BIOSIGNALS 2009 Abstracts


Full Papers
Paper Nr: 11
Title:

ARRAY-BASED GENOME COMPARISON OF ARABIDOPSIS ECOTYPES USING HIDDEN MARKOV MODELS

Authors:

Ali Banaei, Andreas Houben, François Roudier, Ivo Grosse, Jens Keilwagen, Marc Strickert, Michael Seifert, Michael Florian Mette and Vincent Colot

Abstract: Arabidopsis thaliana is an important model organism in plant biology with a broad geographic distribution including ecotypes from Africa, America, Asia, and Europe. The natural variation of different ecotypes is expected to be reflected to a substantial degree in their genome sequences. Array comparative genomic hybridization ( ACGH ) can be used to quantify the natural variation of different ecotypes at the DNA level. Besides, such ACGH data provides the basics to establish a genome-wide map of DNA copy number variation for different ecotypes. Here, we present a new approach based on Hidden Markov Models (HMMs) to predict copy number variations in ACGH experiments. Using this approach, an improved genome-wide characterization of DNA segments with decreased or increased copy numbers is obtained in comparison to the routinely used segMNT algorithm. The software and the data set used in this case study can be downloaded from http://dig.ipk-gatersleben.de/HMMs/ACGH/ACGH.html.

Paper Nr: 21
Title:

MAPPING LANDMARKS ON TO THE FACE

Authors:

B. J. Boom, G. M. Beumer and R. N. J. Veldhuis

Abstract: Landmarking can be formalised as calculating the Maximum A-posteriori Probability (MAP) of a set of landmarks given an image (texture) containing a face. In this paper a likelihood-ratio based landmarking method is extended to a MAP-based landmarking method. The approach is validated by means of experiments. The MAP approach turns out to be advantageous, particularly for low quality images, in which case the landmarking accuracy improves significantly.

Paper Nr: 22
Title:

THERMOGRAPHIC BODY TEMPERATURE MEASUREMENT USING A MEAN-SHIFT TRACKER

Authors:

Guillaume-Alexandre Bilodeau, J. M. Pierre Langlois, Lionel Carmant, Maxime Levesque and Pablo Lema

Abstract: In epilepsy research, using a wide range of sensors can help to automatically detect the occurrence of seizures and to understand their underlying mechanisms. One such sensor is a thermographic camera that can measure the surface temperature of the body. This sensor may have an important role in investigating seizures as studies have shown that they can affect the body temperature of a patient. Furthermore, it has also been shown that kainic acid, a drug used to provoke seizures in animals, has an impact on rat body temperature. Consequently, there is a need to continuously measure the evolution of the body temperature of an animal during seizures. In this paper, we present our developed methodology to measure the temperature of a moving rat using a thermographic camera. To accurately measure the body temperature, we propose a methodology using a Mean-Shift tracker. The obtained measures are compared with a ground truth. The method is tested on a 2-hour video, and it is shown that the Mean-Shift tracker achieves an RMS error of approximately 0.1ºC.

Paper Nr: 27
Title:

MODELING AND SIMULATION OF BIODEGRADATION OF XENOBIOTIC POLYMERS BASED ON EXPERIMENTAL RESULTS

Authors:

Fusako Kawai and Masaji Watanabe

Abstract: Biodegradation of polyethylene glycol is studied mathematically. A mathematical model for depolymerization process of exogenous type is described. When a degradation rate is a product of a time factor and a molecular factor, a time dependent model can be transformed into a time independent model, and techniques developed in previous studies can be applied to the time independent model to determine the molecular factor. The time factor can be determined assuming the exponential growth of the microbial population. Those techniques are described, and numerical results are presented. A comparison between a numerical result and an experimental result shows that the mathematical method is appropriate for practical applications.

Paper Nr: 40
Title:

VARIABLE SUBSET SELECTION FOR BRAIN-COMPUTER INTERFACE - PCA-based Dimensionality Reduction and Feature Selection

Authors:

J. H. Correia, M. Kamrunnahar, N. S. Dias, P. M. Mendes and S. J. Schiff

Abstract: A new formulation of principal component analysis (PCA) that considers group structure in the data is proposed as a Variable Subset Selection (VSS) method. Optimization of electrode channels is a key problem in brain-computer interfaces (BCI). BCI experiments generate large feature spaces compared to the sample size due to time limitations in EEG sessions. It is essential to understand the importance of the features in terms of physical electrode channels in order to design a high performance yet realistic BCI. The VSS produces a ranked list of original variables (electrode channels or features), according to their ability to discriminate between tasks. A linear discrimination analysis (LDA) classifier is applied to the selected variable subset. Evaluation of the VSS method using synthetic datasets selected more than 83% of relevant variables. Classification of imagery tasks using real BCI datasets resulted in less than 16% classification error.

Paper Nr: 44
Title:

MFCC-BASED REMOTE PATHOLOGY DETECTION ON SPEECH TRANSMITTED THROUGH THE TELEPHONE CHANNEL - Impact of Linear Distortions: Band Limitation, Frequency Response and Noise

Authors:

Corinne Fredouille, Juan I. Godino-Llorente, Nicolás Sáenz-Lechón, Rubén Fraile and Víctor Osma-Ruiz

Abstract: Advances in speech signal analysis during the last decade have allowed the development of automatic algorithms for a non-invasive detection fo laryngeal pathologies. Performance assessment of such techniques reveals that classification success rates over 90% are achievable. Bearing in mind the extension of these automatic methods to remote diagnosis scenarios, this paper analyses the performance of a pathology detector based on Mel Frequency Cepstral Coefficients when the speech signal has undergone the distortion of an analogue communications channel, namely the phone channel. Such channel is modeled as a concatenation of linear effects. It is shown that while the overall performance of the system is degraded, success rates in the range of 80% can still be achieved. This study also shows that the performance degradation is mainly due to band limitation and noise addition.

Paper Nr: 46
Title:

CHROMOSOME REGION RECOGNITION WITH LOCAL BAND PATTERNS

Authors:

Chieko Hamada, Tetsuo Kinoshita and Toru Abe

Abstract: To make the visual examination of a chromosome image for various chromosome abnormalities, individual chromosome regions have to be extracted from the subject image and classified into the distinct chromosome types. To improve the accuracy and flexibility in this process, we propose a subregion (local band pattern) based method for recognizing chromosome regions in the image. This method regards each chromosome region as a series of subregions, and iterates a search for subregions in the image consecutively. Consequently, chromosome region classification is performed simultaneously with its extraction for each subregion. Since the dimensions and intensities of chromosome regions vary with every image, effective subregion searches require templates whose dimensions and intensities correspond with those of chromosome regions in the image. To develop an effective subregion search, we also propose a method for adjusting the dimensions of templates to those of chromosome regions in the image and adapting the intensities in the image to those of the templates.

Paper Nr: 50
Title:

A COMPUTATIONAL SALIENCY MODEL INTEGRATING SACCADE PROGRAMMING

Authors:

Anne Guérin-Dugué, Nathalie Guyader and Tien Ho-Phuoc

Abstract: Saliency models have showed the ability of predicting where human eyes fixate when looking at images. However, few models are interested in saccade programming strategies. We proposed a biologically-inspired model to compute image saliency maps. Based on these saliency maps, we compared three different saccade programming models depending on the number of programmed saccades. The results showed that the strategy of programming one saccade at a time from the foveated point best matches the experimental data from free viewing of natural images. Because saccade programming models depend on the foveated point where the image is viewed at the highest resolution, we took into account the spatially variant retinal resolution. We showed that the predicted eye fixations were more effective when this retinal resolution was combined with the saccade programming strategies.

Paper Nr: 51
Title:

IMPROVING CLASSIFICATION FOR BRAIN COMPUTER INTERFACES USING TRANSITIONS AND A MOVING WINDOW

Authors:

Inés M. Galván, José M. Valls and Ricardo Aler

Abstract: The context of this paper is the brain-computer interface (BCI), and in particular the classification of signals with machine learning methods. In this paper we intend to improve classification accuracy by taking advantage of a feature of BCIs: instances run in sequences belonging to the same class. In that case, the classification problem can be reformulated into two subproblems: detecting class transitions and determining the class for sequences of instances between transitions. We detect a transition when the Euclidean distance between the power spectra at two different times is larger than a threshold. To tackle the second problem, instances are classified by taking into account, not just the prediction for that instance, but a moving window of predictions for previous instances. Our results are competitive with those obtained in the BCI III competition.

Paper Nr: 53
Title:

MODIFICATIONS AND IMPROVEMENTS ON IRIS RECOGNITION

Authors:

André Lourenço, Artur Ferreira, Bárbara Pinto and Jorge Tendeiro

Abstract: Iris recognition is a well-known biometric technique. John Daugman has proposed a method for iris recognition, which is divided into four steps: segmentation, normalization, feature extraction and matching. In this paper, we evaluate, modify and extend John Daugman’s method. We study the images of CASIA and UBIRIS databases to establish some modifications and extensions on Daugman’s algorithm. The major modification is on the computationally demanding segmentation stage, for which we propose a template matching approach. The extensions on the algorithm address the important issue of pre-processing, that depends on the image database, being especially important when we have a non infra-red red camera (e.g. a WebCam). For this typical scenario, we propose several methods for reflexion removal and pupil enhancement and isolation. The tests, carried out by our C# application on grayscale CASIA and UBIRIS images, show that our template matching based segmentation method is accurate and faster than the one proposed by Daugman. Our fast pre-processing algorithms efficiently remove reflections on images taken by non infra-red cameras.

Paper Nr: 58
Title:

COMPUTATIONAL MUSCLE REFLEX MODEL OF WHEELCHAIR USERS TRAVELING IN MOTOR VEHICLES - Evaluation of the Motion and the Myoelectric Potential of People with Disabilities

Authors:

Emiko Kikuchi, Junichi Takano and Shigeru Aomura

Abstract: In this study, a human motion computer model in a wheelchair was developed to evaluate the effectiveness of a seatbelt for people with disabilities traveling in a motor vehicle. The human model was composed of two rigid links and three masses. This model was characterized with muscle reflection defined by Hill’s equation. A sudden stop experiment by using a carriage on which a wheelchair was fixed with a subject was performed to obtain the human muscle parameters and to evaluate the model. Volunteer subjects including people with disabilities participated in the experiment. The motion and muscle activity of a subject wearing a seatbelt were simulated by this model. The muscle reflection of people with disabilities was stronger than that of normal people in the case of not using a seatbelt, but in the case of using a seatbelt the muscle reflection of people with disabilities was similarly weak with normal people. The result of computer simulation showed that a seatbelt is more important for people with disabilities than for normal people.

