Natallia Katenka

Local Vote Decision Fusion for Target Detection in Wireless Sensor Networks

This paper examines the problem of target detection by a wireless sensor network. Sensors acquire measurements emitted from the target that are corrupted by noise, and initially make individual decisions about the presence/absence of the target. We propose the local vote decision fusion algorithm, in which sensors first correct their decisions using decisions of neighboring sensors, and then make a collective decision as a network. An explicit formula that approximates the systems decision threshold for a given false alarm rate is derived using limit theorems for random fields, which provides a theoretical performance guarantee for the algorithm. We examine both distance- and nearest-neighbor-based versions of the local vote algorithm for grid and random sensor deployments and show that, in many situations, for a fixed-system false alarm, the local vote correction achieves significantly higher target detection rate than decision fusion based on uncorrected decisions. The algorithm does not depend on the signal model and is shown to be robust to different types of signal decay. We also extend this framework to temporal fusion, where information becomes available over time.

Intrusion as (anti)social communication: characterization and detection

A reasonable definition of intrusion is: entering a community to which one does not belong. This suggests that in a network, intrusion attempts may be detected by looking for communication that does not respect community boundaries. In this paper, we examine the utility of this concept for identifying malicious network sources. In particular, our goal is to explore whether this concept allows a core-network operator using flow data to augment signature-based systems located at network edges. We show that simple measures of communities can be defined for flow data that allow a remarkably effective level of intrusion detection simply by looking for flows that do not respect those communities. We validate our approach using labeled intrusion attempt data collected at a large number of edge networks. Our results suggest that community-based methods can offer an important additional dimension for intrusion detection systems.

Enhancement of radiation effect on cancer cells by gold-pHLIP

Significance Nanometer-sized gold particles are shown to increase the effectiveness of radiation in killing cancer cells. Improved radiation effectiveness allows less radiation to be used, reducing adverse effects to patients. Alternatively, more cancer killing could be possible while using current radiation doses. Here we used pH Low-Insertion Peptide (pHLIP) to tether gold nanoparticles to membranes of cancer cells. This increases their effectiveness because the radiation/particle effect is very localized. We find that pHLIP significantly increases the amount of gold particles in cancer cells, as well as the amount of cancer cell death from radiation. This methodology is promising for clinical research, as previous results show efficient targeting of gold nanoparticles to tumors by pHLIP. Previous research has shown that gold nanoparticles can increase the effectiveness of radiation on cancer cells. Improved radiation effectiveness would allow lower radiation doses given to patients, reducing adverse effects; alternatively, it would provide more cancer killing at current radiation doses. Damage from radiation and gold nanoparticles depends in part on the Auger effect, which is very localized; thus, it is important to place the gold nanoparticles on or in the cancer cells. In this work, we use the pH-sensitive, tumor-targeting agent, pH Low-Insertion Peptide (pHLIP), to tether 1.4-nm gold nanoparticles to cancer cells. We find that the conjugation of pHLIP to gold nanoparticles increases gold uptake in cells compared with gold nanoparticles without pHLIP, with the nanoparticles distributed mostly on the cellular membranes. We further find that gold nanoparticles conjugated to pHLIP produce a statistically significant decrease in cell survival with radiation compared with cells without gold nanoparticles and cells with gold alone. In the context of our previous findings demonstrating efficient pHLIP-mediated delivery of gold nanoparticles to tumors, the obtained results serve as a foundation for further preclinical evaluation of dose enhancement.

Robust Target Localization From Binary Decisions in Wireless Sensor Networks

Wireless sensor networks (WSNs) are becoming important tools in various tasks, including monitoring and tracking of spatially occurring phenomena. These networks offer the capability of densely covering a large area, but at the same time are constrained by the limiting sensing, processing and power capabilities of their sensors. To complete the task at hand, the information collected by the sensor nodes needs to be appropriately fused. In this article we study the problems of estimating the location of a target and estimating its signal intensity. The proposed algorithms are based on the local vote decision fusion (LVDF) mechanism, where sensors first correct their original decisions using decisions of neighboring sensors. These corrected decisions are more accurate and robust and improve detection; however, they are correlated, which makes maximum likelihood estimation intractable. We adopt a pseudolikelihood formulation and examine several variants of localization and signal estimation algorithms based on original and corrected decisions using direct optimization methods, as well as an EM approach. Uncertainty assessments about the parameters of interest are provided using a parametric bootstrap technique. An extensive simulation study of the developed algorithms, along with several benchmarks, establishes the overall superior performance of the LVDF-based algorithms, especially in low signal-to-noise ratio environments. Extensions to tracking moving targets and localizing multiple targets also are considered.

Inference and characterization of multi-attribute networks with application to computational biology

Our work is motivated by and illustrated with application of association networks in computational biology, specifically in the context of gene/protein regulatory networks. Association networks represent systems of interacting elements, where a link between two different elements indicates a sufficient level of similarity between element attributes. While in reality relational ties between elements can be expected to be based on similarity across multiple attributes, the vast majority of work to date on association networks involves ties defined with respect to only a single attribute. We propose an approach for the inference of multi-attribute association networks from measurements on continuous attribute variables, using canonical correlation and a hypothesis-testing strategy. Within this context, we then study the impact of partial information on multi-attribute network inference and characterization, when only a subset of attributes is available. We consider in detail the case of two attributes, wherein we examine through a combination of analytical and numerical techniques the implications of the choice and number of node attributes on the ability to detect network links and, more generally, to estimate higher-level network summary statistics, such as node degree, clustering coefficients, and measures of centrality. Illustration and applications throughout the paper are developed using gene and protein expression measurements on human cancer cell lines from the NCI-60 database.

