Christos Panagiotou

On the Detection of Myocadial Scar Based on ECG/VCG Analysis

In this paper, we address the problem of detecting the presence of a myocardial scar from the standard electrocardiogram (ECG)/vectorcardiogram (VCG) recordings, giving effort to develop a screening system for the early detection of the scar in the point-of-care. Based on the pathophysiological implications of scarred myocardium, which results in disordered electrical conduction, we have implemented four distinct ECG signal processing methodologies in order to obtain a set of features that can capture the presence of the myocardial scar. Two of these methodologies are: 1) the use of a template ECG heartbeat, from records with scar absence coupled with wavelet coherence analysis and 2) the utilization of the VCG are novel approaches for detecting scar presence. Following, the pool of extracted features is utilized to formulate a support vector machine classification model through supervised learning. Feature selection is also employed to remove redundant features and maximize the classifiers performance. The classification experiments using 260 records from three different databases reveal that the proposed system achieves 89.22% accuracy when applying tenfold cross validation, and 82.07% success rate when testing it on databases with different inherent characteristics with similar levels of sensitivity (76%) and specificity (87.5%).

Performance evaluation of a WSN system for distributed event detection using fuzzy logic

The research field of event detection in realistic WSN environments has attracted a lot of interest, with health monitoring being one of its most pronounced applications. Although efforts related to the healthcare applications exist in the current literature, there is a significant lack of investigation on the performance of such systems, when applied in error prone and limited resource wireless environments. This paper aimed to address this need by porting a Fuzzy Inference System (FIS) to a WSN simulation framework. The considered FIS is implemented on TelosB motes and evaluates the health status of a monitored person, in an energy conserving manner. A distributed implementation of the above FIS is also proposed, comprising an additional contribution of this paper, based on an objective function, attempting to reduce the network congestion and balance the energy consumption between network nodes. This work presents a thorough performance evaluation of the FIS under the distributed and the centralized approach, while varying the communication conditions and highlighting the advantages of the distributed execution of the FIS, leading to packet loss gain and transmission gain up to 67% and 25% respectively. The networking benefits from the distributed approach are reflected to the FIS performance. Respective results and comparative evaluation against Matlab simulations reveal strong dependencies of the applications performance to critical WSN network parameters.

Data merge: A data aggregation technique for wireless sensor networks

The Wireless Sensor Networks (WSNs) have limited power and communication capabilities, combined with the requirement for long network lifetime. To increase it, methods to reduce energy consumption are highly required. To achieve this goal, we study a data aggregation technique without size reduction, i.e. data merge. It is a generic technique, since it is also usable in applications with heterogeneous data and requirements for high accuracy. This study presents the impact of the data merge technique on WSNs applications executed under various realistic data flow scenarios, traffic loads and wait time intervals. Our results show significant reductions in both packet loss and radio energy consumption.

Segmentation and reassembly data merge (SaRDaM) technique for Wireless Sensor Networks

The Wireless Sensor Networks (WSNs) have limited power capabilities whereas they require long network lifetime. To increase the latter, techniques to reduce energy consumption are highly required. This study proposes such a technique which explores the benefits of data aggregation without size reduction, under various parameters in data flows scenarios. The proposed technique exploits the potential of utilizing the maximum allowed packet size. Actual data from incoming packets are appended to already buffered packets, until the maximum packet size or a maximum waiting time is reached. Hence, the number of transmissions is reduced and redundant header transmissions are avoided, leading to an overall gain in energy consumption. Our study explores the impact of the proposed data merge mainly in all-to-one data flow scenarios executed under various traffic loads, wait time intervals and number of merging nodes. Our results show gain up to 56% in packet loss and 46% in energy consumption compared to a direct forwarding of packets.

