Search

Purpose: Prostate cancer aggressiveness and appropriate therapy are routinely determined following biopsy sampling. Current clinical and pathologic parameters are insufficient for accurate risk prediction leading primarily to overtreatment and also missed opportunities for curative therapy. Experimental Design: An 8-biomarker proteomic assay for intact tissue biopsies predictive of prostate pathology was defined in a study of 381 patient biopsies with matched prostatectomy specimens. A second blinded study of 276 cases validated this assay's ability to distinguish “favorable” versus “nonfavorable” pathology independently and relative to current risk classification systems National Comprehensive Cancer Network (NCCN and D'Amico). Results: A favorable biomarker risk score of ≤0.33, and a nonfavorable risk score of >0.80 (possible range between 0 and 1) were defined on “false-negative” and “false-positive” rates of 10% and 5%, respectively. At a risk score ≤0.33, predictive values for favorable pathology in very low-risk and low-risk NCCN and low-risk D'Amico groups were 95%, 81.5%, and 87.2%, respectively, higher than for these current risk classification groups themselves (80.3%, 63.8%, and 70.6%, respectively). The predictive value for nonfavorable pathology was 76.9% at biomarker risk scores >0.8 across all risk groups. Increased biomarker risk scores correlated with decreased frequency of favorable cases across all risk groups. The validation study met its two coprimary endpoints, separating favorable from nonfavorable pathology (AUC, 0.68; P < 0.0001; OR, 20.9) and GS-6 versus non–GS-6 pathology (AUC, 0.65; P < 0.0001; OR, 12.95). Conclusions: The 8-biomarker assay provided individualized, independent prognostic information relative to current risk stratification systems, and may improve the precision of clinical decision making following prostate biopsy. Clin Cancer Res; 21(11); 2591–600. ©2015 AACR.

Background: Key challenges of biopsy-based determination of prostate cancer aggressiveness include tumour heterogeneity, biopsy-sampling error, and variations in biopsy interpretation. The resulting uncertainty in risk assessment leads to significant overtreatment, with associated costs and morbidity. We developed a performance-based strategy to identify protein biomarkers predictive of prostate cancer aggressiveness and lethality regardless of biopsy-sampling variation. Methods: Prostatectomy samples from a large patient cohort with long follow-up were blindly assessed by expert pathologists who identified the tissue regions with the highest and lowest Gleason grade from each patient. To simulate biopsy-sampling error, a core from a high- and a low-Gleason area from each patient sample was used to generate a ‘high' and a ‘low' tumour microarray, respectively. Results: Using a quantitative proteomics approach, we identified from 160 candidates 12 biomarkers that predicted prostate cancer aggressiveness (surgical Gleason and TNM stage) and lethal outcome robustly in both high- and low-Gleason areas. Conversely, a previously reported lethal outcome-predictive marker signature for prostatectomy tissue was unable to perform under circumstances of maximal sampling error. Conclusions: Our results have important implications for cancer biomarker discovery in general and development of a sampling error-resistant clinical biopsy test for prediction of prostate cancer aggressiveness.

In this paper, we develop a new dimension reduction technique, using the projection pursuit approach, to identify underlying physical mechanisms of the flow. By first applying the standard dimension reduction technique–proper orthogonal decomposition (POD), a low-dimensional subspace is defined. POD models of pressure fields have, in the past, been challenged with questions of interpretation in terms of flow mechanisms. To address this issue, the projection criterion is applied to decompose independent physical mechanisms in the non-Gaussian pressure field. This procedure leads to a non-orthogonal subspace decomposition that provides a suitable subspace for identification of physical mechanisms in intermittent flows. This approach provides a new tool to further our understanding of the fundamental nature of intermittent and independent phenomena in fluid flows. Finally, this technique is tested with experimental data collected at Texas Tech University's Wind Engineering Research Field Laboratory.

In this paper, we introduce two statistical techniques to detect coherent structures in long time-series: an intermittency rate estimator and an event-counting technique. The intermittency rate estimator calculates the fraction of time occupied by coherent structures, in contrast to incoherent noise. The event-counting approach estimates the number of discrete structures that occurred. In both cases, the wavelet transform and extensions of the coherent structure detector (CSD) are used to develop rigorous statistical approaches. The techniques are demonstrated on a 7 h wind time-series collected at the Wind Engineering Research Field Laboratory at Texas Tech.