Paper Nr: 61
Title:

DRUG ADDICTION: A COMPUTATIONAL MULTISCALE MODEL COMBINING NEUROPSYCHOLOGY, COGNITION AND BEHAVIOR

Authors:

Yariv Z. Levy, Dino Levy, Jerrold S. Meyer and Hava T. Siegelmann

Abstract: According to the United Nations, approximately 24.7 million people used amphetamines, 16 million used cocaine, and 12 million used heroin in 2006/07 (Costa, 2008). Full recovery from drug addiction by chemical treatment and/or social and psychological support is uncertain. The present investigation was undertaken to expand our understanding of the factors that drive the dynamics of addiction. A new multiscale computational model is presented which integrates current theories of addiction, unlike previous models, considers addiction as a reversible process (Siegelmann, 2008). Explicit time dependency is added to the inhibition and the compulsion processes. Preliminary computational predictions of drug-seeking behavior are presented and potential correlation with experimental data is discussed. Validation of the model appears promising, however additional investigation is required.

Paper Nr: 62
Title:

3D RECONSTRUCTION FOR TEXTURELESS SURFACES - Surface Reconstruction for Biological Research of Bryophyte Canopies

Authors:

Eduard Sojka, Jan Gaura, Michal Krumnikl and Oldřich Motyka

Abstract: This paper covers the topic of three dimensional reconstruction of small textureless formations usually found in biological samples. Generally used reconstructing algorithms do not provide sufficient accuracy for surface analysis. In order to achieve better results, combined strategy was developed, linking stereo matching algorithms with monocular depth cues such as depth from focus and depth from illumination. Proposed approach is practically tested on bryophyte canopy structure. Recent studies concerning bryophyte structure applied various modern, computer analysis methods for determining moss layer characteristics drawing on the outcomes of a previous research on surface of soil. In contrast to active methods, this method is a non-contact passive, therefore, it does not emit any kind of radiation which can lead to interference with moss photosynthetic pigments, nor does it affect the structure of its layer. This makes it much more suitable for usage in natural environment.

Paper Nr: 68
Title:

DETECTION OF THE CRITICAL POINT INTERVAL OF POSTURAL CONTROL STRATEGY USING WAVELET TRANSFORM ANALYSIS

Authors:

David J. Hewson, Hichem Snoussi, Jacques Duchêne and Neeraj Kumar Singh

Abstract: Postural balance is often studied in order to understand the effect of sensory degradation with age. The aim of this study was to develop a new method of detecting the critical point interval (CPI) at which sensory feedback is used as part of a closed-loop postural control strategy. Postural balance was evaluated using centre of pressure (COP) displacements measured using a force plate for 17 control subjects and 10 elderly subjects under control (eyes open) and experimental (eyes closed, vibration) conditions. A modified local-maximum-modulus wavelet transform analysis using the power spectrum of COP signals was used to calculate the critical point when closed-loop control occurs. Lower values of CPI are associated with increased closed-loop postural control, indicating a quicker response to sensory input. This strategy of postural control will require greater energy expenditure due to the repeated muscular interventions in order to remain stable. The CPI for elderly subjects occurred significantly quicker than for control subjects, indicating that posture was more closely controlled. Similar results were observed for eyes closed and vibration conditions. The CPI parameter offers a new method of detecting differences in postural control between different experimental conditions or changes due to ageing.

Paper Nr: 80
Title:

CLOSURE IN ARTIFICIAL CELL SIGNALLING NETWORKS - Investigating the Emergence of Cognition in Collectively Autocatalytic Reaction Networks

Authors:

James Decraene

Abstract: Cell Signalling Networks (CSNs) are complex biochemical networks responsible for the coordination of cellular activities in response to internal and external stimuli. We hypothesize that CSNs are subsets of collectively autocatalytic reaction networks. The signal processing or cognitive abilities of CSNs would originate from the closure properties of these systems. We investigate how Artificial CSNs, regarded as minimal cognitive systems, could emerge and evolve under this condition where closure may interact with evolution. To assist this research, we employ a multi-level concurrent Artificial Chemistry based on the Molecular Classifier Systems and the Holland broadcast language. A critical issue for the evolvability of such undirected and autonomous evolutionary systems is to identify the conditions that would ensure evolutionary stability. In this paper we present some key features of our system which permitted stable cooperation to occur between the different molecular species through evolution. Following this, we present an experiment in which we evolved a simple closed reaction network to accomplish a pre-specified task. In this experiment we show that the signal-processing ability (signal amplification) directly resulted from the evolved systems closure properties.

Paper Nr: 84
Title:

EMD-BASED ANALYSIS OF RAT EEG DATA FOR SLEEP STATE CLASSIFICATION

Authors:

Damien Gervasoni, Edmundo Pereira de Souza Neto, Patrice Abry, Paulo Gonçalves, Pierre-Hervé Luppi and Süleyman Baykut

Abstract: In this paper Empirical Mode Decomposition (EMD)-based features from single-channel electroencephalographic (EEG) data are proposed for rat’s sleep state classification. The classification performances of the EMD-based features and some classical power spectrum density (PSD)-based features are compared. Supported by experiments on real EEG data, we demonstrate that classification performances can significantly improve, by simply substituting EMD to PSD in features extraction. This is in noticeably due to the natural adaptivity of EMD which show more robust to subjects variability.

Paper Nr: 85
Title:

SEVERE APNOEA DETECTION USING SPEAKER RECOGNITION TECHNIQUES

Authors:

Doroteo T. Toledano, Eduardo López, José Alcazar, Jose Luis Blanco, Luis A. Hernández and Ruben Fernández

Abstract: The aim of this paper is to study new possibilities of using Automatic Speaker Recognition techniques (ASR) for detection of patients with severe obstructive sleep apnoea (OSA). Early detection of severe apnoea cases can be very useful to give priority to their early treatment optimizing the expensive and time-consuming tests of current diagnosis methods based on full overnight sleep in a hospital. This work is part of an on-going collaborative project between medical and signal processing communities to promote new research efforts on automatic OSA diagnosis through speech processing technologies applied on a carefully designed speech database of healthy subjects and apnoea patients. So far, in this contribution we present and discuss several approaches of applying generative Gaussian Mixture Models (GMMs), generally used in ASR systems, to model specific acoustic properties of continuous speech signals in different linguistic contexts reflecting discriminative physiological characteristics found in OSA patients. Finally, experimental results on the discriminative power of speaker recognition techniques adapted to severe apnoea detection are presented. These results obtain a correct classification rate of 81.25%, representing a promising result underlining the interest of this research framework and opening further perspectives for improvement using more specific speech recognition technologies.

Paper Nr: 87
Title:

AUTOMATIC DETECTION OF ATRIAL FIBRILLATION AND FLUTTER - A Tachogram-based Algorithm for Mobile Devices

Authors:

Christophe Kunze, Malte Kirst and Stefanie Kaiser

Abstract: Two versions of a new detector for automatic real-time detection of atrial fibrillation and atrial flutter in non-invasive ECG signals are introduced. The methods are based on beat to beat variability, tachogram analysis and simple signal filtering. The implementation on mobile devices is made possible due to the low demand on computing power of the employed analysis procedures. The proposed algorithms correctly identified 436 of 440 five minute episodes of atrial fibrillation or flutter and also correctly identified up to 302 of 342 episodes of no atrial fibrillation, including normal sinus rhythm as well as other cardiac arrhythmias. These numbers correspond to a sensitivity of 99.1% and a specificity of 88.3%.

Paper Nr: 88
Title:

APPLICATION OF NEURAL NETWORKS IN AID FOR DIAGNOSIS FOR PATIENTS WITH GLAUCOMA

Authors:

Dário A. B. Oliveira, Mariana M. B. Oliveira, Marley M. B. R. Vellasco and Riuitiro Yamane

Abstract: Glaucoma is an ophthalmologic disease very difficult to diagnose in the earlier phase. Additionally, exams and methods used to give reliable information for correct diagnosis are usually very expensive. Therefore, other methods less expensive and also reliable must be proposed as an auxiliary tool to Glaucoma diagnosis. This paper analyzes the performance of neural networks as an auxiliary tool for the diagnosis of patients with glaucoma, avoiding the use of data only available in expensive exams. The analysis considers two different kinds of neural networks (Multi-Layer Perceptron (MLP) and Probabilistic Neural Networks (PNN)) and two different methods variable selection: a random and iterative method; and the Least Square Extrapolation (LSE) method. The paper also evaluates the benefits of applying principal components analysis (PCA) to the database. The results obtained were very good, attaining an accuracy of more than 90% of correct classification of all cases present in our database. It confirms the real possibility of using neural networks as an auxiliary and inexpensive tool to help in Glaucoma diagnosis.

Paper Nr: 92
Title:

AN EFFICIENT AND ROBUST TECHNIQUE OF T WAVE DELINEATION IN ELECTROCARDIOGRAM

Authors:

P. Gupta and Yogendra Narain Singh

Abstract: This study presented an efficient and robust method for the automatic delineation of T wave in the single-lead electrocardiogram. The method attained optimum performance using the fusion of delineation results obtained from a pair of new approaches. The first approach utilizes the advantage of time derivative and determines T wave ends using derivative curve analysis. The effect of local noise on the ECG signal is seized using a nonderivative approach which performs T wave delineation using the analysis of its waveform curvature toward the ends. Using the assumption that beginning and end of T wave exhibit the convex shape, this approach determines minimum radius of curvature of the convex regions at both ends. It is formally shown that the time instance corresponding to minimum radius of curvature coincides with T wave ends. The delineation results obtained from both the approaches are fused to achieve the optimum performance. The delineator attained a detection sensitivity of 99.9%, positive predictivity of 99.1% and an accuracy of 99.01% over the first lead of physionet QT database (20 records of 1, 000 beats each). The delineation errors are found well within the referenced intercardiologist observations, especially for T wave end. The mean error and standard deviation are found smaller than 10 ms which outperformed in comparison to other published results.

Paper Nr: 95
Title:

TOWARDS SPEAKER-ADAPTIVE SPEECH RECOGNITION BASED ON SURFACE ELECTROMYOGRAPHY

Authors:

Michael Wand and Tanja Schultz

Abstract: We present our recent advances in silent speech interfaces using electromyographic signals that capture the movements of the human articulatory muscles at the skin surface for recognizing continuously spoken speech. Previous systems were limited to speaker- and session-dependent recognition tasks on small amounts of training and test data. In this paper we present speaker-independent and speaker-adaptive training methods which for the first time allows us to use a large corpus of data from many speakers to reliably train acoustic models. On this corpus we compare the performance of speaker-dependent and speaker-independent acoustic models, carry out model adaptation experiments, and investigate the impact of the amount of training data on the overall system performance. In particular, since our data corpus is relatively large compared to previous studies, we are able for the first time to train an EMG recognizer with context-dependent acoustic models. We show that like in acoustic speech recognition, context-dependent modeling significantly increases the recognition performance.

