they are currently insufficient computational models for understanding neural information processing. Here, we show how neural circuits can directly encode cognitive processes via simple neurobiological principles. To illustrate, we implemented this model in a non-gradient-based machine learning algorithm to train deep neural networks called essence neural networks (ENNs). Neural information processing in ENNs is intrinsically explainable, even on benchmark computer vision tasks. ENNs can also simulate higher cognitive functions such as deliberation, symbolic reasoning and out-of-distribution generalization. ENNs display network properties associated with the brain, such as modularity, distributed and localist firing, and adversarial robustness. ENNs establish a broad computational framework to decipher the neural basis of cognition and pursue artificial general intelligence. The authors demonstrate how neural systems can encode cognitive functions, and use the proposed model to train robust, scalable deep neural networks that are explainable and capable of symbolic reasoning and domain generalization.

Distributed in-memory caching is a key component of modern Internet services. Such caches are often accessed via remote procedure call (RPC), as RPC frameworks provide rich support for productionization, including protocol versioning, memory efficiency, auto-scaling, and hitless upgrades. However, full-featured RPC limits performance and scalability as it incurs high latencies and CPU overheads. Remote Memory Access (RMA) offers a promising alternative, but meeting productionization requirements can be a significant challenge with RMA-based systems due to limited programmability and narrow RMA primitives. This paper describes the design, implementation, and experience derived from CliqueMap, a hybrid RMA/RPC caching system. CliqueMap has been in production use in Google's datacenters for over three years, currently serves more than 1PB of DRAM, and underlies several end-user visible services. CliqueMap makes use of performant and efficient RMAs on the critical serving path and judiciously applies RPCs toward other functionality. The design embraces lightweight replication, client-based quoruming, self-validating server responses, per-operation client-side retries, and co-design with the network layers. These foci lead to a system resilient to the rigors of production and frequent post deployment evolution.

The guest editors introduce the Top Picks and Honorable Mentions from the 2015 computer architecture conferences.

The lack of memory safety in languages such as C and C++ is a root source of exploitable security vulnerabilities. This article presents Watchdog, a hardware approach that eliminates such vulnerabilities by enforcing comprehensive memory safety. Inspired by prior software-only mechanisms, Watchdog maintains bounds and identifier metadata with pointers, propagates them on pointer operations, and checks them on pointer dereferences. Checking this bounds and identifier metadata provides both precise, byte-granularity buffer-overflow protection and protection from use-after-free errors, even in the presence of reallocations. Watchdog stores pointer metadata in a disjoint shadow space to provide comprehensive protection and ensure compatibility with existing code. To streamline implementation and reduce runtime overhead, Watchdog uses micro-operations to implement metadata access and checking, eliminates metadata copies via a register renaming scheme, and uses a dedicated identifier cache to reduce checking overhead.

The tight thermal constraints of mobile devices, which limit sustainable performance, and the bursty nature of interactive mobile applications call for a new design focus: enhancing user responsiveness rather than sustained throughput. To that end, this article explores computational sprinting, wherein a mobile device temporarily exceeds sustainable thermal limits to provide a brief, intense burst of computation in response to user input. By enabling tenfold more computation within the timescale of human patience, sprinting has the potential to fundamentally change the user experience that an interactive application can provide.

Languages such as C and C++ use unsafe manual memory management, allowing simple bugs (i.e., accesses to an object after deallocation) to become the root cause of exploitable security vulnerabilities. This paper proposes Watchdog, a hardware-based approach for ensuring safe and secure manual memory management. Inspired by prior software-only proposals, Watchdog generates a unique identifier for each memory allocation, associates these identifiers with pointers, and checks to ensure that the identifier is still valid on every memory access. This use of identifiers and checks enables Watchdog to detect errors even in the presence of reallocations. Watchdog stores these pointer identifiers in a disjoint shadow space to provide comprehensive protection and ensure compatibility with existing code. To streamline the implementation and reduce runtime overhead: Watchdog (1) uses micro-ops to access metadata and perform checks, (2) eliminates metadata copies among registers via modified register renaming, and (3) uses a dedicated metadata cache to reduce checking overhead. Furthermore, this paper extends Watchdog's mechanisms to detect bounds errors, thereby providing full hardware-enforced memory safety at low overheads.

The world's offshore continental margins contain vast reserves of gas hydrate, a frozen form of natural gas that is embedded in cold, near-seafloor strata. Published estimates suggest that the energy represented by gas hydrate may exceed the energy available from conventional fossil fuel by a factor of 2 or more. Understanding marine hydrate systems has become critical for long-term worldwide energy planning. Groups in several nations are attempting to evaluate the resource and to define seafloor stability problems across hydrate accumulations.

