DCBIOSTEC 2015 Abstracts

Short Papers
Short Papers
Paper Nr: 4
Title:

A Stationary Bike in Virtual Reality - Rhythmic Exercise and Rehabilitation
Authors:

Justyna Maculewicz, Stefania Serafin and Lise Kofoed
Abstract: This paper presents a project dedicated to the development of a rhythmic rehabilitation device for physical and neurological rehabilitation with the use of a stationary bike and a virtual reality (VR) setup augmented with soundscape sounds to create higher level of immersion. We discuss the benefits of the use of visual and auditory feedback and identify possible challenges for development. In addition, we comment on existing systems, which combine a stationary bike and VR (in any form) and match clinical groups that could benefit from this rehabilitation approach. Finally, we discuss the role of sound and music in enhancement of fitness performance and potential of benefit of use in neurorehabilitation.
Download
Paper Nr: 5
Title:

A Web Application for Automatic Analysis on Life-style Factors Affecting Personal Health from Self-tracking Data - Towards User-side Statistics Free Personal Data Analysis
Authors:

Zilu Liang
Abstract: The advent of commercial portable sensing devices has enabled many non-experts to collect their own data, and there has been a boom in health-centric self-monitoring and tracking. However, huge amount of these data remain unanalyzed simply because many of the data owners have no idea what to do with the large amount of data they have collected. Existing data analysis software tools were designed for statisticians or data scientists who have strong background in related fields. These tools are thus not usable for non-experts who have limited or no knowledge on statistics and programming. As more and more people start to collect their own data, it becomes important to solve the following problem: How to empower non-experts with an effective and easily usable tool to analyze their personal data? This project aims to address this problem by developing an online data analysis software tool to help non-experts gain insights from their personal data at simply few clicks, which requires no statistics background from the users. I have developed a prototype of the proposed web application. It is expected that the developed data analysis web application will not only help individuals identify the critical life-style factors that affect their health conditions and thus make it possible for them to personalize their healthcare plans for the best health outcome, but also help reduce public health cost and potential financial lost associated with poor health of the working population.
Download
Paper Nr: 6
Title:

Cardiac Arrhythmias Classification and Compression using a Hybrid Technique
Authors:

Asiya Al-Busaidi, Lazhar Khriji and Abdulnasir Y. Hossen
Abstract: This research work discusses the challenges and limitations of real-time analysis methods of low-powered wireless sensor networks for health monitoring. The work focuses on compression of ECG signal and classification of cardiac arrhythmias. Since, the discrimination of cardiac arrhythmia is still an open research field and many classification techniques have not been tested on ECG signals yet, more investigation on hybrid classification method will be conducted. The hybrid compression and classification techniques showed promising performance compared to the classical techniques. The main significant contribution of this work is to integrate the compression and classification algorithms with less number of steps to reduce the computational load and complexity of the system. Initial prospective is to apply compression on the decomposed coefficients and then after decompressing those coefficients, features are extracted from them and used for classification.
Download
Paper Nr: 8
Title:

Use of the Heart Rate Variability as a Diagnostic Tool
Authors:

Raquel Gutiérrez Rivas, Juan Jesús García Domínguez and William P. Marnane
Abstract: The electrocardiographic signal represents the electrical activity of the heart. It has several nodes able to generate synchronized electrical impulses to sequentially activate its valves. All this impulses overlapped form the well-known QRS complex (Figure 1). Usually, the position of the R peak is taken as the instant in which the heartbeat has place. Thus, to determine the heart rate it is necessary to find all the R peaks present during the measurement of the ECG signal. Heart Rate (HR) is controlled by the Autonomous Nervous System (ANS), which is composed by the Sympathetic Nervous System (SNS) and the Parasympathetic Nervous System (PNS). Both of them, SNS and PNS, respond to the necessities of the rest of physiologic systems (thermoregulatory, vasomotor, respiratory, central nervous, etc. systems) which make possible to correlate variations in the HR with the performance of all those systems. In short, due to the easiness with which is possible to obtain the ECG signal, and taking into consideration that is taken through a non-invasive measurement, several parameters of it have been studied for helping to the diagnosis of several diseases and as a tool to study the patients’ fitness. However, to study carefully the performance of those physiological systems, in most of the cases it is not only enough to know just the HR, but also the Heart Rate Variability (HRV), which is the focus of the study carried out through this thesis.
Download
Paper Nr: 9
Title:

Development of an Automated System for Ex Vivo Measuring the Neuro Muscular Junction Functionality
Authors:

Simona Pisu, Emanuele Rizzuto, Antonio Musarò and Zaccaria Del Prete
Abstract: The loss of functional connection between muscle and nerve is a crucial biological mechanism involved in several neuromuscular diseases, as Amyotrophic Lateral Sclerosis (ALS). ALS is a neurodegenerative disease associated with motor neuron degeneration, muscle atrophy and paralysis. In this context, the aim of this work is to characterize the functionality of the communication between muscle and nerve in mouse models by the development of new automated experimental methodologies. We developed an ex vivo technique based on the comparison between muscle contractile response due to membrane stimulation and muscle contractile response due to nerve stimulation. Since this latter stimulation bypasses the neuronal signalling, any difference between the two contractile responses may be related to NMJ alterations. Once the system and the stimulation protocol were set we started investigating the SOD1(G93A) mouse, the most studied animal model of ALS. Preliminary results from the transgenic model are in accordance with the literature, showing muscle contraction defects and NMJ impairment.
Download
Paper Nr: 11
Title:

Towards Creating an Iso-semantic Lexicon Model using Computational Semantics and Sublanguage Analysis Within Clinical Subdomains for Medical Language Processing
Authors:

B. S. Begum Durgahee and Adi Gundlapalli
Abstract: Although the widespread adoptions of Electronic Health Records (EHR) have made clinical data available in electronic format, a significant amount of important information is represented in unstructured narrative (free text) form. This complicates the use of these clinical data for decision support and research. Recent efforts have been aimed at applying natural language processing (NLP) and information extraction (IE) techniques to clinical text. A common practice is to manually construct semantic lexicons and use those to identify and extract clinical entities for specific tasks such as cohort identification and phenotyping. Besides requiring intensive manual, linguistic and medical knowledge, the vocabulary tends to be restricted to specific institutions and groups of users. There is no standardized way of building lexicons and this impedes the performance of the NLP or IE, due to inconsistent word usage. The objective of the proposed research study is to find a method of leveraging semantic lexicons to enable sharing of lexicons for information extraction from clinical text. Current NLP tools are mostly focused on clinical entity extraction by mapping textual elements to available ontologies. This method is insufficient due to ontology incompleteness and context dependent entities. Hence, there is a need for deeper understanding of relations among these entities in order to expand existing dictionaries accordingly. Lexico-semantic relations and patterns from heterogeneous clinical text will be detected in terms of sublanguage specific patterns. The discovered significant relations and patterns will be used with unsupervised methods, formal concept analysis, distributive analysis techniques and existing ontologies to inform the design of a learning-based system for automatic construction of clinical ontology-based lexicons. The Semantic Web technologies will be investigated to build a common ontology-based lexicons, using ontological and lexical representations. The ultimate goal of using Semantic Web technologies will be to interlink lexical resources with Biomedical ontologies in a computable form for sharing. This research proposal will contribute to the discovery of new concepts and relations in the clinical domain of interest. While automating the ontology-based lexicon construction with minimal supervised learning, we intend to enhance word sense and improve the text processing to retrieve accurate results. The resulting ontology-based semantic lexicon model will provide a new perspective towards standardizing semantic lexicons to facilitate content interoperability for clinical text mining and natural language processing tasks. Such a model will be helpful in predictive modeling studies for personalized healthcare to provide better health care with more efficient use of limited resources.
Download