Paper Nr: 96
Title:

ECG-BASED AUTHENTICATION - Bayesian vs. Nearest Neighbour Classifiers

Authors:

Ana Fred and Carla Oliveira

Abstract: This paper presents an approach for human authentication based on electrocardiogram (ECG) waveforms. ECG data was collected from 24 individuals during the realization of cognitive tests, where subjects held a surface mount triode placed on the V2 pre cordial derivation. Authentication is based on MAP, One-Class and 1-NN classifiers. Results show that ECG-based authentication may be a feasible tool for biometric systems. The One-Class classifier with class normalization has presented enhanced performance, with an equal error rate of 3.5%.

Paper Nr: 101
Title:

YEAST METABOLIC STATE IDENTIFICATION BY FIBER OPTICS SPECTROSCOPY

Authors:

C. C. Castro, J. S. Silva, R. C. Martins and V. V. Lopes

Abstract: In this manuscript we explore the feasibility of using LWUV-VIS-SWNIR (340 - 1100 nm) spectroscopy to classify Saccharomyces cerevisiae colony structures in YP agar and YPD agar, under different growth conditions, such as: i) no alcohol; ii) 1 % (v/v) Ethanol; iii) 1 % (v/v) 1-Propanol; iv) 1 % (v/v) 1- butanol; v) 1 % (v/v) Isopropanol; vi) 1 % (v/v) (±)-1-Phenylethanol; vii) 1 % (v/v) Isoamyl alcohol; viii) 1 % (v/v) tert-Amyl alcohol (2-Methyl-2-butanol); and ix) 1 % (v/v) Amyl alcohol. Results show that LWUV-VISSWNIR spectroscopy has the potential for yeasts metabolic state identification once the spectral signatures of colonies differs from each others, being possible to acheive 100% of classification in UV-VIS and VISSWNIR. The UV-VIS region present high discriminant information (350-450 nm), and different responses to UV excitation were obtained. Therefore, high precision is obtained because UV-VIS and VIS-NIR exhibit different kinds of information. In the future, high precision analytical chemistry techniques such as mass spectroscopy and molecular biology transcriptomic studies should be performed in order to understand the detailed cell metabolism and genomic phenomena that characterize the yeast colony state.

Paper Nr: 104
Title:

GENETIC OPTIMIZATION OF CEPSTRUM FILTERBANK FOR PHONEME CLASSIFICATION

Authors:

Diego H. Milone, Hugo L. Rufiner, John C. Goddard and Leandro D. Vignolo

Abstract: Some of the most commonly used speech representations, such as mel-frequency cepstral coefficients, incorporate biologically inspired characteristics into artificial systems. Recent advances have been introduced modifying the shape and distribution of the traditional perceptually scaled filterbank, commonly used for feature extraction. Some alternatives to the classic mel scaled filterbank have been proposed, improving the phoneme recognition performance in adverse conditions. In this work we propose an evolutionary strategy as a way to find an optimal filterbank. Filter parameters such as the central and side frequencies are optimized. A hidden Markov model classifier is used for the evaluation of the fitness for each possible solution. Experiments where conducted using a set of phonemes taken from the TIMIT database with different additive noise levels. Classification results show that the method accomplishes the task of finding an optimized filterbank for phoneme recognition.

Paper Nr: 124
Title:

A BIOMETRIC IDENTIFICATION SYSTEM BASED ON THYROID TISSUE ECHO-MORPHOLOGY

Authors:

Ana L. N. Fred and José C. R. Seabra

Abstract: This paper proposes a biometric system based on features extracted from the thyroid tissue accessed through 2D ultrasound. Tissue echo-morphology, which accounts for the intensity (echogenicity), texture and structure has started to be used as a relevant parameter in a clinical setting. In this paper, features related to texture, morphology and tissue reflectivity are extracted from the ultrasound images and the most discriminant ones are selected as an input for a prototype biometric identification system. Several classifiers were tested, with the best results (90% identification rate) being achieved with the maximum a posteriori classifier. Another classifier which only takes into account the reflectivity parameter achieved a reasonable identification rate of 70%. This suggests that the acoustic impedance (reflectivity) of the tissue is a good parameter to discriminate between individuals. This paper shows the effectiveness of the proposed classification, which can be used not only as a new biometric modality but also as a diagnostic tool.

Short Papers
Paper Nr: 1
Title:

SELF-ORGANIZING DSP CIRCUITS

Authors:

André Stauffer and Joël Rossier

Abstract: Living organisms are endowed with three structural principles: multicellular architecture, cellular division, and cellular differentiation. Implemented in digital according to these principles, our DSP circuits present self-organizing mechanisms like configuration, cloning, cicatrization, and regeneration. These mechanisms are made of simple processes such as growth, load, branching, repair, reset, and kill. The description of a configurable molecule implementing the self-organizing mechanisms and its application to a multiplier function constitute the core of this paper.

Paper Nr: 23
Title:

USING SUPPORT VECTOR MACHINES (SVMS) WITH REJECT OPTION FOR HEARTBEAT CLASSIFICATION

Authors:

Adel Belouchrani, Ahmed Amirou and Zahia Zidemal

Abstract: In this paper, we introduce a new system for ECG beat classification using Support Vector Machines (SVMs) classifier with a double hinge loss. This classifier has the option to reject samples that cannot be classified with enough confidence. Specifically in medical diagnoses, the risk of a wrong classification is so high that it is convenient to reject the sample. After ECG preprocessing, feature selection and extraction, our decision rule uses dynamic reject thresholds following the cost of rejecting a sample and the cost of misclassifying a sample. Significant performance enhancement is observed when the proposed approach was tested with the MIT/BIH arrythmia database. The achieved results are represented by the error reject tradeoff and a sensitivity higher than 99%, being competitive to other published studies.

Paper Nr: 25
Title:

MULTI-MODAL FUSION OF SPEECH-GESTURE USING INTEGRATED PROBABILITY DENSITY FUNCTION

Authors:

Chang-Seok Bae, Chi-Geun Lee, Jin-Tae Kim and Mun-Sung Han

Abstract: Recently, multi-modal recognition has become a hot topic in the field of Ubiquitous, Speech and gesture recognition, especially, are the most important modalities of human-to-machine interaction. Although speech recognition has been explored extensively and successfully developed, it still encounters serious errors in noisy environments. In such cases, gestures, a by-product of speech, can be used to help interpret the speech. In this paper, we propose a method of multi-modal fusion recognition of speech-gesture using integrated discrete probability density function omit estimated by a histogram. The method is tested with a microphone and a 3-axis accelerator in a real-time experiment. The test has two parts : a method of add-and-accumulate speech and gesture probability density functions respectively, and a more complicated method of creating new probability density function from integrating the two PDF’s of speech and gesture.

Paper Nr: 26
Title:

MULTI-CHIRP SIGNAL SEPARATION

Authors:

B. Dugnol, C. Fernández, G. Galiano and J. Velasco

Abstract: Assuming that an specific audio signal, such as recordings of animal sounds, may be modelled as an addition of nonlinear chirps, we use the quadratic energy distribution corresponding to the Chirplet Transform of the signal to produce estimates of the corresponding instantaneous frequencies, chirp-rates and amplitudes at each instant of the recording and design an algorithm for tracking and separating the chirp components of the signal. We demonstrate the accuracy of our algorithm applying it to some synthetic and field recorded signals.

Paper Nr: 30
Title:

SUBSET SELECTION OF MYOELECTRIC CHANNELS - A Genetic Algorithm for Subset Selection of Myoelectric Channels for Patients Following TMR Surgery

Authors:

Gernot Kvas and Rosemarie Velik

Abstract: State of the art self powered prostheses make use of the surface myoelectric signal for motor control. With increasing height of the amputation, control by residual muscles becomes less intuitive and physiologic. Targeted muscle reinnervation (TMR), a surgery technique to increase the number of control sites available in combination with multichannel surface electromyography allows for improved prosthesis control. This paper presents a genetic algorithm that determines a channel subset with high classification accuracy out of a given number of channels recorded from the reinnervated area of a patient.

Paper Nr: 31
Title:

A MULTIVARIATE STATISTICAL ANALYSIS OF MUSCULAR BIOPOTENTIAL FOR HUMAN ARM MOVEMENT CHARACTERIZATION

Authors:

Carlos Eduardo Thomaz, Giuliano Alves da Silva and Maria Cláudia Ferrari de Castro

Abstract: Pattern recognition of electromyographic signals consists of a hard task due to the high dimensionality of the data and noise presence on the acquired signals. This work intends to study the data set as a multivariate pattern recognition problem by applying linear transformations to reduce the data dimensionality. Five volunteers contributed in a previous experiment that acquired the myoelectrical signals using surface electrodes. Attempts to analyse the groups of acquired data by means of descriptive statistics have shown to be inconclusive. This works shows that the use of multivariate statistical techniques such as Principal Components Analysis (PCA) and Maximum uncertainty Linear Discriminant Analysis (MLDA) to characterize the acquired set of signals through low dimensional scatter plots provides a new understanding of the data spread, making easier its analysis. Considering the arm horizontal movement and the acquired set of data used in this research, a multivariate linear separation between the patterns of interest quantified by the distance of Bhattacharyya suggests that it’s possible not only to characterize the angular joint position, but also to confirm that different movements recruit similar amounts of energy to be executed.

Paper Nr: 32
Title:

TRANSFER FUNCTION OF THE HEART RATE CONTROL SYSTEM WITH RESPIRATORY INPUT - The Classical Engineering Approach

Authors:

Bronya Vaschillo and Evgeny G. Vaschillo

Abstract: The classic “control system theory” approach was used to find the transfer function (TF) of the HR control system with respiratory input. Eight healthy subjects, ages 19–40 participated in the study. Paced breathing at seven frequencies in 0.5– 0.04 Hz range was used as sine-wave stimuli to assess the HR control system. The sine-wave HR oscillation in response to each stimulus as the system’s output was recorded. Amplitude and phase TFs were calculated for each frequency separately. The Fourier filtration procedure was used for TF calculation. Experimentally obtained TFs revealed the same important features in all participants: 1) the amplitude TF had its peak in a narrow frequency range around 0.1 Hz; 2) the phase TF successively changed from positive to negative when breathing frequency increased, and passed “0” at frequency where the amplitude TF peaked; 3) the peak frequency and magnitude were unique for each participant. These features are evidence for the resonance property of the HR control system at a frequency around 0.1 Hz. This study suggests that accurate identification of an individual’s resonance frequency can be found using the TF features of HR control system and controlled breathing techniques.