We discuss, in a two-part article, the benefits of 90°-phase wavelets in stratigraphic and lithologic interpretation of seismically thin beds. In Part 1, seismic models of Ricker wavelets with selected phases are constructed to assess interpretability of composite waveforms in increasingly complex geologic settings. Although superior for single surface and thick-layer interpretation, zero-phase seismic data are not optimal for interpreting beds thinner than a wavelength because their antisymmetric thin-bed responses tie to the reflectivity series rather than to impedance logs. Nonsymmetrical wavelets (e.g., minimum-phase wavelets) are generally not recommended for interpretation because their asymmetric composite waveforms have large side lobes. Integrated zero-phase traces are also less desirable because they lose high-frequency components in the integration process. However, the application of 90°-phase data consistently improves seismic interpretability. The unique symmetry of 90°-phase thin-bed response eliminates the dual polarity of thin-bed responses, resulting in better imagery of thin-bed geometry, impedance profiles, lithology, and stratigraphy. Less amplitude distortion and less stratigraphy-independent, thin-bed interference lead to more accurate acoustic impedance estimation from amplitude data and a better tie of seismic traces to lithology-indicative wireline logs. Field data applications are presented in part 2 of this article.

Multicomponent seismic data provide the potential to better characterize the subsurface than do P-wave data alone. Converted waves (C-waves) and pure shear waves can be used to validate bright spot reflections, provide Poisson's ratio estimates, provide reflection images where the P-wave reflectivity contrast is small, image through gas clouds where the P-wave signal is attenuated/scattered, and detect/characterize fractures through shear-wave splitting.

As dependence on database management systems and Web servers increases, so does the need for them to run reliably and efficiently-goals that rigorous simulations can help achieve. Execution-driven simulation models system hardware. These simulations capture actual program behavior and detailed system interactions. The authors have developed a simulation methodology that uses multiple simulations, pays careful attention to the effects of scaling on workload behavior, and extends Virtutech AB's Simics full system functional simulator with detailed timing models. The Wisconsin Commercial Workload Suite contains scaled and tuned benchmarks for multiprocessor servers, enabling full-system simulations to run on the PCs that are routinely available to researchers.

We develop a specification methodology that documents and specifies a cache coherence protocol in eight tables: the states, events, actions, and transitions of the cache and memory controllers. We then use this methodology to specify a detailed, modern three-state broadcast snooping protocol with an unordered data network and an ordered address network that allows arbitrary skew. We also present a detailed specification of a new protocol called multicast snooping (Bilir et al., 1999) and, in doing so, we better illustrate the utility of the table-based specification methodology. Finally, we demonstrate a technique for verification of the multicast snooping protocol, through the sketch of a manual proof that the specification satisfies a sequentially consistent memory model.

We develop an availability solution, called SafetyNet, that uses a unified, lightweight checkpoint/recovery mechanism to support multiple long-latency fault detection schemes. At an abstract level, SafetyNet logically maintains multiple, globally consistent checkpoints of the state of a shared memory multiprocessor (i.e., processors, memory, and coherence permissions), and it recovers to a pre-fault checkpoint of the system and re-executes if a fault is detected. SafetyNet efficiently coordinates checkpoints across the system in logical time and uses "logically atomic" coherence transactions to free checkpoints of transient coherence state. SafetyNet minimizes performance overhead by pipelining checkpoint validation with subsequent parallel execution.We illustrate SafetyNet avoiding system crashes due to either dropped coherence messages or the loss of an interconnection network switch (and its buffered messages). Using full-system simulation of a 16-way multiprocessor running commercial workloads, we find that SafetyNet (a) adds statistically insignificant runtime overhead in the common-case of fault-free execution, and (b) avoids a crash when tolerated faults occur.

Lack of memory safety in C is the root cause of a multitude of serious bugs and security vulnerabilities. Numerous software-only and hardware-based schemes have been proposed to enforce memory safety. Among these approaches, pointer-based checking, which maintains per-pointer metadata in a disjoint metadata space, has been recognized as providing comprehensive memory safety. Software approaches for pointer-based checking have high performance overheads. In contrast, hardware approaches introduce a myriad of hardware structures and widgets to mitigate those performance overheads.This paper proposes WatchdogLite, an ISA extension that provides hardware acceleration for a compiler implementation of pointer-based checking. This division of labor between the compiler and the hardware allows for hardware acceleration while using only preexisting architectural registers. By leveraging the compiler to identify pointers, perform check elimination, and insert the new instructions, this approach attains performance similar to prior hardware-intensive approaches without adding any hardware structures for tracking metadata.

