Duckwon Chung

A New Approach for Mammogram Image Classification Using Fractal Properties

Abstract Accurate classification of images is essential for the analysis of mammograms in computer aided diagnosis of breast cancer. We propose a new approach to classify mammogram images based on fractal features. Given a mammogram image, we first eliminate all the artifacts and extract the salient features such as Fractal Dimension (FD) and Fractal Signature (FS). These features provide good descriptive values of the region. Second, a trainable multilayer feed forward neural network has been designed for the classification purposes and we compared the classification test results with K-Means. The result reveals that the proposed approach can classify with a good performance rate of 98%.

ANALYSIS OF ELECTROCARDIOGRAM SIGNALS OF ARRHYTHMIA AND ISCHEMIA USING FRACTAL AND STATISTICAL FEATURES

In this study, we present a three-stage method for detecting abnormalities and classifying electrocardiogram (ECG) beats using a k-nearest neighbor (k-NN) classifier and Gaussian mixture model (GMM). In the first stage, a signal filtering method is used to remove the ECG beat baseline wander. In the second stage, features are extracted based on Higuchis fractal dimension (HFD) and statistical features. In the third stage, k-NN and GMM are used as classifiers to classify arrhythmia and ischemia. A total of 30,000 ECG segments obtained from the MIT-BIH Arrhythmia and European ST-T Ischemia databases were used to quantify this approach. 60% of the beats were used for training the classifier and the remaining 40%, for validating it. An overall accuracy of 99% and 98.24% was obtained for k-NN and GMM, respectively. This result is significantly better than that of currently used state-of-the-art classification approaches for arrhythmia and ischemia.

A neural network approach to mammogram image classification using fractal features

This paper presents a novel method based on fractal features for the classification of mammogram images. For recognition of regions and objects in the natural scenes, there is always a need for features, which are invariant, and they provide a good set of descriptive values for the region. There are numerous methods available to estimate parameters from the images of the fractal surface. In this paper we perform mammogram image classification based on fractal dimension and fractal signature. The result has shown the potential usefulness of the fractal features for image analysis. A trainable multilayer feed forward neural network has been designed for the classification of images. Experimental results shown that the proposed system can well perform a classification rate of 98%.

An Awareness Approach to Analyze ECG Streaming Data

Real-time remote health monitoring systems are experiencing tremendous advancement resulting from improvements in low power, reliable sensors; yet they are still constrained to low-level interpretation. Automatic data analysis continues to be a tedious task due to a lack of efficient, reliable platforms for data analysis. In this paper, we present a system for monitoring patients remotely by emphasizing the strength of Complex Event Processing (CEP) and Situation Awareness. In this approach, the system makes decisions in a declarative way, which helps medical experts to understand the situation in a more realistic manner. The primary objective of this paper is to explicate the different components inside the system. To verify the technical feasibility of each component, the proposed system is implemented and tested using ECG data.

Middleware Integration of DDS and ESB for Interconnection between Real-Time Embedded and Enterprise Systems

As technology evolves, the border of embedded system and enterprise system is gradually meaningless. ESB is a well-known middleware for integration of enterprise services and systems, and DDS is a promising real-time communication middleware for embedded systems. This paper introduces structurally integrated architecture between DDS middleware and ESB middleware for interconnection between real-time embedded systems and enterprise information systems. An integrated architecture of DDS and ESB allows integration of embedded system environment and enterprise system environment. It supports integration of diverse embedded devices from ubiquitous systems with real-time communication and allows effective data processing using enterprise services, which are generated from embedded systems. We design DDS binding component for ESB to allow DDS communication to enterprise services. This DDS-ESB integration architecture could be applied in the field that generates huge of data for monitoring and processes data for providing user services in real-time, such as u-City or home network.

Performance Evaluation for Efficient Complex Event Processing System According to Event Types and Environment

Complex event processing (CEP) is a relatively new technology for processing and analyzing multiple events from distributed sources, with the objective of extracting useful information from them. We need architecture of efficient complex event transporting and processing system according to event types and environment because of these events generate in various environment. It is not only generated in simple system but also generated complex system such as U-City, integrated meteorological information system (IMIS) that needs integration with other system using enterprise service bus (ESB) for extending system and enterprise event processing. In this paper, we propose the guideline of applying method for architecture of system that processes complex event using CEP engine and optimized transfer protocol according to characteristic of generated events and environment through performance test.

Applied Self-Healing: Self-Aware Service Request Optimization for Gateway Stability in Ubiquitous Zone-Based Networks

In this paper, we present an application of self-healing systems in the form of service request optimization algorithm for service gateways. This algorithm preestimates the load of the gateway and accepts the service request according to the capability of the worker thread. The experimental and simulation results show that the performance of load control algorithm works well for managing the service request. Our study has revealed that normally the networks fail to entertain the requests after the saturation point. Whereas the algorithm proposed in this paper, virtually accepts the service requests almost always which is what is essentially required from hybrid network in a ubiquitous environment. The algorithm proposed is already embedded and are in work in the IT infrastructure of major banks of Korea.

Road traffic big data collision analysis processing framework

With the advancement of sensor technologies, big data processing becomes a new paradigm for large scale information processing. Big data comes from different sources, such as from traffic information, social sites, mobile phone GPS signals and so on. In this paper, we propose a new architecture for distributed processing that enables big data processing on the road traffic data and its related information analysis. We applied Hadoop and HBase that can store and analyze real-time collision data in a distributed processing framework. This framework is designed as flexible and scalable framework using distributed CEP that process massive real-time traffic data and ESB that integrates other services. We tested the proposed framework on road traffic data on a 45-mile section of I-880N freeway CA, USA. By integrating freeway traffic big data and collision data over a ten-year period (1TB Size), we obtained the collision probability.

WiP Abstract: Cyber-Physical Systems for Real Time Cardiac Monitoring

The aim of this paper is to propose a cyber-physical systems(CPS) for monitoring the cardiac rhythm of a patient who had recently suffered from chronic heart disease. Doctor can monitor and collaborate with other doctors in virtual space and analyze the patient behavior virtually thus, provide intelligent decision in a timely manner.

A Comparative Performance Analysis of CBD Application Framework with JBean Application Framework for Improved Distributed Application Development

We observe that the e-business systems development frameworks tradeoff performance at the expense of flexibility. In this paper, we present a performance comparison of JBean application framework with a well-known framework, Struts. JBean is a flexible and extensible CBD application framework. Our experiment results show the significance of JBean application framework over contemporary CBD application frameworks.