Paper Nr: 33
Title:

ASSESSMENT AND COMPARISON OF TIME REALIGNMENT METHODS FOR SUPERVISED HEART BEAT CLASSIFICATION

Authors:

G. de Lannoy, J. Delbeke and M. Verleysen

Abstract: A reliable diagnosis of cardiac diseases can sometimes only be obtained by observing the heart of a patient for a long time period where every single heart beat is of importance. Computer-aided classification of heart beats is therefore of great help. The classification of the complete heart beat has many advantages compared to a classification of the QRS complex only or feature extraction methods. Nevertheless, the task is challenging because of the time-varying property of the heart beats. In this work, four time-alignment methods are evaluated and compared in the context of supervised heart beat classification. Among the four methods are three time series resampling methods by linear interpolation, cubic splines interpolation and trace segmentation. The fourth method is a realignment algorithm by dynamic time warping. The multiple sources of artifacts are filtered by discrete wavelet transform. As it only relies on a dissimilarity measure, the $k-$nearest neighbor classifier is a suitable choice for supervised classification of time series like ECG signals in multiple classes. Two different experiments corresponding to inter-patient and intra-patient classification are conducted on representative dataset built from the standard public MIT-BIH arrhythmia database.

Paper Nr: 35
Title:

EXTRACTION OF FETAL ECG FROM ABDOMINAL SIGNAL

Authors:

D. V. Prasad and R. Swarnalatha

Abstract: Fetal ECG monitoring is essential for identification of fetal distress. The assessment of the QRS waveform of fetal ECG is good analysis tool. Extraction of fetal ECG from abdominal signals is difficult. This paper presents a method for extracting fetal ECG (FECG) from composite abdominal signal. This method is applied to composite abdominal signal containing maternal ECG and fetal ECG. Adaptive filtering techniques along with denoising techniques were used to extract fECG. This method leads to enhancement of fetal ECG by cancelling maternal ECG. The results were validated using real signals. The thoracic signal is purely that of mother (mECG) while the abdominal signal contains both mothers and fetus ECG signals (mECG + fECG). The results clearly show the effectiveness of the method in extracting fECG.

Paper Nr: 36
Title:

A NOVEL SIMILARITY METRIC FOR RETINAL IMAGES BASED AUTHENTICATION

Authors:

C. Mariño, M. Ortega, M. G. Penedo and M. J. Carreira

Abstract: In biometrics the identity of an individual is verified using some physiologic or behavioural feature. In a typical authentication process involving some biometric trait, the biometric pattern for the user is extracted (a set of feature landmarks, a characteristic vector etc...). A similarity score is calculated between these patterns to determine if they belong to the same individual or not. This work presents an analysis of similarity metrics for an authentication system in which retinal vessel feature points are used as biometric pattern. The VARIA database of retinal images is used. A new metric is defined weighting the matched points information with the previously defined metrics. The obtained results show a large stretchment of the confidence gap between the matching scores of patterns from the same individual and the matching scores of patterns from different ones.

Paper Nr: 38
Title:

PERSPECTIVES OF ELECTROCOAGULATION IN WATER DISINFECTION

Authors:

A. Darchen, C. Ricordel and D. Hadjiev

Abstract: Today effective electrochemical disinfection system seems to be an alternative to conventional chlorination, ozone or UV processes. This paper reports on a series of experiments evaluating the disinfection efficiency of electrocoagulation cell using Al electrodes. Solutions contaminated by E. coli and surface waters were used as test media. The process significantly outperformed chemical coagulation reducing the amount of aluminum used. In all cases, the treated solutions were effectively disinfected and almost total removal of coliforms and algae were rapidly performed. A relationship for the disinfection rate of E. coli cells has been proposed. This equation takes into account the electrophoretic migration of the cells and the subsequent release of discharged cells into the bulk.

Paper Nr: 41
Title:

EFFICIENT SOURCE CODING IN A THRESHOLDING-BASED ECG COMPRESSOR USING THE DISCRETE WAVELET TRANSFORM

Authors:

Carlos Hernando Ramiro, Eduardo Moreno-Martínez, Fernando Cruz Roldán, José Sáez Landete and Manuel Blanco Velasco

Abstract: The aim of electrocardiogram (ECG) compression is to achieve as much compression as possible while the significant information for diagnosis purposes is preserved in the reconstructed signal. The source coding stage allows us to modify the compression ratio without quality degradation through a lossless encoder. In this work, the performance of this stage is analyzed in a compression scheme that has already presented good results among those from the state of the art. The compressor is based on discrete wavelet transform, thresholding and two-role encoder. The study consists of fixing all the stages except the source coding one in order to obtain an upper compression ratio bound. The assessment is based on the entropy of the independent symbols and the minimum expected length of the codewords. The results reveal a gap to improve the compression ratio, so from the previous entropy study an alternative compression method is proposed. For this purpose the symbols probabilities are analyzed through the normalized histogram. Thus, a Huffman encoder instead of the two-role one is applied in the new compressor to attain the maximum compression ratio. In this way a significant improvement is obtained without decreasing the original retrieved quality.

Paper Nr: 43
Title:

RECORDING EEG DURING REPETITIVE TRANS-CRANIAL MAGNETIC STIMULATION

Authors:

Gary Hasey, Hubert de Bruin, Mark Archambeault and Trung Trinh

Abstract: This paper discusses several issues related to recording EEG during repetitive trans-cranial magnetic stimulation (rTMS). The objective of recording EEG is to obtain magnetically evoked and event related potentials. The issue of electrode heating is discussed and experimental results presented that show graphite as well as fully notched or “C” silver, gold or silver-silver chloride are suitable for current rTMS protocols. Standard silver or gold cup electrodes may cause excessive scalp heating. Removal or reduction of the magnetically induced stimulus artefact is also discussed. A new system is presented that uses sample and hold circuitry to block most of the artefact allowing the researcher to record ipsi- and contra-lateral evoked potentials occurring within the first few milliseconds of the magnetic stimulus.

Paper Nr: 47
Title:

TOWARDS UNIFIED ANALYSIS OF EEG AND FMRI - A Comparison of Classifiers for Single-trial Pattern Recognition

Authors:

Johan Wessberg, Malin Åberg, Simon Bergstrand and Timo Niiniskorpi

Abstract: Pattern recognition methods, which recently have shown promising potential in the analysis of neurophysiological data, are typically model-free and can thus be applied in the analysis of any type of signal. This study demonstrates the feasibility of, after suitable pre-processing steps, applying identical state-of-the-art pattern recognition method to single-trial classification of brain state data acquired with the fundamentally different techniques EEG and fMRI.We investigated linear and non-linear support vector machines (SVM) and artificial neural networks (ANNs), and it was found that the SVM is highly suitable for the classification of both fMRI and EEG single patterns. However, the non-linear classifiers performed better than the linear ones on the EEG data (linear ANN: 66.2%, SVM: 78.9% vs. non-linear ANN: 71.8%, SVM: 83.2%), whereas the opposite was true for the fMRI dataset (linear ANN: 74.4%, SVM: 77.2% vs. non-linear ANN: 70.5%, SVM: 74.2%). The exciting possibility of concurrent EEG and fMRI registration warrants a need for a unified analysis method for both modalities, and we propose pattern recognition for this purpose. The ability to identify cortical patterns on a single-trial basis allows for brain computer interfaces, lie detection, bio-feedback, the tracking of mental states over time, and in the design of interactive, dynamic fMRI and EEG studies.

Paper Nr: 48
Title:

PARTICLE SWARM FEATURE SELECTION FOR FMRI PATTERN CLASSIFICATION

Authors:

Johan Wessberg, Malin Björnsdotter Åberg and Timo Niiniskorpi

Abstract: The application of pattern recognition to functional magnetic resonance imaging (fMRI) data enables exiting possibilities, including mind-reading and brain-machine interfacing. This paper presents a novel brain state identification approach, which, using an algorithm based on particle swarm optimization (PSO) in conjunction with a classifier of choice, identifies important brain voxels – thus both maximizing the classification performance and identifying physiologically relevant areas of the brain. For classifiers, we have investigated simple multiple linear regression (MLR) with thresholding and linear support vector machines (SVMs). Applying the PSO algorithm to single-subject, 2D data from a pleasant touch study, originally containing 5650 voxels, voxel subsets of mean size 64.8 and 132.6 voxels with classification accuracies of 73.1% and 77.0%, respectively for MLR and SMVs, was obtained. Similarly, on group level 3D data from a fingertapping study, with a total volume of 61078 voxels, a classification score of 83.5% was achieved on 89 voxels using the linear regression approach. For both datasets, the identified voxels agreed well with both general linear model T-maps and physiologically expected regions of activation. The PSO is thus effective in the identification of high-performing voxel subsets for fMRI volume classification, and also provides physiological information about brain processing related to the experimental conditions. Moreover, the PSO is a user-friendly algorithm, requiring little input from the user in terms of parameter specification.

Paper Nr: 59
Title:

INTRA-PATIENT REGISTRATION METHODS FOR THORACIC CT EXAMS

Authors:

João Cancela, José Silvestre Silva and Luísa Teixeira

Abstract: Now-a-days CT scanners provide detailed morphological information of pulmonary structures, with great importance to the diagnostic and follow-up of oncological diseases. When a patient with lung cancer is submitted to several CT exams during a period of time; these exams need an appropriate registration to quantify or visualize the tumour’s evolution. We propose a new method for 3D intra-patient registration of thoracic CT exams and compare its results with several 3D registration methods. The performance of these registration methods is analysed, computing several normalized figures of merit; we also explore these metrics to check which is more sensible to changes in CT exams due to the presence of lung tumours. The results with several cases of intra-patient, intra-modality registration show that the proposed method provides an accurate registration which is needed for the quantitative tracking of lesions that may effectively assist the follow-up process of oncological patients.

Paper Nr: 70
Title:

HOW MUCH BOVINE RHODOPSIN CRYSTAL STRUCTURE IS USEFUL FOR MODELING HUMAN GPCRS? - β2-Adrenergic Receptor as a Test Case

Authors:

Anwar Rayan, Jamal Raiyn and Mohamed Hegaze

Abstract: Availability of realistic models for human G-Protein Coupled Receptors (hGPCRs) will aid structure-based drug design (SBDD), thus shortening the time period needed for drug development and minimizing cross-reactivity of drugs with other hGPCRs. Many researchers have constructed models for hGPCRs with homology modeling techniques based on the X-ray structure of bovine rhodopsin and recently to β2-adrenergic receptor which are the only two GPCRs that have high resolution crystal structures. In this study, we evaluate the usefulness of the bovine rhodopsin crystal structures for modeling hGPCRs by analysis of large database of human G-protein coupled receptors that are members of family A (rhodopsin family). The recently released structure of β2-adrenergic receptor was used as a test case for validation purposes of our findings. From pair-wise sequence alignment of each of the receptors in the database to bovine rhodopsin, we come to the conclusion that only for few hGPCRs, X-ray structure of rhodopsin could be used as a template for modeling the trans-membrane domains (TMDs).The detailed analysis of the whole database shows that in general, similarity to bovine rhodopsin is found more in the middle/endoplasmic part than in the exoplasmic part. The shift in the cytoplasmic end of TMD-6 that has been seen in the recently released crystal structure of β2-adrenergic receptor could be understood well from our bioinformatics study. On the basis of our results from this research, we propose to regard specific parts from the endoplasmic domain of the rhodopsin helices as appropriate template for constructing models of other GPCRs, while most of the exoplasmic parts of GPCRs in this family require other techniques for their modeling, due to the low sequence similarity between the receptors and rhodopsin in that region.