Comparative Study of Tumor Targeting and Biodistribution of pH (Low) Insertion Peptides (pHLIP(®) Peptides) Conjugated with Different Fluorescent Dyes.

PurposeAcidification of extracellular space promotes tumor development, progression, and invasiveness. pH (low) insertion peptides (pHLIP® peptides) belong to the class of pH-sensitive membrane peptides, which target acidic tumors and deliver imaging and/or therapeutic agents to cancer cells within tumors.ProceduresEx vivo fluorescent imaging of tissue and organs collected at various time points after administration of different pHLIP® variants conjugated with fluorescent dyes of various polarity was performed. Methods of multivariate statistical analyses were employed to establish classification between fluorescently labeled pHLIP® variants in multidimensional space of spectral parameters.ResultsThe fluorescently labeled pHLIP® variants were classified based on their biodistribution profile and ability of targeting of primary tumors. Also, submillimeter-sized metastatic lesions in lungs were identified by ex vivo imaging after intravenous administration of fluorescent pHLIP® peptide.ConclusionsDifferent cargo molecules conjugated with pHLIP® peptides can alter biodistribution and tumor targeting. The obtained knowledge is essential for the design of novel pHLIP®-based diagnostic and therapeutic agents targeting primary tumors and metastatic lesions.

Tracking Multiple Targets Using Binary Decisions From Wireless Sensor Networks

This article introduces a framework for tracking multiple targets over time using binary decisions collected by a wireless sensor network, and applies the methodology to two case studies—an experiment involving tracking people and a dataset adapted from a project tracking zebras in Kenya. The tracking approach is based on a penalized maximum likelihood framework, and allows for sensor failures, targets appearing and disappearing over time, and complex intersecting target trajectories. We show that binary decisions about the presence/absence of a target in a sensors neighborhood, corrected locally by a method known as local vote decision fusion, provide the most robust performance in noisy environments and give good tracking results in applications.

Joint modeling of time to recurrence and cancer stage at recurrence in oncology trials

ABSTRACT This research was motivated by a clinical trial with bladder cancer patients who went through a surgery and were followed up for cancer recurrence. One of the main objectives of the trial was to evaluate the time to cancer recurrence in patients in control and experimental groups. At the time of recurrence, the disease stage was also evaluated. Because the stage of cancer at recurrence significantly impacts future treatment and patient prognosis of survival, analyzing the time to cancer recurrence and the stage at recurrence jointly provides more clinically relevant information than analyzing the time to recurrence alone. In this paper, we propose a stochastic model for the joint distribution of time to recurrence and cancer stage that (1) accounts for the recurrence caused by cancer cells surviving a treatment or a surgery and for the recurrence caused by spontaneous carcinogenesis, and (2) incorporates parameters that have biological meaning. To estimate the parameters, we use the maximum-likelihood method combined with the EM algorithm. To demonstrate the performance of our modeling, we evaluate the data from a clinical trial in patients with bladder cancer. We also use simulations to assess the sensitivity of the method.

Epidemiological Study of Browser-Based Malware for University Network with Partially Observed Flow Data

The presence of personal financial data, intellectual property, and classified documents on University computer systems makes them particularly attractive to hackers, but not well prepared for their attacks. The University of Rhode Island (URI) is one of the few institutions collecting network traffic data (NetFlow) for inference and analysis of normal and potentially malicious activity. This research focuses on web-based traffic with client-server architecture and adopts simple probability-based transmission models to explore the vulnerability of the URI web-network to anticipated threats. The fact that the URI firewall captures only traffic data in- and out- of URI necessitates the modeling of internal un-observed traffic. Relying on a set of intuitive assumptions, we simulate the spread of infection on the dynamic bipartite graph inferred from observed external and modeled unobserved internal web-browsing traffic and evaluate the susceptibility of URI nodes to threats initiated by random clients and clients from specific countries. Overall, the results suggest higher rates of infection for client nodes compared to servers with maximum rates achieved when infection is initiated randomly. Remarkably, very similar rates are observed when infection is initiated from 100 different clients from each of selected countries (e.g., China, Germany, UK) or from one most active node from Denmark. Interestingly, the daily analysis over a three-month period reveals that the simulated infection rates that are not consistent with the intensity of the flow traffic may indicate the presence of compromised node activity and possible intrusion.

Assortative Mixture of English Parts of Speech

Network data analysis is an emerging area of study that applies quantitative analysis to complex data from a variety of application fields. Methods used in network data analysis enable visualization of relational data in the form of graphs and also yield descriptive characteristics and predictive graph models. This paper presents an application of network data analysis to the authorship attribution problem. Specifically, we show how a representation of text as a word graph produces the well documented feature sets used in authorship attribution tasks such as the word frequency model and the part-of-speech (POS) bigram model. Analysis of these models along with word graph characteristics provides insights into the English language. Particularly, analysis of the nominal assortative mixture of parts of speech, a statistic that measures the tendency of words of the same POS in the word network to be connected by an edge, reveals regular structural properties of English grammar.