MS defense: Detection of Unsafe Action in Laparoscopic Cholecystectomy Video

MS defense

Detection of Unsafe Action in Laparoscopic Cholecystectomy Video

Ashwini Lahane

10:00am Thursday, 28 April 2011, ITE 346

Wellness and healthcare are central to the lives of all people. Information technology has already contributed in significant ways towards enhancement of healthcare delivery and to improving the quality of life. And it will continue to do so with the development of “smarter” technologies and environments. Recent years have seen context awareness as one of the most important aspects in the emerging pervasive computing paradigm. We focus our work on situation awareness; a more holistic variant of context awareness where situations are regarded as logically aggregated contexts. We demonstrate an application of situation aware computing in healthcare. We primarily focus on laparoscopic cholecystectomy, a complex yet commonly performed surgical procedure. The outcome of the surgery is influenced greatly by the training, skill, and judgment of the surgeon. Many surgical simulators have been developed to meet the training and practice needs of the surgeons. However few systems provide feedback during the actual surgery. We present a method to detect a situation, that shows possibility of injury to an artery by analyzing the laparoscopic cholecystectomy surgical video. The system can be used to provide feedback to the operating surgeon in case of a possible risk. We have also built a prototype to demonstrate the use of our system in telemedicine, in the form of a web service.

Thesis Committee:

  • Dr. Yelena Yesha (chair)
  • Dr. Anupam Joshi
  • Dr. Milton Halem
  • Dr. Michael Grasso


Complex-valued Adaptive Signal Processing: Applications in Medical Imaging

EE Graduate Seminar

Complex-valued Adaptive Signal Processing:
Applications in Medical Imaging

Prof. Tulay Adali
Machine Learning for Signal Processing Laboratory

1-2pm Friday, 29 April 2011, ITE 237

Complex-valued signals arise frequently in applications as diverse as communications, radar, geophysics, optics, and biomedicine, as most practical modulation formats are of complex type and applications such as radar and magnetic resonance imaging lead to data that are inherently complex valued. The complex domain, however, presents unique challenges for signal processing, in particular for adaptive nonlinear processing, and as a result, until recently, most algorithms derived for the complex domain have taken engineering shortcuts limiting their usefulness. The most common one among those has been assuming the circularity of the signal, thus ignoring the information conveyed by the phase. Similarly when taking gradients in the complex domain, a "split" approach that performs optimization separately with respect to the real and imaginary variables has been the dominant practice.

There have been important advances in the area within the last decade that clearly demonstrate that noncircularity is an intrinsic characteristic of many signals of practical interest, and when taken into account, the methods developed for their processing may provide significant performance gains. Similarly, it has been shown that using Wirtinger calculus, all calculations can be carried out in a manner similar to real-valued calculus while keeping all the computations in the complex domain.

In this talk, after a brief introduction to optimization using Wirtinger calculus and statistics in the complex domain, and then I will give examples from some of the recent work conducted at the MLSP-Lab.

MS defense: Boosting Base Station Anonymity in Wireless Sensor Networks

MS Defense

Exploiting Architectural Techniques for Boosting
Base Station Anonymity in Wireless Sensor Networks

Zhong Ren

2:00pm Thursday, 28 April 2011, ITE 346

Wireless Sensor Networks (WSNs) can be deployed to serve mission-critical applications in hostile environments such as battlefield and territorial borders. In these setups, the WSN may be subject to attacks in order to disrupt the network operation. The most effective way for an adversary to do so is by targeting the Base-Station (BS), where the sensor data are collected in the field. By identifying and locating the BS, the adversary can launch attacks to damage or disrupt the operation of the BS. Therefore, maintaining the BS anonymity is of utmost importance in WSNs.

In this thesis we propose three novel approaches to boost the anonymity of the BS nodes to protect them from potential threats. We first explore the deployment of more BS nodes. We compare the BS anonymity of one versus multiple stationary BS under different network topologies. Our results show that having more base-stations can boost both the average and max anonymity of BS nodes. We further provide guidelines on a cost versus anonymity trade-off to determine the most suitable BS count for a network. Second we exploit the mobility of base-stations and explore the effect of relocating some of the existing BS nodes to the lowest anonymity regions. Our results show that having one mobile BS can dramatically boost the anonymity of the network and moving multiple BS does not provide much value. Finally, we propose to pursue dynamic sensor to cluster re-association to confuse the adversary. This can be employed when base-stations cannot safely move.

Committee members:

  • Mohamed Younis (Chair)
  • Yun Peng
  • Charles Nicholas

CSEE students presenting at the 2011 Graduate Research Conference

Congratulations to the record number of CSEE graduate students who will make oral or poster presentations at the 2011 UMBC Graduate Research Conference on Friday April 29th. The presentations will take place in two sessions, one 9:00 to 10:50 and another 11:00 to 12:50. Poster sessions will be on the seventh floor of the library and oral presentations in UC 312. See the GRC site for a program booklet with abstracts and the session schedules.