Paper Nr: 74
Title:

DYNAMIC MULTIMEDIA CONTENT DELIVERY BASED ON REAL-TIME USER EMOTIONS - Multichannel Online Biosignals Towards Adaptative GUI and Content Delivery

Authors:

Eugénio Oliveira, Luís Paulo Reis and Vasco Vinhas

Abstract: Recently topics such as affective computing and multichannel multimedia distribution have gained the attention and investment of both industry and academics. The proposed system joins these domains so that ubiquitous system can be potentiated by means of online user emotion assessment based on real-time user’s biosignals. It was used IAPS as a emotional library for controlled visual stimuli and biosignals were collected in real-time - heartbeat rate and skin conductance - in order to online assess the user's emotional state through Russell’s Circumplex Model of Affect. To improve usability and session setup, a distributed architecture was used so that software models might be physically detached. The conducted experimental sessions and the validation interviews supported the system's efficiency not only in real-time discrete emotional state assessment but also considering the emotion inducing process. The future work consists in replicating the success in multi-format multimedia contents without pre-defined emotional metadata.

Paper Nr: 76
Title:

AUTOMATED EMG-SIGNAL PATTERN CLUSTERING BASED ON ICA DECOMPOSITION

Authors:

Abbas K. Abbas and Rasha Bassam

Abstract: Adaptive independent component analysis is interactive method for processing and classifying EMG signals pattern through short steps of ICA algorithms. In this work the efficiency and presentation of EMG signal decomposition and classification with adaptive ICA algorithm was investigated and presented. Single and multiple fibers motor unit action potentials (MUAP) patterns were tested and identified. Applying a fixed point modified ICA method, instead of much decomposition and pattern clustering algorithm localization of the action-potential source in the vicinity of specific neuromuscular diseases was achieved. ICA has its flex-ibility for robustly classify and identify the MUAP’s signal stochastic sources with a linear way and localizing the blind source for bioelectric potential. The utilization of adaptive ICA as an embedded clustering algorithm for separating a blind signal source will assist in construction an automated EMG signal diagnosis system with aid of new computerized real time signal processing technique. From the proposed system a stable and robust EMG classifying system based on multiple MUAP’s intensity were developed and tested through a standardization of clinical EMG signal acquisition and processing.

Paper Nr: 83
Title:

USEFULNESS OF BRAIN SIGNALS FOR THE DETECTION OF LOSS OF CONSCIOUSNESS IN ANESTHESIA - Overview of the Problem and Results from a Clinical Study

Authors:

Carmen González, Catarina S. Nunes, Joaquim Mendes, Marina Mendes and Pedro Amorim

Abstract: Loss of consciousness (LOC) detection is essential for better anesthesia guidance. Clinical signs and brain monitoring are currently used in operating rooms to assess the state of consciousness. However, a patient-independent, accurate and reliable indicator of LOC is not currently available. We studied 69 patients undergoing general anesthesia, investigating a possible relationship between loss of consciousness and BIS and EMG signals registered during induction. Neither BIS and EMG values at LOC, nor their abrupt fall proved to be good indicators of loss of consciousness. Further work needs to be done in order to reliably detect loss of consciousness.

Paper Nr: 89
Title:

UNIFIED ICA-SPM ANALYSIS OF FMRI EXPERIMENTS - Implementation of an ICA Graphical User Interface for the SPM Pipeline

Authors:

Carsten Haagen Nielsen, Jonas Henriksen, Kristoffer Hougaard Madsen, Lars Kai Hansen, Peter Mondrup Rasmussen and Troels Bjerre

Abstract: We present a toolbox for exploratory analysis of functional magnetic resonance imaging (fMRI) data using independent component analysis (ICA) within the widely used SPM analysis pipeline. The toolbox enables dimensional reduction using principal component analysis, ICA using several different ICA algorithms, selection of the number of components using the Bayesian information criterion (BIC), visualization of ICA components, and extraction of components for subsequent analysis using the standard general linear model. We demonstrate how the toolbox is capable of identifying activity and nuisance effects in fMRI data from a visual experiment.

Paper Nr: 93
Title:

ECG CLASSIFICATION AND ANALYSIS IN A ZIGBEE WIRELESS SENSOR NETWORK

Authors:

Cosmin Rotariu, Des Chambers and Enda Barrett

Abstract: Wireless technology has become ubiquitous in our daily lives. From 802.11 to Bluetooth we have become familiar with new technologies and expectations are rife as to its potential. The medical world is potentially lucrative for the use of such technology. The ability to improve patient comfort, monitor patients remotely and increase device mobility should all contribute handsomely to patient life quality. It also offers the unique opportunity to monitor ambulatory patients in a real-time environment. Outlined is an approach to integrate an Electrocardiogram (ECG) classifier into an overall wireless patient monitoring system enabling real-time classification and analysis of ECG data. Our research has shown that it is possible to use the open source classifier (Hamilton, 2002) in a wireless sensor network for beat detection and arrhythmia classification. We have tested the classifier with up to 80 simulated sensors proving that its lightweight implementation enables it to cope perfectly with only minor modifications needed. It was found that the addition of multiples of sensors produced on average 0.01% performance degradation.

Paper Nr: 94
Title:

IN-SITU, REAL-TIME BIOREACTOR MONITORING BY FIBER OPTICS SENSORS

Authors:

A. A. Vicente, J. A. Teixeira, J. S. Silva, R. G. Silva and R. C. Martins

Abstract: One of the most studied bioprocesses is fermentation by yeasts. Although this is true, there is still the lack of real-time instrumentation that is capable of providing detailed information on metabolic state of fermentations. In this research we explore the possibility of using UV-VIS-SWNIR spectroscopy as a high-output, non-destructive and multivariate methodology of monitoring beer fermentation. We herein report the implementation of the a fibber optics sensor and the capacity for detecting key parameters by partial least squares regression for biomass, extract, pH and total sugars. Results show that UV-VIS-SWNIR is a robust technique for monitoring beer fermentations, being able to provide detailed information spectroscopic fingerprinting of the process. Calibrations were possible to obtain for all the studied parameters with R2 of 0.85 to 0.94 in the UV-VIS region and 0.95 to 0.97 in the VIS-SWNIR region. This preliminary study allowed to conclude that further improvements in experimental methodology and signal processing may turn this technique into a valuable instrument for detailed metabolic studies in biotechnology.

Paper Nr: 97
Title:

IEETA BRAIN COMPUTER INTERFACE - Towards a Rapid Prototyping and Multi-Application System

Authors:

Filipe M. Silva, João P. Silva Cunha and Virgílio F. Bento

Abstract: Recent advances in computer hardware and signal processing assert that controlling certain functions by thoughts may represent a landmark in the way we interact with many output devices. This paper exploits the possibility of achieving a communication channel between the brain and a mobile robot through the modulation of the electroencephalogram (EEG) signal during motor imagery tasks. A major concern was directed towards designing a generalized and multi-purpose framework that supports rapid prototyping of various experimental strategies and operating modes. Preliminary results of brain-state estimation using EEG signals recorded during a self-paced left/right hand movement task are also presented. The user successfully learned to operate the system and how to better perform the motor-related tasks based on outcomes produced by its mental focus.

Paper Nr: 99
Title:

ON THE INFLUENCE OF LOW FREQUENCY COMPONENTS IN THE WEIGHT BEHAVIOUR OF THE LMS ALGORITHM

Authors:

A. K. Barros, D. S. Brito, E. Aguiar, F. Lucena and R. C. S. Freire

Abstract: The Least Mean Square (LMS) algorithm is a very important tool in the estimation and filtering of biomedical signals. Amongst these signals are the periodic and quasiperiodic. For example, the LMS algorithm was used to estimate the coefficients of the Fourier series at a given frequency or even in a spectral analysis. In this paper we study the behavior of the weights of the LMS algorithm when the signal to be estimated acts at very low frequencies. We prove theoretically that lower frequency noise affects the estimation of the weights at higher frequencies. We carried out simulations and showed that experimental findings are in agreement with the theoretical results. Moreover, we exemplify the problem with electrocardiogram signals (ECG).

Paper Nr: 105
Title:

A NEW ACCURATE METHOD OF HARMONIC-TO-NOISE RATIO EXTRACTION

Authors:

Ricardo J. T. de Sousa

Abstract: In this paper, an accurate method that estimates the HNR from sustained vowels based on harmonic structure modeling is proposed. Basically, the proposed algorithm creates an accurate harmonic structure where each harmonic is parameterized by frequency, magnitude and phase. The harmonic structure is then synthesized and assumed as the harmonic component of the speech signal. The noise component can be estimated by subtracting the harmonic component from the speech signal. The proposed algorithm was compared to others HNR extraction algorithms based on spectral, cepstral and time domain methods, and using different performance measures.

Paper Nr: 107
Title:

TOWARDS AN IMMUNE-INSPIRED TEMPORAL ANOMALY DETECTION ALGORITHM BASED ON TUNABLE ACTIVATION THRESHOLDS

Authors:

Jorge Carneiro, Manuel Correia and Mário Antunes

Abstract: The detection of anomalies in computer environments, like network intrusion detection, computer virus or spam classification, is usually based on some form of pattern search on a database of “signatures” for known anomalies. Although very successful and widely deployed, these approaches are only able to cope with anomalous events that have already been seen. To cope with these weaknesses, the “behaviour” based systems has been deployed. Although conceptually more appealing, they have still an impractical high rate of false alarms. The vertebrate Immune System is an emergent and appealing metaphor for new ideas on anomaly detection, being already adopted some algorithms and theoretical theories in particular fields, such as network intrusion detection. In this paper we present a temporal anomaly detection architecture based on the Grossman’s Tunable Activation Threshold (TAT) hypothesis. The basic idea is that the repertoire of immune cells is constantly tuned according to the cells temporal interactions with the environment and yet retains responsiveness to an open-ended set of abnormal events. We describe some preliminary work on the development of an anomaly detection algorithm derived from TAT and present the results obtained thus far using some synthetic data-sets.