Nine-component (9-C) 3-D seismic reflection data are acquired using orthogonal shear-wave sources and orthogonal horizontal geophones. Shear-wave sources are oriented inline and crossline (SI and SX) to the receiver lines, as are the horizontal geophones (RI and RX). Four shear-wave data sets result (SI,RI; SI,RX; SX,RI,SX,RX). These inline-crossline coordinates are referred to as field coordinates. A main interpretation emphasis for these data has been to infer the presence of vertical cracks using shear-wave birefringence. Nonzero crossterms (SI,RX; SX,RI) have been used as an indicator that shear-wave birefringence (splitting) is present.

Lung clearance of 99mTc-labeled diethylenetriamine pentaacetate (DTPA) is a sensitive test of altered alveolar epithelial permeability that has been found to be increased in smokers of tobacco, as well as a small number of healthy smokers of crack cocaine, suggesting the possibility of subclinical crack-related lung injury.To evaluate further whether habitual smoking of cocaine alone alters alveolar permeability, whether crack smoking adds to or potentiates the effects of tobacco and/or marijuana, and whether experimental cocaine smoking acutely alters DTPA lung clearance.Observational cohort study (habitual cocaine smoking) and single-blind crossover study (experimental cocaine administration).Fourteen habitual smokers of cocaine alone (CS), 19 smokers of cocaine and tobacco (CTS), 3 smokers of cocaine and marijuana, 12 smokers of cocaine, tobacco, and marijuana (CMTS), and 5 smokers of marijuana plus tobacco (MTS). Results obtained in the crack-smoking subjects were compared with data previously obtained in 10 nonsmokers (NS), 9 smokers of tobacco alone (TS), 10 smokers of marijuana alone (MS), and 4 additional MTS.Subjects underwent measurements of DTPA radioaerosol lung clearance after refraining from marijuana and/or cocaine for12 h and from tobacco for2 h. Ten of the 48 crack users were tested on two days 1 to 2 weeks apart within 2 h of experimental smoking of three physiologically active or inactive doses (total 98.8+/-15.5 or 8.5+/-2.5 mg, respectively) of cocaine base. Lung clearance half-times (T1/2) were computed from time-activity curves for each lung.T1/2 values for each lung in CS and MS were comparable to those of NS, while TS, MTS, CTS, and CMTS had significantly shorter clearance rates than NS (p0.01; three-way analysis of variance). No additive or interactive effects on T1/2 were noted among tobacco, cocaine, and/or marijuana. No acute effect of experimental cocaine smoking on T1/2 was noted.Whereas regular smoking of tobacco alone or with other substances increases alveolar epithelial permeability, habitual smoking of cocaine and/or marijuana has no measurable effect on alveolar permeability in the absence of tobacco nor any additive effect to that of tobacco alone.

CMOS scaling trends have led to an inflection point where thermal constraints (especially in mobile devices that employ only passive cooling) preclude sustained operation of all transistors on a chip --- a phenomenon called "dark silicon." Recent research proposed computational sprinting --- exceeding sustainable thermal limits for short intervals --- to improve responsiveness in light of the bursty computation demands of many media-rich interactive mobile applications. Computational sprinting improves responsiveness by activating reserve cores (parallel sprinting) and/or boosting frequency/voltage (frequency sprinting) to power levels that far exceed the system's sustainable cooling capabilities, relying on thermal capacitance to buffer heat. Prior work analyzed the feasibility of sprinting through modeling and simulation. In this work, we investigate sprinting using a hardware/software testbed. First, we study unabridged sprints, wherein the computation completes before temperature becomes critical, demonstrating a 6.3x responsiveness gain, and a 6% energy efficiency improvement by racing to idle. We then analyze truncated sprints, wherein our software runtime system must intervene to prevent overheating by throttling parallelism and frequency before the computation is complete. To avoid oversubscription penalties (context switching inefficiencies after a truncated parallel sprint), we develop a sprint-aware task-based parallel runtime. We find that maximal-intensity sprinting is not always best, introduce the concept of sprint pacing, and evaluate an adaptive policy for selecting sprint intensity. We report initial results using a phase change heat sink to extend maximum sprint duration. Finally, we demonstrate that a sprint-and-rest operating regime can actually outperform thermally-limited sustained execution.