Network driven cache behavior in wireless sensor networks

It is becoming increasingly apparent that approaches and methodologies traditionally used in on-chip and off-chip cache memories can offer significant benefits to a wide range of wireless network scenarios. A prominent relative area is Wireless Sensor Networks characterized by a wide range of advantages but, at the same time, extremely limited and scarce resources. However, most of the relative efforts presented so far do not consider the unique conditions and respective challenges imposed by wireless communication medium even more in industrial environments. This paper presents our initial study in developing cross-layer approaches among MAC layer and cache management in three axis: (i) adaptive cache invalidation technique, (ii) application and network driven cache replacement techniques and (iii) adaptive cache prefetching techniques. Additionally, simulation based performance evaluation is presented revealing promising results regarding resource conservation due to the use of proposed methodologies in wireless networks of limited resources.

Priority Handling Aggregation Technique (PHAT) for Wireless Sensor Networks

Wireless Sensor Networks (WSNs) have limited power capabilities, whereas they serve applications which usually require specific packets, i.e. High Priority Packets (HPP), to be delivered before a deadline. Hence, it is essential to reduce the energy consumption and to have real-time behavior. To achieve this goal we propose a hybrid technique which explores the benefits of data aggregation without data size reduction in combination with prioritized queues. The energy consumption is reduced by appending data from incoming packets with already buffered Low Priority Packets (LPP). The real-time behavior is achieved by directly forwarding the HPP to the next node. Our study explores the impact of the proposed hybrid technique in several all-to-one data flow scenarios with various traffic loads, wait time intervals and percentage of HPP. Our results show gain up to 23,3% in packet loss and 36,6% in energy consumption compared with the direct forwarding of packets.

Performance enhancement in WSN through data cache replacement policies

It is increasingly regarded that data local caching in WSN can benefit the resource conservation and performance. A critical challenge is that cached data comprise only a small percentage of the data that a WSN node may require, thus leading to the replacement of data items. Consequently, replacement policy comprises critical factors in order to maximize performance benefits. Equally important in WSN is the interplay of such policies, application and network characteristics. Therefore, this paper aims to experimentally evaluate prominent replacement policies performance with respect to critical application and network parameters. An already presented cache model is developed and appropriately extended in the context of the prominent WSN TelosB platform. Valuable insights are provided concerning the behavior of considered policies in typical WSN scenarios and important trade-offs are revealed.

Sleep monitoring classification strategy for an unobtrusive EEG system

The advances in the wearable devices and Artificial Intelligence domains highlight the need for ICT systems that aim in the improvement of humans quality of life. In this paper we present the sleeping tracking component of an activity and sleeping tracking system. We present the sleep quality assessment based on EEG processing and support vector machines with sequential minimal optimization classifiers (SVM-SMO). The performance of the system demonstrated by respective experiments (accuracy: 83% and kappa coeff: 72%) exhibits significant prospects for the assessment of sleep quality and the further validation through an evaluation study.

Detection of myocardial scar from the VCG using a supervised learning approach

This paper addresses the possibility of detecting presence of scar tissue in the myocardium through the investigation of vectorcardiogram (VCG) characteristics. Scarred myocardium is the result of myocardial infarction (MI) due to ischemia and creates a substrate for the manifestation of fatal arrhythmias. Our efforts are focused on the development of a classification scheme for the early screening of patients for the presence of scar. More specifically, a supervised learning model based on the extracted VCG features is proposed and validated through comprehensive testing analysis. The achieved accuracy of 82.36% (sensitivity 84.31%, specificity 77.36%) indicates the potential of the proposed screening mechanism for detecting the presence/absence of scar tissue.

An agent based middleware imposing intelligence over critical infrastructures utilizing Wireless Sensor Networks

A modular architecture for power constrained embedded devices which leverages an agent based middleware in order to impose intelligence over critical infrastructures which require real time actions is always desirable in the case of Wireless Sensor Networks. The main objective of such architecture is the integration of a Wireless Sensor Network with the internet. This objective indicates the need of Ipv6 ready devices as well as the utilization of newly emerged standards such as the 6lowpan which forms an abstraction layer between the Medium Access control and the IP layers.