Paper Nr: 108
Title:

INSECT NAVIGATION BY POLARIZED LIGHT

Authors:

F. J. Smith

Abstract: Many insects can navigate accurately using the polarized light from the sky. A study of a large number of experiments on the behaviour and anatomy of insects has led to a simple algorithm for navigation by skylight, suitable for a robot or drone in lightly clouded skies The algorithm is based on the special ability of insect eyes to measure the position of the 4 points in the sky at which the polarization angle, i.e. the angle χ between the polarized E-vector and the meridian, equals ±π/4. The azimuths of these 4 points are almost invariant to variable cloud cover, provided that polarized light is still detectable below the clouds. It is shown that the sum of these 4 azimuths can be turned into a celestial compass in a few short steps and a simulation shows that the compass is accurate as well as simple and well suited for a robot or drone. It can also explain many of the experimental results published on insect navigation.

Paper Nr: 112
Title:

DISCRIMINATION OF HEART SOUNDS USING CHOAS ANALYSIS IN VARIOUS SUBBANDS

Authors:

A. Sá e Melo, D. Kumar, J. Henriques, J. Habetha, M. Antunes and P. Carvalho

Abstract: Discrimination among different types of heart sounds has a significant impact in designing pHealth systems based upon this bio-signal, since (i) it enables the optimal selection and tuning of the analysis algorithms and (ii) it may be applied as a first level strategy for heart dysfunction diagnosis. In this paper we introduce an algorithm for heart sound type discrimination into three classes: healthy heart sounds, heart sounds with murmur produced by native heart valves and heart sounds produced by prosthetic mechanical heart valves. The algorithm is based on a nonlinear dynamical model of phase space reconstruction for various frequency bands. For each frequency sub-band the chaotic nature and the complexity of the signal is assessed using the largest Lyapunov exponents (LLE) and the correlation dimension (CD). The effectiveness of the method has been tested with heart sounds of 45 subjects (15 subjects of each class). It was concluded that LLEs and the CDs exhibit complementary significance in the discrimination among different classes of heart sounds.

Paper Nr: 114
Title:

EEG-BASED SPEECH RECOGNITION - Impact of Temporal Effects

Authors:

Anne Porbadnigk, Jan Calliess, Marek Wester and Tanja Schultz

Abstract: In this paper, we investigate the use of electroencephalograhic signals for the purpose of recognizing unspoken speech. The term unspoken speech refers to the process in which a subject imagines speaking a given word without moving any articulatory muscle or producing any audible sound. Early work by Wester (Wester, 2006) presented results which were initially interpreted to be related to brain activity patterns due to the imagination of pronouncing words. However, subsequent investigations lead to the hypothesis that the good recognition performance might instead have resulted from temporal correlated artifacts in the brainwaves since the words were presented in blocks. In order to further investigate this hypothesis, we run a study with 21 subjects, recording 16 EEG channels using a 128 cap montage. The vocabulary consists of 5 words, each of which is repeated 20 times during a recording session in order to train our HMM-based classifier. The words are presented in blockwise, sequential, and random order. We show that the block mode yields an average recognition rate of 45.50%, but it drops to chance level for all other modes. Our experiments suggest that temporal correlated artifacts were recognized instead of words in block recordings and back the above-mentioned hypothesis.

Paper Nr: 116
Title:

A BIOLOGICALLY INSPIRED HARDWARE MODULE FOR EMBEDDED AUDITORY SIGNAL PROCESSING APPLICATIONS

Authors:

Chris Melhuish, Mokhtar Nibouche, Tony Pipe and Xin Yang

Abstract: This paper presents a fully parameterised and highly scalable design prototype of FPGA (field programmable gate array) implementation of a biologically inspired auditory signal processing system. The system has been captured and simulated using system-level integrated design tools, namely, System GeneratorTM and AccelDSPTM both from XilinxTM . The implemented hardware auditory periphery model consists of two sub-models—the Patterson’s Gammatone filter bank and the Meddis’ inner hair cell. The prototype has been successfully ported onto a VirtexTM –II Pro FPGA. Ultimately, it can be used as a front-end apparatus in a variety of embedded auditory signal processing applications.

Paper Nr: 118
Title:

CLASSIFYING AYURVEDIC PULSE SIGNALS VIA CONSENSUS LOCALLY LINEAR EMBEDDING

Authors:

Amod Jog, Anant Madabhushi, Aniruddha Joshi and Sharat Chandran

Abstract: In this paper, we present a novel method for analysis of Ayurvedic pulse signals via a recently developed nonlinear dimensionality reduction scheme called Consensus Locally Linear Embedding (C-LLE). Pulse Based Diagnosis (PBD) is a prominent method of disease detection in Ayurveda, the system of Indian traditional medicine. Ample anecdotal evidence suggests that for several conditions, PBD, based on sensing changes in the patient’s pulse waveform, is superior to conventional allopathic diagnostic methods. PBD is an inexpensive, non-invasive, and painless method; however, a lack of quantification and standardization in Ayurveda, and a paucity of expert practitioners, has limited its widespread use. The goal of this work is to develop the first Computer-Aided Diagnosis (CAD) system able to distinguish between normal and diseased patients based on their PBD. Such a system would be inexpensive, reproducible, and facilitate the spread of Ayurvedic methods. Digitized Ayurvedic pulse signals are acquired from patients using a specialized pulse waveform recording device. In our experiments we considered a total of 50 patients. The 50 patients comprised of two cohorts obtained at different frequencies. The first cohort comprised 24 patients that were normal or diseased (slipped disc (backache), stomach ailments) while the second consists of a set of 26 patients who were normal or diseased (diabetic, with skin disorders, slipped disc (backache) and stress related headaches). In this study, we consider the C-LLE scheme which non-linearly projects the high-dimensional Ayurvedic pulse data into a lower dimensional space where a consensus clustering scheme is employed to distinguish normal and abnormal waveforms. C-LLE differs from other linear and nonlinear dimensionality reduction schemes in that it respects the underlying nonlinear manifold structure on which the data lies and attempts to directly estimate the pairwise object adjacencies in the lower dimensional embedding space. A major contribution of this work is that it employs non-Euclidean similarity measures such as mutual information and relative entropy to estimate object similarity in the high-dimensional space which are more appropriate for measuring the similarity of the pulse signals. Our C-LLE based CAD scheme results in a classification accuracy of 80.57% using relative entropy as the signal distance measure in distinguishing between normal and diseased patients for the first cohort, based on their Ayurvedic pulse signal. For the 500Hz data we got a maximum of 88.34% accuracy with C-LLE and relative entropy as a distance measure. Furthermore, C-LLE was found to outperform LLE, Isomap, PCA across multiple distance measures for both cohorts.

Paper Nr: 123
Title:

CLASSIFICATION OF MASS SPECTROMETRY DATA - Using Manifold and Supervised Distance Metric Learning

Authors:

Andrew H. Sung, Bernardete M. Ribeiro, Mengyu Qiao and Qingzhong Liu

Abstract: Mass spectrometry becomes the most widely used measurement in proteomics research. The quality of the feature set and applied learning classifier determine the reliability of the prediction of disease status. A well-known approach is to combine peak detection and support vector machine recursive feature elimination (SVMRFE). To compare the feature selection and to search for alternative learning classifier, in this paper, we employ a distance metric learning to classification of proteomics mass spectrometry (MS) data. Experimental results show that distance metric learning is promising for the classification of proteomics data; the results are comparable to the best results by applying SVM to the SVMRFE feature sets. Results also indicate that the good potential of manifold learning for feature reduction in MS data analysis.

Paper Nr: 132
Title:

A DYNAMICAL MODEL FOR PULSATILE FLOW ESTIMATION IN A LEFT VENTRICULAR ASSIST DEVICE

Authors:

Abdul-Hakeem H. AlOmari, Andrey V. Savkin, Dean Karantonis, Einly Lim and Nigel H. Lovell

Abstract: In this paper, we propose a dynamical model for pulsatile flow estimation of an iRBP. Noninvasive measurements of the motor power (VI) and pump impeller rotational speed (w) were acquired from the pump controller and used together with blood hematocrit (HCT) values as inputs to the model. A circulatory mock loop was operated with different aqueous glycerol solutions, mimicking different values of viscosities equivalent to the range of 20 - 50% of human blood HCT, to generate pulsatile flow data. Linear regression between estimated pulsatile flow (Qest ) and measured flow (Qmeas) obtained from the mock loop resulted in a highly significant correlation (R2 = 0.957) and mean absolute error of e = 0.364 L/min. Also, R2 = 0.902 and e = 0.317 L/min were obtained when our model was validated using six sets of ex vivo porcine data. Furthermore, in steady state, the solution of the presented model can be described by a previously designed and verified static model.

Paper Nr: 133
Title:

NON-INVASIVE SEPSIS PATIENT CLASSIFICATION USING LEAST SQUARES SUPPORT VECTOR MACHINE

Authors:

Andrey V. Savkin, Collin H. H. Tang and Paul M. Middleton

Abstract: Sepsis is a systemic inflammatory response to serious infection. Without proper identification and treatment at its early stage, this syndrome can deteriorate within hours to a more devastating state. In this paper, it was hypothesized that early identification of sepsis stages can be achieved through the evaluation of patients’ autonomic neural activity by means of power spectral analysis. Least squares support vector machine (LSSVM) was utilized to classify sepsis patients into systemic inflammatory response syndrome (SIRS) and severe sepsis groups, based on the measured normalized low-frequency (LFn) components of heard period (RRi) and pulse transit time (PTT) time series. Polar-like transformation of LFn pair of RRi and PTT provides another two distinctive features into the construction of input space. Age factor was also used as an attribute in sepsis classification. The performance of the proposed LSSVM with two different kernels: cubic-polynomial and Gaussian radial basis function (RBF), was evaluated using 5-fold cross-validation technique. From the study, LSSVM with RBF kernel was found to be an effective classifier in the identification of sepsis syndrome progression, with the classification accuracy, sensitivity, and specificity: 93.32%, 99.87%, and 79.29% respectively.