A practical approach to linear prestack seismic inversion in the context of a locally 1-D earth is employed to use amplitude variation with offset (AVO) information for the direct detection in hydrocarbons. The inversion is based on the three‐term linearized approximation to the Zoeppritz equations. The normal‐incidence compressional‐wave reflection coefficient [Formula: see text] models the background reflectivity in the absence of hydrocarbons and incorporates the mudrock curve and Gardner’s equation. Prediction‐error parameters, [Formula: see text] and [Formula: see text], represent perturbations in the normal‐incidence shear‐wave reflection coefficient and the density contribution to the normal incidence reflectivity, respectively, from that predicted by the mudrock curve and Gardner’s equation. This prediction‐error approach can detect hydrocarbons in the absence of an overall increase in AVO, and in the absence of bright spots, as expected in theory. Linear inversion is applied to a portion of a young, Tertiary, shallow‐marine data set that contains known hydrocarbon accumulations. Prestack data are in the form of angle stack, or constant offset‐to‐depth ratio, gathers. Prestack synthetic seismograms are obtained by primaries‐only ray tracing using the linearized approximation to the Zoeppritz equations to model the reflection amplitudes. Where the a priori assumptions hold, the data are reproduced with a single parameter [Formula: see text]. Hydrocarbons are detected as low impedance relative to the surrounding shales and the downdip brine‐filled reservoir on [Formula: see text], also as positive perturbations (opposite polarity relative to [Formula: see text]) on [Formula: see text] and [Formula: see text]. The maximum perturbation in [Formula: see text] from the normal‐incidence shear‐wave reflection coefficient predicted by the a priori assumptions is 0.08. Hydrocarbon detection is achieved, although the overall seismic response of a gas‐filled thin layer shows a decrease in amplitude with offset (angle). The angle‐stack data (70 prestack ensembles, 0.504–1.936 s time range) are reproduced with a data residual that is 7 dB down. Reflectivity‐based prestack seismograms properly model a gas/water contact as a strong increase in AVO and a gas‐filled thin layer as a decrease in AVO.

We have evaluated the performance of an illicit substance detection system that performs image reconstruction using the maximum likelihood algebraic reconstruction algorithm, a few number of projections, and relatively coarse projection and pixel resolution. This evaluation was done using receiver operator curves and simulated data from the fast-neutron transmission spectroscopy system operated in a mode to detect explosives in luggage. The results show that increasing the number of projection angles is more important than increasing the projection resolution, the reconstructed pixel resolution, or the number of iterations in the maximum likelihood algorithm. A 100% detection efficiency with essentially no false positives is possible for a square block of RDX explosive, a projection resolution of 2 cm, a reconstructed pixel size of 2/spl times/2 cm, and five projection angles. For rectangular shaped explosives more angles are required to obtain the same system performance.

International audience; The growing complexity of hardware optimizations employed by multiprocessors leads to subtle distinctions among allowed and dis-allowed behaviors, posing challenges in specifying their memory models formally and accurately, and in understanding and analyzing the behav-ior of concurrent software. This complexity is particularly evident in the IBM R Power Architecture R , for which a faithful specification was pub-lished only in 2011 using an operational style. In this paper we present an equivalent axiomatic specification, which is more abstract and con-cise. Although not officially sanctioned by the vendor, our results indicate that this axiomatic specification provides a reasonable basis for reasoning about current IBM R POWER R multiprocessors. We establish the equiva-lence of the axiomatic and operational specifications using both manual proof and extensive testing. To demonstrate that the constraint-based style of axiomatic specification is more amenable to computer-aided ver-ification, we develop a SAT-based tool for evaluating possible outcomes of multi-threaded test programs, and we show that this tool is signifi-cantly more efficient than a tool based on an operational specification.

The clinical diagnosis of pulmonary embolism (PE) remains difficult despite years of investigation. The clinical signs and symptoms of pulmonary embolism are numerous, but they are not diagnostically accurate. Radionuclide ventilation perfusion imaging (VQ) has become an important adjunct in screening patients for suspected PE. This study re-evaluates the predictive capabilities of various clinical signs and symptoms in a population of patients in whom angiography was thought to be necessary because of clinical concern for PE. One hundred one patients with suspected PE were retrospectively identified. Clinical information on them was complete, and laboratory studies, VQ imaging, and pulmonary angiography were performed. Thirty-seven clinical signs, symptoms, and other characteristics were individually compared for patients whose angiographic results were positive or negative for PE. Modifying or interactive effects on PE were then examined with logistic regression analysis. Several clinical conditions, including immobilization and recent surgery, were significantly associated with PE. Shortness of breath and history of smoking were significantly associated with negative results on pulmonary angiography. A "high probability" VQ scan was highly predictive (P.0001) of positive results on angiography. When interactive factors for PE were examined, lack of shortness of breath and arterial pHor = 7.45, lack of shortness of breath and respiratory rate23, diaphoresis in a nonsmoker and immobilized female, were significantly associated with PE. This study again documents the difficulty in using clinical criteria--including signs, symptoms, and laboratory determinations--to predict PE accurately, even in a population in which PE was of clinical concern irrespective of the VQ scan results.(ABSTRACT TRUNCATED AT 250 WORDS)

EG&G Idaho, Inc. has initiated a program to develop safe, efficient, cost-effective treatment methods for the stabilization of some of the hazardous and mixed wastes generated at the Idaho National Engineering Laboratory.Laboratory-scale testing has shown that Extraction Procedure toxic wastes can be successfully stabilized by solidification, using various binders to produce nontoxic, stable waste forms for safe, long-term disposal as either landfill waste or low-level radioactive waste, depending upon the radioactivity content. This paper presents the results of drum-scale solidification testing conducted on hazardous, low-level incinerator fly ash generated at the Waste Experimental Reduction Facility. The drum-scale test program was conducted to verify that laboratory-scale results could be successfully adapted into a production operation.