In this paper, we describe a method for nonlinear fuzzy regression using neural network models. In earlier work, strong assumptions were made on the form of the fuzzy number parameters: symmetric triangular, asymmetric triangular, quadratic, trapezoidal, and so on. Our goal here is to substantially generalize both linear and nonlinear fuzzy regression using models with general fuzzy number inputs, weights, biases, and outputs. This is accomplished through a special training technique for fuzzy number neural networks. The technique is demonstrated with data from an industrial quality control problem.

This paper presents a practical algorithm for training neural networks with fuzzy number weights, inputs, and outputs. Typically, fuzzy number neural networks are difficult to train because of the many α-cut constraints implied by the fuzzy weights. A transformation is used to eliminate these constraints and allow use of standard unconstrained optimization methods. The algorithm is demonstrated on a three-layer network.

Fuzzy fault trees provide a powerful and computationally efficient technique for developing fuzzy probabilities based on independent inputs. The probability of any event that can be described in terms of a sequence of independent unions, intersections, and complements may be calculated by a fuzzy fault tree. Unfortunately, fuzzy fault trees do not provide a complete theory: many events of substantial practical interest cannot be described only by independent operations. Thus, the standard fuzzy extension (based on fuzzy fault trees) is not complete since not all events are assigned a fuzzy probability. Other complete extensions have been proposed, but these extensions are not consistent with the calculations from fuzzy fault trees. We propose a new extension of crisp probability theory. Our model is based on n independent inputs, each with a fuzzy probability. The elements of our sample space describe exactly which of the n input events did and did not occur. Our extension is complete since a fuzzy probability is assigned to every subset of the sample space. Our extension is also consistent with all calculations that can be arranged as a fault tree. Our approach allows the reliability analyst to develop complete and consistent fuzzy reliability models from existing crisp reliability models. This allows a comprehensive analysis of the system. Computational algorithms are provided both to extend existing models and develop new models. The technique is demonstrated on a reliability model of a three-stage industrial process.

The spectrum is commonly used to characterize the energy content and the longitudinal scale of turbulence in the wind field. However, this incoherent signal model is not always sufficient to identify intermittent coherent structures in the wind that may be responsible for wind damage. These periods of gusts may be viewed as coherent structures occurring in an otherwise incoherent signal. In this paper, we describe a method for using wavelet transforms to detect coherent structures in the wind field. To detect coherent gusts, we use the Haar wavelet since the Haar analyzing function detects the magnitude and length scale of sustained changes in signal level. A statistical technique is developed to separate coherent structures in the wavelet transform from fluctuations due to incoherent noise.

This paper studies a 1989 measles epidemic on the Texas Tech campus. The infection is modeled in a population with a social contact structure. In the model, students are assigned to classes, dormitory complexes, dormitory rooms, and social groups in the dining halls. A single infected individual is introduced, and the resulting infection path is described as a random graph.

Optimal control of perturbed Hamiltonian systems in $\Re^2$ is studied. Systems are considered with a control term scaling with the size of a small perturbing noise. The dynamics are shown to converge in a certain sense to a diffusion on a graph. Using the approach developed in [M. I. Freidlin and A. D. Wentzell, Mem. Amer. Math. Soc., 109 (1994), pp. 1--82] and [M. I. Freidlin and A. D. Wentzell, Ann. Probab., 21 (1993), pp. 2215--2245] for random perturbations of Hamiltonian systems, a convergence theorem is discussed. An optimal control theorem is then developed to maximize the expected exit time from a domain. This control is asymptotically robust for small noise. Several examples are provided.

A reaction - diffusion equation is studied with linear dynamics in the medium and a nonlinear boundary condition. The diffusion occurs in a domain that is perforated by periodically placed holes. On the boundary of these holes, the nonlinear boundary condition is applied. Wavefront propagation is established by use of a Feynman - Kac representation of the solution and development of an appropriate large deviations principle. The location of the wavefront is described in terms of the Legendre transform of the solution to an eigenvalue problem. The proof uses periodicity in an essential way by application of Doeblin's condition to the associated random process on the torus.