Paper Nr: 134
Title:

THE POWER SPECTRA RESPONSE OF STROKE VOLUME AND ARTERIAL BLOOD PRESSURE VARIABILITY SIGNALS TO AUTONOMIC NERVOUS SYSTEM MODULATION OF THE HEART

Authors:

Abdul-Hakeem H. AlOmari, Andis Graudins and Andrey V. Savkin

Abstract: This study presents results that describe the short term oscillations in SBPV and SVV signals due to calcium channel blockers poisoning with verapamil treated with continuous infusion of levosimendan. In addition, we used average spectra of these oscillations to observe the activity and sympathovagal balance of the autonomic nervous system. Then, we compared the average spectra obtained from both signals. The frequency contents of the average spectra of SVV and SBPV signals to levosimendan treatment of verapamil-poisoned rats were analyzed and related to the activity of sympathetic and parasympathetic tones. In control group, average spectra of SVV and SBPV exhibited a low-frequency band (LF: 0.03−0.8 Hz) peaked at ~ 0.4 Hz and a highfrequency (HF: 0.8−3.0 Hz) peaked at ~ 1 Hz. LF peak was abolished after verapamil infusion. The LF component of both spectra was observed to recover after continuous infusion of levosimendan. Additionally, a new frequency component was observed at 1.5 Hz in the spectrum of SBPV. Significant correlations were found between bands of the average spectra in both signals in all groups of treatment studied in this paper. Our results revealed that, like SBPV, SVV can herald useful information regarding the sympathovagal balance and cardiac output improvements.

Paper Nr: 136
Title:

DOUBLE PULSE TRANSMISSION - DEAD ZONE DECREASING IN ULTRASOUND IMAGING

Authors:

Andrzej Nowicki, Ihor Trots and Marcin Lewandowski

Abstract: This study investigates a new composing method of double transmission of short coded sequences based on well-known Golay complementary codes, which allows to obtain the higher signal-to-noise ratio (SNR) and decrease dead zone area. The proposed method can potentially find application in small parts ultrasonography and play an important role in examination of superficial structures, e.g. in dermatology, ophthalmology, etc., where using longer coded sequences leads to increase of a dead zone and single pulse transmission of short sequences does not assure sufficient SNR. This paper discusses the results obtained during the examination of four different length pairs of Golay coded sequences excited at 3.7 MHz: the single 64-bits pair of Golay sequences and combined sequences consisting of two 8, 16, and 32-bits Golay codes separated in time. The experimental results have shown that double pulse transmission allows to suppress considerably the noise level, the SNR increases by 5.7 dB in comparison with the single pulse transmission of Golay sequences of the same length. The presented results of this work demonstrate the advantage of double pulse transmission method which enhances SNR while maintaining the dead zone short.

Paper Nr: 138
Title:

EFFECT OF SURFACE ELECTRODE ORIENTATION ON INDEPENDENT COMPONENT ANALYSIS FOR FEATURE EXTRACTION OF SURFACE MOTOR UNIT ACTION POTENTIAL

Authors:

Jun Akazawa, Kotaro Minato, Masaki Yoshida, Ryuhei Okuno, Takaharu Ikeuchi, Takemasa Okamoto and Tetsuo Sato

Abstract: Recently, application of Independent Component Analysis (ICA) has been reported for effective decomposi-tion of surface electromyogram (SEMG) signals into a train of surface motor unit action potentials (SMUAPs) of a single motor unit (MU). Results of ICA were not always sufficient as the feature extraction of SMUAP at first dorsal interosseous muscle (FDI). The purpose of this study is to propose an effective method for feature extraction of SMUAP by simulation study of focusing on the effects of electrode orientation. SEMG signals were created with the model and application of ICA was applied to the signals. The present study showed that the useful and actual method of ICA application was to repeat measurement of SEMG signals with varying the electrode orientation, and then to select the better signals for the feature extraction by executing ICA algorithm.

Paper Nr: 141
Title:

PREREQUISITES FOR AFFECTIVE SIGNAL PROCESSING (ASP)

Authors:

Egon L. van den Broek, Jennifer A. Healey, Joris H. Janssen and Joyce H. D. M. Westerink

Abstract: Although emotions are embraced by science, their recognition has not reached a satisfying level. Through a concise overview of affect, its signals, features, and classification methods, we provide understanding for the problems encountered. Next, we identify the prerequisites for successful Affective Signal Processing: validation (e.g., mapping of constructs on signals), triangulation, a physiology-driven approach, and contributions of the signal processing community. Using these directives, a critical analysis of a real-world case is provided. This illustrates that the prerequisites can become a valuable guide for Affective Signal Processing.

Paper Nr: 143
Title:

BRAIN ACTIVITY DETECTION - Statistical Analysis of fMRI Data

Authors:

Alicia Quirós Carretero and Raquel Montes Diez

Abstract: We are concerned with modelling and analysing fMRI data. An fMRI experiment is a series of images obtained over time under two different conditions, in which regions of activity are detected by observing differences in blood magnetism due to hemodynamic response. In this paper we propose a spatiotemporal model for detecting brain activity in fMRI. The model makes no assumptions about the shape or form of activated areas, except that they emit higher signals in response to a stimulus than non-activated areas do, and that they form connected regions. The Bayesian spatial prior distributions provide a framework for detecting active regions much as a neurologist might; based on posterior evidence over a wide range of spatial scales, simultaneously considering the level of the voxel magnitudes together with the size of the activated area. A single spatiotemporal Bayesian model allows more information to be obtained about the corresponding magnetic resonance study. Despite higher computational cost, a spatiotemporal model improves the inference ability since it takes into account the uncertainty in both the spatial and the temporal dimension. A simulated study is used to test the model applicability and sensitivity.

Paper Nr: 144
Title:

DATA REDUCTION OR DATA FUSION IN BISOGINAL PROCESSING?

Authors:

David Sommer, Martin Golz and Udo Trustschel

Abstract: When subjects are monitored over long time spans and when several biosignals are derived a large amount of data has to be processed. In consequence, the number of features which has to be extracted is mostly very restricted in order to avoid the so-called “curse of high dimensionality”. Donoho (Donoho, 2000) stated that this applies only if algorithms perform local in order to search systematically for general discriminant functions in a high-dimensional space. If they take into account a concept for regularization between locality and globality “blessings of high dimensionality” are to be expected. The aim of the present study is to examine this on a particular real world data set. Different biosignals were recorded during simulated overnight driving in order to detect driver’s microsleep events (MSE). It is investigated if data fusion of different signals reduces dete¬ction errors or if data reduction is beneficial. This was realized for nine electroencephalography, two electro¬oculography, and for six eyetracking signals. Features were extracted of all signals and were processed dur¬ing a training process by computational intelligence methods in order to find a discriminant function which separates MSE and Non-MSE. The true detection error of MSE was estimated based on cross-validation. Results indicate that fusion of all signals and all features is most beneficial. Feature reduction was of limited success and was slightly beneficial if Power Spectral Densities were averaged in many narrow spectral bands. In conclusion, the processing of several biosignals and the fusion of many features by computational intelligence methods has the potential to establish a reference standard (gold standard) for the detection of extreme fatigue and of dangerous microsleep events which is needed for upcoming Fatigue Monitoring Technologies.

Posters
Paper Nr: 8
Title:

HYBRID PARAMETERIZATION SYSTEM FOR WRITER IDENTIFICATION

Authors:

Carlos F. Romero, Carlos M. Travieso, Jesús B. Alonso and Miguel A. Ferrer

Abstract: In this paper, we present a hybrid parameterization system from classical and graphologist features, as the existing percentage of cohesion in the writing of each individual, as well as the smaller and greater axes of the ovals and loops. They have been used on the writer identification together with other parameters applied to handwritten words. That set of characteristics has been tested with our off-line database, which consists of 70 writers with 10 samples per writer and as well each sample is composed of 34 words. We have got a success rate of 96%, applying as classifier Neural Network, and after, the technique of “more voted” algorithm, with 10 Neural Networks.

Paper Nr: 14
Title:

HOW MUCH SEQUENCE IDENTITY GUARANTEE GOOD MODELS IN HOMOLOGY MODELING - Proteins from Serine Protease Family as a Test Case?

Authors:

Anwar Rayan and Jamal Raiyn

Abstract: Homology modelling is utilized to predict the 3-D structure of a given protein (target) based on its sequence alignment to a protein whose structure (template) has been experimentally determined. The use of such technique is already rewarding and increasingly widespread in biological research and drug development. The accuracy of the predictions as commonly accepted is dependent on the score of target protein - template sequence identity. Given the sequence identity score of pairs of proteins, certain questions are raised as to whether we can assess or quantitate the quality of the obtained model. Also, whether we should choose, the protein with the highest sequence identity as a template. The answer to these questions is critical since only with such determinations, we could decide how to choose the template and to which usage the model is reliable. We intend in the paper to assess the accuracy of sequence identity-based homology modeling by analyzing a database of 4560 pair-wise sequence and structural alignments. The decision making process regarding to which parts of the known protein to perform structural alignment is not trivial and clearer rules should be extracted.

Paper Nr: 15
Title:

AN EFFICIENT AND EFFECTIVE IMAGE SEGMENTATION INTERACTIVE TOOL

Authors:

Dorit S. Hochbaum

Abstract: This paper describes a utilization of a very efficient polynomial time algorithm, discovered by Hochbaum (2001), for segmentation tool through Markov Random Fields. The tool allows flexible choice of input parameters, controlling the output within an interactive tool with dynamic features and easily modified parameters.

Paper Nr: 18
Title:

AUTOMATIC RECOGNITION OF LEAVES BY SHAPE DETECTION PRE-PROCESSING WITH ICA

Authors:

Carlos M. Travieso, Jesús B. Alonso, Jordi Solé-Casals, Juan Carlos Briceño and Miguel A. Ferrer

Abstract: In this work we present a simulation of a recognition process with perimeter characterization of a simple plant leaves as a unique discriminating parameter. Data coding allowing for independence of leaves size and orientation may penalize performance recognition for some varieties. Border description sequences are then used to characterize the leaves. Independent Component Analysis (ICA) is then applied in order to study which is the best number of components to be considered for the classification task, implemented by means of an Artificial Neural Network (ANN). Obtained results with ICA as a pre-processing tool are satisfactory, and compared with some references our system improves the recognition success up to 80.8% depending on the number of considered independent components.

Paper Nr: 20
Title:

COHERENCY AND SHARPNESS MEASURES BY USING ICA ALGORITHMS - An Investigation for Alzheimer's Disease Discrimination

Authors:

Andrzej Cichocki, François Vialatte, Jordi Solé-Casals and Zhe Chen

Abstract: In this paper, we present a comprehensive study of different Independent Component Analysis (ICA) algorithms for the calculation of coherency and sharpness of electroencephalogram (EEG) signals, in order to investigate the possibility of early detection of Alzheimer's disease (AD). We found that ICA algorithms can help in the artifact rejection and noise reduction, improving the discriminative property of features in high frequency bands (specially in high alpha and beta ranges). In addition to different ICA algorithms, the optimum number of selected components is investigated, in order to help decision processes for future works.