The finite length of seismic arrays results in smeared estimates of the plane‐wave decomposition of the recorded wave field. The smearing is a fundamental property of plane‐wave decomposition; it can be reduced but not eliminated. If not accounted for, this smearing can have a significant effect upon the waveform inversion of plane‐wave seismograms. This paper illustrates the effects of limited spatial aperture on the plane‐wave decomposition and demonstrates that these truncation artifacts can be described by a linear ray parameter‐dependent filter. Experiments with synthetic data indicate that including the effects of limited aperture in the forward model of an iterative inversion produces a stable inversion result without significantly reducing the information content of the data. The method is illustrated with synthetic data derived for an earth consisting of elastic, isotropic, flat layers.

Background: The operative management of children with combined gastroesophageal reflux and delayed gastric emptying is controversial. This study measures the long-term follow-up of gastric emptying in children who have undergone gastroesophageal fundoplication combined with antroplasty. Methods: Fifteen randomly selected children with gastroesophageal reflux and scintigraphically demonstrated delayed gastric emptying underwent fundoplication and antroplasty. Each patient had another gastric emptying scintigraphic study performed an average of 3.6 years postoperation. Results: All patients reported improvement of their symptoms compared with before the operation, and none required further medical therapy for gastroesophageal reflux or experienced dumping syndrome. Eleven of the 15 patients had significant long-term improvement of their gastric emptying postoperatively. The mean percent of isotope meal remaining in the stomach at 90 minutes improved from 72% preoperatively to 40% postoperatively ( P = 0.0005). Conclusions: Gastric emptying in children with gastroesophageal reflux and delayed gastric emptying is significantly improved for several years in three-fourths of patients after fundoplication and antroplasty. Fundoplication and concomitant antroplasty are recommended for symptomatic children with documented gastroesophageal reflux and delayed gastric emptying.

Improvements in semiconductor technology now enable chip multiprocessors (CMPs). As many future computer systems will use one or more CMPs and support shared memory, such systems will have caches that must be kept coherent. Coherence is a particular challenge for multiple-CMP (M-CMP) systems. One approach is to use a hierarchical protocol that explicitly separates the intra-CMP coherence protocol from the inter-CMP protocol, but couples them hierarchically to maintain coherence. However, hierarchical protocols are complex, leading to subtle, difficult-to-verify race conditions. Furthermore, most previous hierarchical protocols use directories at one or both levels, incurring indirections - and thus extra latency - for sharing misses, which are common in commercial workloads. In contrast, this paper exploits the separation of correctness substrate and performance policy in the recently-proposed token coherence protocol to develop the first M-CMP coherence protocol that is flat for correctness, but hierarchical for performance. Via model checking studies, we show that flat correctness eases verification. Via simulation with micro-benchmarks, we make new protocol variants more robust under contention. Finally, via simulation with commercial workloads on a commercial operating system, we show that new protocol variants can be 10-50% faster than a hierarchical directory protocol.

Application of multicomponent seismic exploration produces multiple images of the same subsurface. For a successful interpretation of multicomponent images, it is crucially important to register them in the same coordinate frame. Accurate registration of time-domain images also provides an effective estimate of the interval VP/VS ratio, a major petrophysical attribute. We propose a multistep approach to image registration, which consists of initial interpretation, amplitude and frequency balancing, registration scan, and least-squares optimization. The approach is applied successfully to a 3C3D seismic dataset from West Texas, where multicomponent image registration provides an insight into a targeted carbonate reservoir facies.

This paper advocates that cache coherence protocols use a bandwidth adaptive approach to adjust to varied system configurations (e.g., number of processors) and workload behaviors. We propose Bandwidth Adaptive Snooping Hybrid (BASH), a hybrid protocol that ranges from behaving like snooping (by broadcasting requests) when excess bandwidth is available to behaving like a directory protocol (by unicasting requests) when bandwidth is limited. BASH adapts dynamically by probabilistically deciding to broadcast or unicast on a per request basis using a local estimate of recent interconnection network utilization. Simulations of a microbenchmark and commercial and scientific workloads show that BASH robustly performs as well or better than the best of snooping and directory protocols as available bandwidth is varied. By mixing broadcasts and unicasts, BASH outperforms both snooping and directory protocols in the mid-range where a static choice of either is inefficient.