Oral presentations:

  • Yousef Ebrahimi, Increasing Transmission Power for Higher Base-station Anonymity in Wireless Sensor Network
  • Zamon Granger, Stabilizing Air Pressure in a Hermetic Environment with Embedded Technology
  • Congchong Liu, Optimizing MapReduce Programming Model through Adaptive Load Balance
  • Kavita Krishnaswamy, Path planning a robotic arm efficiently
  • Varish Mulwad, Generating knowledge from tables
  • Darshana Dalvi, Genome-based Clinical Decision Support System
  • Ashwini Lahane, Detection of unsafe actions in laparoscopic cholecystectomy surgical videos
  • Karuna Pande Joshi, Lifecycle of virtualized services on the cloud
  • Jon Ward, ICF: A Physical Layer Metric for Measuring Anonymity in Wireless Sensor Networks
  • Theoplis E. Stewart Sr., Smart Grid: What are some underlying implications to overlaying digital technology onto the electric power grids?

Poster presentations:

  • Sai Ma, Tom Eichele and Nicolle M. Correa, Hierarchical and graphical analysis of fMRI network connectivity in healthy and schizophrenic groups.
  • Shiming Yang , Near Real"time Data Assimilation for the HYSPLIT Aerosol Dispersion Model.
  • Yu Wang , A Framework for GPU 3D Model Reconstruction using Structure-from-Motion
  • Akshaya Iyengar, Estimating Temporal Boundaries of Events using Social Media Data
  • Brice Cannon, Paveen Apiratikul and John Hryniewicz, Characterization of semiconductor optical waveguides
  • Brian White Jr., The Mobile Memory
  • Xing Chen, Standoff photoacoustic chemical detection using quantum cascade lasers
  • Yi Xin, On Gas Detection and Concentration Estimation via Mid-IR-based Gas Detection System Analysis Model
  • Robert J. Weiblen, Calculation of the expected bandwidth for a mid-infrared supercontinuum source based on As2S3 chalcogenide photonic crystal fibers
  • Pramod Jagtap, Privacy Preservation in Context Aware Geo-social Networking Applications

MS defense: Privacy Preservation in Context-Aware Systems

Mobile devices will provide better services if then can model, recognize and adapt to their users' context.

MS Defense

Privacy Preservation in Context-Aware Systems

Pramod Jagtap

1:00pm Wednesday, 27 April 2011, ITE 210

Recent years have seen a confluence of two major trends – the increase of mobile devices such as smart phones as the primary access point to networked information and the rise of social media platforms that connect people. Their convergence supports the emergence of a new class of context-aware geosocial networking applications. While existing systems focus mostly on location, our work centers on models for representing and reasoning about a more inclusive and higher-level notion of context, including the user’s location and surroundings, the presence of other people and devices, feeds from social networking systems they use, and the inferred activities in which they are engaged. A key element of our work is the use of collaborative information sharing where devices share and integrate knowledge about their context. This introduces the need for privacy and security mechanisms. We present a framework to provide users with appropriate levels of privacy to protect the personal information their mobile devices are collecting including the inferences that can be drawn from the information. We use Semantic Web technologies to specify high-level, declarative policies that describe user’s information sharing preferences. We have built a prototype system that aggregates information from a variety of sensors on the phone, online sources, and sources internal to the campus intranet, and infers the dynamic user context. We show how our policy framework can be effectively used to devise better privacy control mechanisms to control information flow between users in such dynamic mobile systems.


Thesis Committee:

  • Dr. Anupam Joshi (chair)
  • Dr. Tim Finin
  • Dr. Yelena Yesha
  • Dr. Laura Zavala


MS defense: A Framework for GPU 3D Model Reconstruction Using Structure-from-Motion

MS Defense

A Framework for GPU 3D Model Reconstruction
Using Structure-from-Motion

Yu Wang

10:00am Thursday, 28 April 2011, ITE 352

A framework for three-dimensional (3D) model reconstruction is described in this thesis. The primary application is scanning forest canopies and assisting scientific applications such as fire hazard evaluation and vegetation biomass estimation, using photos taken from a radio-controlled helicopter, although the methods apply to any series of photographs taken along a path. The approach is based on the fact that the photos are taken in a continuous path, thus taking advantage of the adjacency between images. The major contributions of this project are 1) introduce a linear time complexity algorithm that reduces number of image pairs to match for datasets obtained in a continuous path. 2) present an optimized high performance framework for GPU 3D reconstruction using structure-from-motion.