Paper Nr: 45
Title:

LOW-COST ADAPTIVE METHOD FOR REAL-TIME ECG BASELINE WANDER REMOVAL WITH REDUCED P AND T WAVE DISTORTION

Authors:

Juan Ignacio Godino-Llorente and Rubén Fraile

Abstract: An adaptive algorithm for the removal of baseline wander in ECG is presented. The scheme is based on a single-tap LMS filter that estimates the baseline signal. The baseline is further processed by a moving average filter. This way, reduced distortion of P-Q and S-T segments is achieved with a low computational cost. Moreover, the proposed system has a short impulse response that makes it appropriate for real-time applications.

Paper Nr: 57
Title:

PULMONARY PRESSURE-VOLUME CURVES OF ELASTASE-TREATED AND CONTROL RATS

Authors:

Angelos Kyriazis, Germán Peces-Barba, Ignacio Rodríguez, Jesús Ruiz-Cabello, Laura del Puerto Nevado and Sandra Pérez-Rial

Abstract: The objective of this experiment is to test if the emphysematous and the control rats can be classified according to the pulmonary pressure-volume curves. Emphysema was simulated by instilling elastase to the rat lungs and leaving them to develop the disease for 6 weeks. The pressure-volume curves were acquired by a customdesigned ventilator. The pressure at the inflection point of the inhalation limb of the curves has been used as a feature to separate the two classes of animals. The extension of emphysema in the rat lungs was assessed post-mortem by measuring the mean surface of the alveoli. This was possible after extracting the lungs, slicing them, photographing them and analysing the pictures. The mean surface of the alveoli distinguishes clearly the control from the emphysematous rats, which verifies the deteriorating effect of elastase over the lung alveoli. However, no clear correlation was found between the pressure-volume curves inflection-points and the animal classes.

Paper Nr: 69
Title:

AUTOMATIC EMOTION INDUCTION AND ASSESSMENT FRAMEWORK - Enhancing User Interfaces by Interperting Users Multimodal Biosignals

Authors:

Eugénio Oliveira, Jorge Teixeira, Luís Paulo Reis and Vasco Vinhas

Abstract: Emotion’s definition, identification, systematic induction and efficient and reliable classification have been themes to which several complementary knowledge areas such as psychology, medicine and computer science have been dedicating serious investments. This project consists in developing an automatic tool for emotion assessment based on a dynamic biometric data acquisition set as galvanic skin response and electroencephalography are practical examples. The output of standard emotional induction methods is the support for classification based on data analysis and processing. The conducted experimental sessions, alongside with the developed support tools, allowed the extraction on conclusions such as the capability of effectively performing automatic classification of the subject’s predominant emotional state. Self assessment interviews validated the developed tool's success rate of approximately 75%. It was also experimentally strongly suggested that female subjects are emotionally more active and easily induced than males.

Paper Nr: 71
Title:

OPTIMIZATION OF EMG-SIGNAL SOURCE CLASSIFICATION BASED ON ADAPTIVE WAVELETS K-MEAN ALGORITHM

Authors:

Abbas K. Abbas, Rana M. Kasim and Rasha Bassam

Abstract: In this paper the optimization of EMG signals segmentation and decomposition based on wavelet represen-tation and k-mean clustering technique is presented. It is shown that wavelet decomposition can be usefull in detecting particular spikes in EMG signals and the presented segmentation algorithm may be useful for the detection of active segments in related MUAP’s action potentials. The algorithms has been tested on the synthetic model signal and on real signals recorded with intramuscular multi-point electrode. The efficiency of EMG signal decomposition and classification with adaptive wavelet algorithm were presented. Single and multiple fibers MUAP patterns were tested and identified. By applying a Debauchies wavelet transformation and k-mean clustering algorithm to localize the action-potential source in the presence of specific neuromuscular diseases like NMI neuropathy, muscular dystrophy and myasthenia gravis (MG), instead of many decomposition and pattern recognition algorithm, wavelets and k-mean clustering have its flexibility for robustly classify and localize the signal stochastic sources with a linear way, in addition to identify the blind source for EMG bioelectric potential.

Paper Nr: 82
Title:

ANALYSIS OF THE HEART RATE VARIABILITY BEFORE AND AFTER ASPHYXIA

Authors:

Federico Cardona Rocha and Fernando S. Schlindwein

Abstract: Over the last two decades there has been a widespread interest in the study of variations in the beat-to-beat timing of the heart, known as heart rate variability (HRV). The studies of heart rate variability have allowed access to useful information about disturbances in autonomic regulation, which are a promising marker to quantify autonomic activity. Heart rate variability has become the conventionally accepted term to describe variations of both instantaneous heart rate and RR intervals (the RR interval is the time interval between two consecutive R-points of the QRS complex) (D. Bajic et al., 2006). The objective of this paper is the analysis and comparison of the HRV before and after asphyxia using data from previous studies where 24 adult Wistar rats were anesthetised and subjected to controlled asphyxia for specified durations (Boardman et al 2002). Preliminary results of our work show a depression of this parameter after long periods of asphyxia, indicating that HRV might be a good marker for assessing injury to the autonomic nervous system due to asphyxia.

Paper Nr: 91
Title:

BIOLOGICAL-VISION INSPIRED DSA SYSTEM FOR UAVS

Authors:

Hyuck M. Kwon, Jie Yang, Larry Paarmann and Wenhao Xiong

Abstract: Uses of unmanned Aerial Vehicles (UAVs) have increased dramatically during the past several years, but currently still do not have convenient access to civil airspace because there is no onboard pilot, and it’s impossible for UAVs to “see and avoid” other aircraft. So a Detect, Sense and Avoid system is needed to provide the UAV with instructions to steer the UAV clear of any potential collision with other traffic. An optical based DSA system is discussed in this paper to provide the UAV with “see and avoid” capability of at least equivalent to a piloted aircraft. DSA minimum detection range assessment, optical resolution requirement and image array size requirement are also discussed in this paper. Also an efficient natural vision system is presented in this paper for DSA system.

Paper Nr: 98
Title:

A PRELIMINARY STUDY ON THE DETECTION OF TRANSCRIPTION FACTOR BINDING SITES

Authors:

Alexandre Perera, Erola Pairo and Santiago Marco

Abstract: Transcription starts when multiple proteins, known as transcription factors recognize and bind to transcription start site in DNA sequences. Since mutation in transcription factor binding sites are known to underlie diseases it remains a major challenge to identify these binding sites. Conversion from symbolic DNA to numerical sequences and genome data make it possible to construct a detector based on a numerical analysis of DNA binding sites. A subspace model for the TFBS is built. TFBS will show a very small distance to this particular subspace. Using this distance binding sites are distinguished from random sequences and from genome data.

Paper Nr: 102
Title:

RESPIRATORY SINUS ARRHYTHMIA IN 10 YEAR OLDS - Normal and Intrauterine Growth Restricted

Authors:

Fernando Schlindwein, Michael Wailoo, Michael Bankart and Taher Biala

Abstract: Frequency domain analysis of RR has been determined by three methods, autoregressive model (AR), Fast Fourier Transform (FFT) and Lomb periodogram for 10 min segments. The first two methods were done after resampling and the third method without resampling RR series of all 75 children. AR was used in this work, and RSA was identified at night time during sleep. The area of the RSA was calculated for every 10 min interval and compared to the overall area of the 10 min segment, then the average RSA of all segments was calculated, as well as the overall percentage of the RSA energy to the total area for the whole period of sleeping. This was done firstly for a sample of Normal and IUGR 10 year olds. Secondly for all the children under study, an independent t-test concluded that there is no significant difference between the IUGR and Normal (p=0.7467).

Paper Nr: 110
Title:

ARTIFICIAL NEURAL NETWORK APPROACH FOR OBESITY-HYPERTENSION CLASSIFICATION

Authors:

Gabriela Postolache, Joaquim Mendes, Octavian Postolache and Pedro Silva Girão

Abstract: One of the newest targets of public health is management of obesity-hypertension. In this paper is presented the use of an artificial neural network based model for objective classification of obesity-hypertension. Different neural network architectures as part of hybrid processing scheme including comparators and competitive processing blocks were developed and tested. The neural network functionality is the classification of the individuals according to the obesity risks. The results show that the neural network classifier is consistent with the standard criteria suggested by the obesity and hypertension guidelines.

Paper Nr: 119
Title:

COMPETITIVE AND COOPERATIVE SOLUTIONS FOR REMOTE EYE-TRACKING

Authors:

Francesco La Rosa, Giancarllo Iannizzotto and Giovanni Crisafulli

Abstract: Reliable detection and tracking of eyes is an important requirement for attentive user interfaces. In this paper, we present an innovative approach to the problem of the eye-tracking. Traditional eye-detectors, chosen for own properties, are combined by two different schemes (competitive and cooperative scheme) to improve own robustness and reliability. To illustrate our work, we introduce a proof-of-concept single camera remote eye-tracker and discuss its implementation and the obtained experimental results.

Paper Nr: 125
Title:

WAVELET BASED EXTRACTION OF BLOOD VESSELS

Authors:

Ali Hojjat and Hammad Omer

Abstract: An algorithm for the segmentation of blood vessels based on the correlation of different wavelet scales is presented. First the wavelet coefficients are computed for a defined number of scales and then the correlation between the corresponding coefficients of two consecutive scales is computed. The normalized product is used as a reference threshold for retaining original wavelet coefficients. If the normalized product is greater than the corresponding original wavelet coefficient, the original coefficient is retained for image reconstruction by inverse wavelet transform, otherwise the coefficient is changed with zero value. Low frequency wavelet coefficients matrix is not used in image reconstruction process as we want only the edge information. The proposed algorithm is quite general and can be used for the extraction of any type of blood vessels and provides very promising results.

Paper Nr: 135
Title:

EXERCISE RATE ESTIMATION USING A TRIAXIAL ACCELEROMETER

Authors:

Andrey V. Savkin, Branko G. Celler, Ning Wang, Steven W. Su and Teddy M. Cheng

Abstract: In this paper, we propose an algorithm for the estimation of exercise rate during a variety of exercises by using measurements from triaxial accelerometry. The algorithm involves the detection of the periodicity of the body’s accelerations, and the detected periods are then fused to form an estimate of exercise rate. Experimental results demonstrate that the algorithm is effective in different modes of exercise. The proposed algorithm will be useful in monitoring training exercises for healthy individuals and rehabilitation exercises for cardiac patients.

Paper Nr: 137
Title:

THE ICA APPROACH FOR REMOVAL OF UNDESIRED COMPONENTS FROM EEG DATA

Authors:

Joanna Górecka

Abstract: The aim of this results of research is to detect and remove selected undesired signals by means of ICA approach. In this paper have been presented the following algorithms BSS: HJ, Infomax and FastICA for separation and removal of selected group of artifacts (eye blinks, muscle activity) from EEG recordings. As it has been proven in experiments, the proposed algorithms can effectively detect and remove these artifacts from EEG recordings.