This article investigates the limiting behavior of a diffusion in a half space with a complicated boundary condition. The boundary condition implements a reflection condition everywhere except a number of small sets or “holes” that meet Dirichlet or mixed boundary conditions. Probabilistic methods associated with the Feynman-Kac formula are used to find the limiting behavior of the diffusion equations as the number of holes gets large and the size of each hole is reduced. With particular scaling homogenization occurs, and we see that the complicated boundary condition is replaced by a simple mixed boundary condition depending on the capacitance and distribution of the holes.

88 Background: Standard clinical and pathological parameters derived from diagnostic biopsy are insufficient to accurately assess final prostate tumor pathology of patients with biopsy Gleason grades 3+3 or 3+4. We developed a novel assay (ProMark) that performs quantitative measurements of 8 protein markers from prostate biopsy FFPE sections. In the validation study, assay risk scores were strongly predictive of final prostate tumor pathology with a C-stat of 0.69 (95%CI = 0.63 – 0.76) (p

The multi-input multi-output (MIMO) technique has been portrayed as the next revolution for wireless technologies. The MIMO technique is an effective mechanism to increase capacity and reliability of wireless communications by taking advantage of its spatial multiplexing gain or diversity gain. However, conventional MIMO is expensive: the form factor of the radio is large, multiple RF front ends are required, multiple antennas can be difficult to place on a mobile platform, and the parallel propagation channels may be correlated. The paper investigates an alternative way of implementing MIMO: cooperative MIMO, where multiple nodes coordinate their transmissions so that cooperative parallel transmissions can be established between multiple source nodes and a destination node, and space/time coding is applied on the transmissions. We not only investigate the capacity gain from using cooperative MIMO, we also use cooperative MIMO to perform fast handoff for wireless networks. The performance of the proposed fast handoff approach is evaluated using modeling and simulation (M&S).

The threat of biological warfare and the emergence of new infectious agents spreading at a global scale have highlighted the need for major enhancements to the public health infrastructure. Early detection of epidemics of infectious diseases requires both real-time data and real-time interpretation of data. Despite moderate advancements in data acquisition, the state of the practice for real-time analysis of data remains inadequate. We present a nonlinear mathematical framework for modeling the transient dynamics of influenza, applied to historical data sets of patients with influenza-like illness. We estimate the vital time-varying epidemiological parameters of infections from historical data, representing normal epidemiological trends. We then introduce simulated outbreaks of different shapes and magnitudes into the historical data, and estimate the parameters representing the infection rates of anomalous deviations from normal trends. Finally, a dynamic threshold-based detection algorithm is devised to assess the timeliness and sensitivity of detecting the irregularities in the data, under a fixed low false-positive rate. We find that the detection algorithm can identify such designated abnormalities in the data with high sensitivity with specificity held at 97%, but more importantly, early during an outbreak. The proposed methodology can be applied to a broad range of influenza-like infectious diseases, whether naturally occurring or a result of bioterrorism, and thus can be an integral component of a real-time surveillance system.

Wireless sensor networks are an enabling technology for many future surveillance-oriented applications. Before a practical wireless sensor network is realized, however, significant challenges must be overcome. Chief among the obstacles to netted sensors is providing low power, robust communications between sensor nodes. Multiple input, multiple output (MIMO) communication promises performance enhancements over conventional single input, single output (SISO) technology for the same radiated power. If leveraged in a sensor network, MIMO may be able to provide significant network performance improvements in power consumption, latency, and network robustness. The application of MIMO to wireless sensor networks was explored in J. Burdin et al. (2005) with regard to probability of cohesion and largest cluster sizes. This paper investigates the benefit of MIMO implementations in multihop wireless sensor networks in terms of the mean path length. We find that MIMO provides improvement to wireless sensor networks, particularly those that have low to midrange node densities