Thesis Committee:

  • Dr. Marc Olano, Chair
  • Dr. Penny Rheingans
  • Dr. Erle Ellis

MS defense: eXtensible Dynamic Form (XDF) for Supplier Discovery

MS Thesis Defense

eXtensible Dynamic Form (XDF) for Supplier Discovery

Yan Kang

1:00PM Tuesday, 26 April 2011, ITE 346

Discovery of suppliers (supplier discovery) is essential for building a flexible network of suppliers in a supply chain. The first step for supplier discovery is to collect manufacturing capabilities of suppliers and requirements of customers. In traditional e-marketplaces, online form interfaces are typically used to collect the requirements and capabilities. However, those forms are mostly lack of flexibility to capture a variety of requirements and capabilities in a structured way. In this thesis, we propose new innovative form architecture called eXtensible Dynamic Form (XDF) to facilitate data collection process of supplier discovery.

This architecture provides several key innovations including: 1) architecture for users (suppliers or customers) to create new structure of form for their own contents; 2) an intelligent search engine facilitating users to reuse the existing form components 3) hierarchical representation of the requirements and capabilities as XML instances. Experimental results demonstrate that the proposed architecture is valuable for facilitating the supplier discovery process.

Thesis Committee:

  • Dr. Charles Nicholas
  • Dr. Yun Peng (Chair)
  • Dr. Yelena Yesha

MS defense: Recovering from Soft Node Failures in Wireless Sensor Networks using Neural Networks

MS Thesis Defense

Recovering from Soft Node Failures in
Wireless Sensor Networks using Neural Networks

Shivvasangari Subramani

9:00am Tuesday, 26 April 2011, ITE 346

In the past few years, wireless sensor networks (WSNs) have become important in different applications because of their robustness in hostile environments. WSNs need to perform in a timely manner in the face of interference, attacks, accidents, and failures. Being a battery operated system, there is a trade-off between performance and energy utilization. In this thesis we focus on WSN accuracy and consider ways to improve the performance of WSNs when sensors become damaged, resulting in poor input signal quality. When all other components of the sensor like the processor, memory, and battery are working, our proposed solution is to learn to undo the damage in a node by training on neighbors sensor values.

Thesis Committee:

  • Dr. Anupam Joshi
  • Dr. Tim Oates (chair)
  • Dr. Mohamed Younis

MS defense: Graph-Theoretic Approach to Collusion Detection in Multi-Agent Systems

MS Thesis Defense

A Graph-Theoretic Approach to
Collusion Detection in Multi-Agent Systems

Peter Hamilton

9:00am Thursday, 28 April 28 2011, ITE 325B

The study of trust and cooperation is a major component of multi-agent systems research. Such work often focuses on how best to estimate the reliability of a specific agent, or how to create strategies and protocols that engender the most cooperation from the most agents. However, when cooperation is not a desired aspect of a multi-agent system, these actions define collusive behavior, which can have a significant impact on the dynamics of the system.

This thesis defines a generic, graph-theoretic approach to collusion detection known as CODING. This approach detects group-based collusion, targeting two basic collusion mechanisms that rely on large numbers of colluding agents for success. CODING analyzes and classifies agent interactions from the system and constructs a series of interaction graphs from this data. These graphs are processed for structures that correspond to collusion mechanisms; the agents composing these structures are reported as colluders. CODING is applied to a game theory domain, in which it must detect agents adhering to group strategies in round-robin tournaments composed of single-player strategies.

Thesis Committee:

  • Dr. Marie desJardins (chair)
  • Dr. Tim Finin
  • Dr. Tim Oates


MS defense: Feature Extraction using a Hierarchical Growing Neural Gas

M.S. Thesis Defense Announcement

Feature Extraction using a Hierarchical Growing Neural Gas

Roger Guseman

12:00pm 25 April 2011, ITE 210

Unsupervised, data-driven, automatic feature extraction from image data is an interesting and difficult problem. High dimensional data, such as images, often contain less information than they do data. For an agent to better reason about this data, finding the "interesting" features in the data is helpful. A current technique, known as the Growing Neural Gas (GNG), is a neural network approach to feature extraction. There are, however, adaptations that can be made to the Growing Neural Gas in order to increase its performance.

Contributions of this work include development of a new neural network algorithm extending the Growing Neural Gas framework, known as the Hierarchical Growing Neural Gas (HGNG), identification of how the parameters of the HGNG affect feature extraction performance, and theoretical and empirical comparisons of performance between the normal GNG and the HGNG neural networks.

Thesis Committee:

  • Dr. Tim Oates (chair)
  • Dr. Tim Finin
  • Dr. Marie desJardins


1 32 33 34 35 36 37