We developed a methodology for manually establishing tie points of depth-equivalent surfaces in P-P and P-S seismic data volumes derived from a 4-C ocean bottom seismic survey using seismic attribute volumes viewed in time slices. These tie points were used as a basis for establishing interpretations of depth-equivalent surfaces throughout the volumes that were then used as a basis for depth registration of several 2-D sections throughout the volume. We examine the Vp/Vs ratios derived from this interpretation. While these ratios are physically reasonable, they are averaged over several stratigraphic sequences and do not provided enough detail to represent the true interval Vp/Vs values of a given sequence. However, we use these Vp/Vs ratios to correct the interpretation and perform an initial registration of the P-S volume to P-P two-way time. An analysis of the registered, or “warped”, P-S volume shows the limitations of this simple technique, and from this we infer what processes must be addressed for a robust method of registration. Introduction The work summarized here is part of a larger study undertaken to develop methods of multicomponent data analysis for the detection of gas hydrate prospects in the northern Gulf of Mexico. Methane gas hydrates deposits occur within a narrow window of near ocean-bottom strata extending no more than a few hundred meters below the seafloor, with size and extent dependent on local temperature and pressure conditions in the region. In many regions of North America, including the southern Gulf of Mexico (GOM), Alaska, and the Atlantic and Pacific coasts of the United States, known gas hydrate deposits can be identified on P-wave seismic surveys by a diagnostic bottom-simulating reflector (BSR) at the base of the frozen hydrates. In the northern GOM, gas hydrates do not exhibit the classic BSR. In this region, the known hydrate deposits have no known distinctive seismic signature, but they do tend to be associated with shallow “gas clouds” which show up as no-data or poor-data quality zones on traditional Pwave seismic surveys. Converted-mode shear waves have already been proven a useful tool for imaging through gas cloud regions (Thomsen et. al, 1997) and specifically in the northern GOM for the purpose of identifying deep targets of interest for oil and gas production (Cafarelli et al, 2000, Knapp et al, 2001). We consider the identification of shallow gas clouds for hydrate detection another possible application for this type of multicomponent seismic data. As part of this study, we introduced the use of a set of multicomponent seismic data designed for deep target exploration as a tool to aid in the identification of possible hydrate prospects and other possible engineering hazards in the upper 1000m of strata across a study region. It should be noted that this study area of the northern GOM shelf is not itself a candidate for hydrate deposits. Large regional 3-D multicomponent data sets of such environments in the deepwater GOM are not available to us at this time. However, the data provided to us for this study provide an opportunity to demonstrate the utility of multicomponent data for the purpose of analyzing nearseafloor sediments and elastic properties which can be extended to other deepwater environments. 4-C Seismic Data The data provided us were from a multi-client 4-C ocean bottom cable survey acquired over 46 OCS blocks of East Cameron South, offshore Louisiana in 1999 and 2000, covering approximately 1000km. The survey is located on the continental shelf in water depths of approximately 100m. The data were processed as a pure-mode P-wave (P-P) volume created from vertical geophone velocity and hydrophone pressure data, and as a radially rotated converted shear mode (P-S or C-wave) volume created from the horizontal velocity phones. For more details of the survey design and processing, refer to Knapp et. al (2001). We were given a P-P wave volume with an extent of 1s in two-way traveltime, and a P-S wave volume extending to 2s so as to allow us enough data to complete our shallow regional studies and protect the confidentiality of the deeper data currently in use for the location of commercial hydrocarbon prospects.

Destination-set prediction can improve the latency/bandwidth tradeoff in shared-memory multiprocessors. The destination set is the collection of processors that receive a particular coherence request. Snooping protocols send requests to the maximal destination set (i.e., all processors), reducing latency for cache-to-cache misses at the expense of increased traffic. Directory protocols send requests to the minimal destination set, reducing bandwidth at the expense of an indirection through the directory for cache-to-cache misses. Recently proposed hybrid protocols trade-off latency and bandwidth by directly sending requests to a predicted destination set.This paper explores the destination-set predictor design space, focusing on a collection of important commercial workloads. First, we analyze the sharing behavior of these workloads. Second, we propose predictors that exploit the observed sharing behavior to target different points in the latency/bandwidth tradeoff. Third, we illustrate the effectiveness of destination-set predictors in the context of a multicast snooping protocol. For example, one of our predictors obtains almost 90% of the performance of snooping while using only 15% more bandwidth than a directory protocol (and less than half the bandwidth of snooping).