Multiple-access interference (MAI) limits the performance of CDMA systems using the conventional receiver, but the optimal receiver is too complex for practical implementation. This paper demonstrates a numerically efficient technique for reducing MAI in CDMA systems. The technique, a nonlinear approach inspired by the Kalman filter, uses decoupled filters to estimate symbols for each user while accomplishing interference cancellation in the innovation term. The filter technique developed here provides a computationally practical approach to multiuser detection for asynchronous long code systems, which significantly outperforms the conventional receiver. Kalman filter approaches to multiuser detection have been applied before (Lim and Ma 2000, Lim et. al. 1998) to binary phase shift keying through implementation of a standard Kalman filter approach. Unfortunately, this leads to high computational complexity. Here, we instead consider a formulation in which the spreading codes are viewed as random and their properties are used to decouple the Kalman filter for each user. Interference cancellation is then implemented through the innovations. A binary minimum mean squared error estimate is made, which results in a nonlinear estimator. The new nonlinear technique is shown to significantly outperform an optimized partial parallel interference cancellation approach, which is the current state of the art technique for asynchronous long code systems. System capacity is then significantly increased.

In this paper, we describe a Kalman filter based technique for multiuser detection of asynchronous CDMA systems. Similar to previously proposed techniques, we use a decoupled form of the Kalman filter, which processes each user independently. We then make the further simplifying assumption that the channel statistics are constant over a single symbol allowing us to reduce the Kalman gain to a constant dependent on the current signal to interference plus noise ratio. We study the performance of the constant gain filter and show that by appropriate choice of the gain you can optimize for either minimum bit error rate or maximum interference reduction.

Long code asynchronous CDMA provides a significant challenge, especially when the system is operating at high capacity. We describe a Kalman filter technique that estimates the channel coefficients while demodulating the multiuser signal. A state-space signal model is derived when multiple sources each provide a pilot and numerous user channels, as described by IS95. A numerically efficient algorithm is proposed which estimates each source disjointly but provides interference cancellation through the innovations. A Rayleigh fading model based on Jakes (1974) model is used to describe a time-varying channel. Performance is then demonstrated at high capacity and compared to the conventional detector.

Because of ease of implementation and robustness to multipath delay spread, orthogonal frequency division multiplexing is an increasing popular technology for ad hoc networks. Unfortunately, OFDM performs poorly in the presence of cochannel interference, which degrades network capacity. In this paper, we develop beamforming techniques which minimize cochannel interference to allow spatial division multiple access. The standard approaches in the literature develop a separate beamformer for each narrowband carrier, with perhaps some grouping based on coherence across frequencies. This approach is both computationally complex and, as shown in this paper, difficult in an outdoor environment with expected Doppler spread. Here, we develop a time domain beamformer with low complexity and more robustness to a changing channel. Performance is demonstrated with multipath angular and Doppler spreads typical of mobile uplinks.

This paper continues the authors' earlier work on the development of a statistical technique for identifying coherent structures based on wavelet transforms. In our earlier work, we proved that our test has the correct size. In this paper, we examine both probabilities of detection and false alarm in the special case of signal plus noise. This analysis demonstrates that the coherent structure detector performs well when compared with standard approaches.

Few training techniques are available for neural networks with fuzzy number weights, inputs, and outputs. Typically, fuzzy number neural networks are difficult to train because of the many /spl alpha/-cut constraints implied by the fuzzy weights. In this paper, we introduce a weight representation that simplifies the constraint equations. A constrained form of backpropagation is then developed for fuzzy number neural networks. Standard backpropagation may be viewed as a constrained optimization of the linearization of the weight function. Our weight representation allows use of the additional /spl alpha/-cut constraints during a weight update.

Fuzzy fault trees provide a powerful and computationally efficient technique for developing fuzzy probabilities based on independent inputs. The probability of any event that can be described in terms of a sequence of independent unions, intersections, and complements may be calculated by a fuzzy fault tree. Unfortunately, fuzzy fault trees do not provide a complete theory: many events of substantial practical interest cannot be described only by independent operations. We introduce an extension of crisp probability theory. Our model is based on n independent inputs, each with a fuzzy probability. The elements of our sample space describe exactly which of the n input events did and did not occur. Our extension is complete, since a fuzzy probability is assigned to every subset of the sample space. Our extension is also consistent with all calculations that can be arranged as a fault tree.

Optimal residence time control is studied as a measure of stabilization. Systems are considered with a control term scaling with the size of a small perturbing noise. The dynamics are shown to converge in a certain sense to a diffusion on a graph. Using the approach developed in Freidlin and Wentzell (1994, 1993) for random perturbations of Hamiltonian systems, a convergence theorem is discussed. An optimal control theorem is then developed to maximize the expected exit time from a domain.