The UCLA Hospital experience with indium-111 labeled white blood cells was reviewed. There were a total of 345 consecutive cases covering a broad range of clinical indications. The overall sensitivity of the method was 79%, specificity was 62%, and accuracy was 73%. The sensitivity for suspected osteomyelitis cases was 84%, with a specificity of 65% and an accuracy of 75%. For other cases sensitivity was 77%, specificity was 60%, and accuracy was 72%. Furthermore, patterns of normal distribution were reviewed.

You have heard about the energy crisis. “The easy oil has been found. We are about to run out of oil. Oil discovered per new field wildcat has been decreasing exponentially.” Does this mean there is no future for exploration geophysics? Quite the contrary. If it were easy to find new oil, there would be no need for exploration geophysics. If we weren’t concerned about running out of oil, we would logically postpone the search.

Macroaggregated albumin (MAA) entrapment in areas of fibrin deposits has been used to detect intravenous thrombosis in the lower extremities. This study reports the correlation of MAA uptake in thrombosis with venography and autologous 125I-labeled fibrinogen uptake in approximately 30 cases. Correlation was good in most patients who underwent at least two of these procedures. The MAA thrombosis scan appears especially accurate in locating thrombosis in the upper thigh or pelvis where fibrinogen uptake is less helpful. The correlation suggests that the MAA scan has very few false negative results. The procedure is easily performed in conjunction with a perfusion lung scan.

Offset‐dependent line‐intersection displays can be of significant direct interpretive value. On our data set from a proprietary location, use of these displays permits a refined structural interpretation and improves the usefulness of amplitude and amplitude‐versus‐offset displays for direct hydrocarbon detection. The degree of data reproducibility at each intersection may be exposed as a function of offset through the use of prestack and partial‐range stack splice and difference displays. At a set of marine line intersections, after only cursory processing, the range of residual amplitudes found by subtracting strike traces from dip traces is from −20 dB to +6 dB relative to the input. The average is about −8 dB. The poorest line intersection ties occur below the edges of deep bright spots, where velocity anomalies leading to time mis‐ties greater than 10 ms can occur. Shallow near‐surface velocity anomalies, out‐of‐plane arrivals including fault plane reflections, and, occasionally, water reverberation variations contribute locally to the data mis‐ties. Location errors are ubiquitous but of secondary importance, and random noise is more than 20 dB down. A simple and inexpensive routine splice display of line intersection data can be of significant value for quality control, for processing evaluation, and for interpretation. Issues of data reliability, such as screening of amplitude‐versus‐offset (AVO) anomalies by shallow gas and transmission‐path problems for large CMP offsets, as well as other problems in operational and research geophysics, may be addressed by analysis of line intersections in the offset domain.

Several particulate radiotracers were studied with respect to their behavior in the vicinity of clots. Experiments were performed in vitro, microscopically and quantitatively, and in vivo on artificially created lesions. Of the particles studied, macroaggregates of albumin and aggregates of stannous hydroxide showed the best accumulation on clots and venous lesions. It was concluded that there is definite demonstrable clot affinity for some particles and not for others. The mechanism remains to be elucidated. Clinical usefulness is to be evaluated.

A new data‐processing technique is presented for the separation of initially up‐traveling (ghost) energy from initially down‐traveling (primary) energy on reflection seismograms. The method combines records from two or more shot depths after prefiltering each record with a different filter. The filters are designed on a least‐mean‐square‐error criterion to extract primary reflections in the presence of ghost reflections and random noise. Filter design is dependent only on the difference in uphole time between shots, and is independent of the details of near‐surface layering. The method achieves wide‐band separation of primary and ghost energy, which results in 10–15 db greater attenuation of ghost reflections than can be achieved with conventional two‐ or three‐shot stacking (no prefiltering) for ghost elimination. The technique is illustrated in terms of both synthetic and field examples. The deghosted field data are used to study the near‐surface reflection response by computing the optimum linear filter to transform the deghosted trace back into the original ghosted trace. The impulse response of this filter embodies the effects of the near‐surface on the reflection seismogram, i.e. the cause of the ghosting. Analysis of these filters reveals that the ghosting mechanism in the field test area consists of both a surface‐ and base‐of‐weathering layer reflector.

Technetium 99m sulfide colloid can be prepared rapidly and conveniently by precipitation from pertechnetate eluted from a Tc99m generator, using rhenium as a carrier, through the action of sodium thiosulfate. The use of presterilized reagents yields a sterile product, which is ready for use without further processing. Gelatin is added as a protective colloid. No toxic materials or odoriferous gases are used. The product contains virtually no soluble technetium and no macroaggregates. Preparation time is less than 10 minutes, and can be done in a small area of any radioisotope clinic. Liver scans of high resolution and excellent counting statistics can be obtained with minimal patient radiation dose.