A new technique is discussed for the training of fuzzy neural networks with general fuzzy number inputs, weights, and outputs. Fuzzy number neural networks are difficult to train because of the many alpha-cut constraints implied by the fuzzy weights. In this paper, an unconstrained representation is used for the fuzzy weights, allowing application of a standard backpropagation approach. The technique is demonstrated on a moderately large problem.

In this paper, we seek evidence of sustained changes in wind speed that are not explained by incoherent variability. Wavelets provide an excellent tool for the detection of coherent structures in the atmosphere. Unfortunately, in signals with a highly stochastic nature, the wavelet transform often replaces a complicated one-dimensional signal representation with an even more complex two-dimensional representation. To account for this, we introduce a statistical technique through the development of a randomized reference distribution that allows us to separate coherent structures from the variability of the transform due to incoherent noise alone.

In a fuzzy number neural network, the inputs, weights, and outputs are general fuzzy numbers. The requirement that F~/sup /spl alpha/(1)//spl sub/F~/sup /spl alpha/(2/) whenever /spl alpha/(1)>/spl alpha/(2) imposes an enormous number of constraints on the weight parameterizations during training. This problem can be solved through a careful choice of weight representation. This new representation is unconstrained, so that standard neural network training techniques may be applied. Unfortunately, fuzzy number neural networks still have many parameters to pick during training, since each weight is represented by a vector. Thus moderate to large fuzzy number neural networks suffer from the usual maladies of very large neural networks. In this paper, we discuss a method for effectively reducing the dimensionality of networks during training. Each fuzzy number weight is represented by the endpoints of its /spl alpha/-cuts for some discretization 0/spl les//spl alpha//sub 1/

5090 Background: Current clinical and pathological parameters are insufficient for accurate prediction of progression risk for patients with biopsy Gleason grades 3+3 or 3+4. Reasons for this inclu...

Current commercial light water reactor cores operate without recylce of fuel, on a once-through fuel cycle. To help conserve the limited nuclear fuel resources, there is interest in increasing the energy yield and, hence, fuel utilization from once-through fuel irradiation. This report evaluates the potential increase in fuel utilization of light water reactor cores operating on a once-through cycle assuming 0.2% enrichment plant tails assay. This evaluation is based on a large number of survey calculations using techniques which were verified by more detailed calculations of several core concepts. It is concluded that the maximum fuel utilization which could be achieved by practical once-through pressurized light water reactor cores with either uranium or thorium is about 17 MWYth/ST U/sub 3/O/sub 8/ (Megawatt Years Thermal per Short Ton of U/sub 3/O/sub 8/). This is about 50% higher than that of current commercial light water reactor cores. Achievement of this increased fuel utilization would require average fuel burnup beyond 50,000 MWD/MT and incorporation of the following design features to reduce parasitic losses of neutrons: reflector blankets to utilize neutrons that would otherwise leak out of the core; fuel management practices in which a smaller fraction of the core is replaced at eachmore » refueling; and neutron economic reactivity control, such as movable fuel control rather than soluble boron control. For a hypothetical situation in which all neutron leakage and parasitic losses are eliminated and fuel depletion is not limited by design considerations, a maximum fuel utilization of about 20 MWYth/ST U/sub 3/O/sub 8/ is calculated for either uranium or thorium. It is concluded that fuel utilization for comparable reactor designs is better with uranium fuel than with thorium fuel for average fuel depletions of 30,000 to 35,000 MWD/MT which are characteristic of present light water reactor cores.« less

Selected prebreeder core concepts are described which could be backfit into a reference light water reactor similar to current commercial reactors, and produce uranium-233 for use in water-cooled breeder reactors. The prebreeder concepts were selected on the basis of minimizing fuel system development and reactor changes required to permit a backfit. The fuel assemblies for the prebreeder core concepts discussed would occupy the same space envelope as those in the reference core but contain a 19 by 19 array of fuel rods instead of the reference 17 by 17 array. An instrument well and 28 guide tubes for control rods have been allocated to each prebreeder fuel assembly in a pattern similar to that for the reference fuel assemblies. Backfit of these prebreeder concepts into the reference reactor would require changes only to the upper core support structure while providing flexibility for alternatives in the type of fuel used.