A new technique has been developed which makes it possible to process a seismic record‐section in such a way that all seismic events with dips in a given range are preserved with no alteration over a wide frequency band, while all seismic events with dips outside the specified range are uniformly and severely attenuated. By applying this process to a noisy record‐section, a record‐section may be obtained which has all events within a specified dip range perfectly preserved, and very high‐velocity noise essentially eliminated, a result which is impossible by simple wave‐number filtering or conventional array usage. In structurally complex areas where several steeply dipping events interfere, the technique may be applied to separate the events with different dips. In areas where a normal‐moveout contrast exists between primaries and multiples, the technique may be used for wide‐band multiple attenuation. By application of a “rotating Pie‐Slice” to micro‐spread noise data, seismic noise may be separated on the basis of propagation velocity, and a clearer picture of the seismic noise problem obtained. The “rotating Pie‐Slice” also provides a means of uncovering diffractions and other steeply‐curved events from a record section. The paper discusses the motivation and implementation of the process and its application to both synthetic and actual data.

The extent and duration of Technetium-99m microaggregates of albumin entrapment in acute intimal arterial ulcerations was investigated by computer videotape analysis to determine the optimal period of time for arterial scanning.

Technetium-99m-sulfur suspension produced by heating of thiosulfate was introduced for clinical use by Patton et al. in 1965(1). At present it is employed extensively for demonstration of liver, spleen, and bone marrow. The original preparation, technetium sulfide-sulfur “colloid,” was co-precipitated with rhenium sulfide and stabilized by the addition of gelatin. Larson and Nelp (2) suggested an alternative preparation of this colloid which was stabilized with dextran instead of with gelatin. After an initial period of using gelatin, dextran was employed at this institution. Over a period of several months it was noted that of several hundred patients, three had reactions of the anaphylactoid type manifested by flushing of the skin and hypotension. Investigation of these symptoms and a review of the literature indicated that this was most likely related to the dextran. This has been confirmed by others (3). As a result of these reactions, the preparation of the colloid was modified by eliminating the dex...

For a 3built12 year period during which 125I was used as the routine agent for thyroid imaging, 95 cases of thyroid disease were reviewed in which a tissue diagnosis was obtained. The scan findings were usually distinctive in the cases of diffuse hyperplasia, but were of little help in separating the carcinomas from the nodular goiters or adenomas. This was true even when attention was confined to those cases presenting as nontoxic solitary or multiple nodules with no clinical evidence of malignancy. These results fall within the spectrum of results reported using 131I, and the two agents are probably of approximately equal value when used with commercially available scanners. The 125I scan showed less accuracy in separating multinodular from uninodular glands than that reported using 99mTcO4, but in our series carcinoma was as frequent in glands which contained multiple nodules on pathologic examination as in glands with solitary nodules.

Since the imaging pattern of 67Ga-citrate is well known, it can be the benchmark for evaluation of potential tumor imaging agents. Tritium labeled 5-fluorouracil, Adriamycin, (doxorubicin/Adria) and Methotrexate/Lederle were compared at 2 h to 67Ga-citrate at 2 and 48 h by percent uptake per gram of various tissues and tumor in rats implanted with Walker 256 carcinosarcoma. 67Ga at 48 h showed approximately three times more uptake in the tumor than 5 fluorouracil and eight times more than the others. 5 Fluorouracil showed approximately three times greater uptake than Methotrexate or doxorubicin and 1.4 times the 2 h uptake of 67Ga. 5 Fluorouracil shows potential as an imaging agent especially where the waiting time between injection and imaging must be short.

The incidence of mural thrombosis following three types of experimental intimal injuries at common carotid arteries, was investigated in a series of 12 dogs. Mural thrombosis was obtained at each of the intimal lesions. Moreover, obstructive thrombosis occurred in three instances. Further investigation of thrombi revealed that partial fragmentation occurred when exposed to hydrodynamic forces. These results suggest that mural thrombosis following carotid endarterectomy in humans, is possibly more frequent and significant than it is generally appreciated, thus leading to the consideration of postoperative anticoagulation after endarterectomy, as a means to minimize the incidence of postoperative embolic cerebrovascular accidents.

The 125 I-fibrinogen uptake test (FUT) was evaluated in a series of 42 subjects to determine its ability to detect active venous thrombosis. Included were patients with recent active thrombosis, patients with long-standing quiescent venous obstruction, and normal controls. In serial tests, increasing isotope uptake was found to occur only in patients with active thrombophlebitis ( P Five patients with known calf vein thrombosis had increasing FUT results at the upper thigh. Phlebography or technetium scans confirmed propagation of the thrombosis. These results suggest a screening value for the FUT in suspected propagating thrombosis. If results are confirmed by subsequent phlebography, major therapeutic changes such as intensification of anticoagulation or venous interruptions can hopefully be made before potentially lethal pulmonary embolism occurs.