Yorgos Goletsis

Automated ischemic beat classification using genetic algorithms and multicriteria decision analysis

Cardiac beat classification is a key process in the detection of myocardial ischemic episodes in the electrocardiographic signal. In the present study, we propose a multicriteria sorting method for classifying the cardiac beats as ischemic or not. Through a supervised learning procedure, each beat is compared to preclassified category prototypes under five criteria. These criteria refer to ST segment changes, T wave alterations, and the patients age. The difficulty in applying the above criteria is the determination of the required method parameters, namely the thresholds and weight values. To overcome this problem, we employed a genetic algorithm, which, after proper training, automatically calculates the optimum values for the above parameters. A task-specific cardiac beat database was developed for training and testing the proposed method using data from the European Society of Cardiology ST-T database. Various experimental tests were carried out in order to adjust each module of the classification system. The obtained performance was 91% in terms of both sensitivity and specificity and compares favorably to other beat classification approaches proposed in the literature.

A novel method for automated EMG decomposition and MUAP classification

OBJECTIVE This paper proposes a novel method for the extraction and classification of individual motor unit action potentials (MUAPs) from intramuscular electromyographic signals. METHODOLOGY The proposed method automatically detects the number of template MUAP clusters and classifies them into normal, neuropathic or myopathic. It consists of three steps: (i) preprocessing of electromyogram (EMG) recordings, (ii) MUAP detection and clustering and (iii) MUAP classification. RESULTS The approach has been validated using a dataset of EMG recordings and an annotated collection of MUAPs. The correct identification rate for MUAP clustering is 93, 95 and 92% for normal, myopathic and neuropathic, respectively. Ninety-one percent of the superimposed MUAPs were correctly identified. The obtained accuracy for MUAP classification is about 86%. CONCLUSION The proposed method, apart from efficient EMG decomposition addresses automatic MUAP classification to neuropathic, myopathic or normal classes directly from raw EMG signals.

Real-Time Driver's Stress Event Detection

In this paper, a real-time methodology for the detection of stress events while driving is presented. The detection is based on the use of physiological signals, i.e., electrocardiogram, electrodermal activity, and respiration, as well as past observations of driving behavior. Features are calculated over windows of specific length and are introduced in a Bayesian network to detect drivers stress events. The accuracy of the stress event detection based only on physiological features, evaluated on a data set obtained in real driving conditions, resulted in an accuracy of 82%. Enhancement of the stress event detection model with the incorporation of driving event information has reduced false positives, yielding an increased accuracy of 96%. Furthermore, our methodology demonstrates good adaptability due to the application of online learning of the model parameters.

A multiscale and multiparametric approach for modeling the progression of oral cancer

BackgroundIn this work, we propose a multilevel and multiparametric approach in order to model the growth and progression of oral squamous cell carcinoma (OSCC) after remission. OSCC constitutes the major neoplasm of the head and neck region, exhibiting a quite aggressive nature, often leading to unfavorable prognosis.MethodsWe formulate a Decision Support System assembling a multitude of heterogeneous data sources (clinical, imaging tissue and blood genomic), aiming to capture all manifestations of the disease. Our primary aim is to identify the factors that dictate OSCC progression and subsequently predict potential relapses of the disease. The discrimination potential of each source of data is initially explored separately, and afterwards the individual predictions are combined to yield a consensus decision achieving complete discrimination between patients with and without a disease relapse. Moreover, we collect and analyze gene expression data from circulating blood cells throughout the follow-up period in consecutive time-slices, in order to model the temporal dimension of the disease. For this purpose a Dynamic Bayesian Network (DBN) is employed which is able to capture in a transparent manner the underlying mechanism dictating the disease evolvement, and employ it for monitoring the status and prognosis of the patients after remission.ResultsBy feeding as input to the DBN data from the baseline visit we achieve accuracy of 86%, which is further improved to complete discrimination when data from the first follow-up visit are also employed.ConclusionsKnowing in advance the progression of the disease, i.e. identifying groups of patients with higher/lower risk of reoccurrence, we are able to determine the subsequent treatment protocol in a more personalized manner.

A multilevel and multistage efficiency evaluation of innovation systems

A framework for estimating national and regional innovation efficiency is presentedThe proposed DEA-based model is formulated as a multiobjective mathematical programMultiple objectives refer to different stages and hierarchies of innovation systemsOrdinal regression analysis examines the influence of additional variablesEfficiency results show significant differences across countries and regions Evaluating the efficiency of innovation systems can serve as a substantial enabling tool for policymaking serving to identify best practices and develop potential improvements of actions and strategies. It also serves to provide valuable insight in understanding the nature and dynamics of innovation process at its different stages and levels. The main aim of the paper is to present an integrated assessment and classification framework for national and regional innovation efficiency. The proposed model is based on Data Envelopment Analysis and is formulated as a multiobjective mathematical program in order to consider the objectives and constraints of the different stages and levels of the innovation process. Additionally, the model copes with DEA inconsistencies when ratio measures are employed. In the second part of the study, a multicriteria decision aid approach, based on an ordinal regression model, is applied in order to study how environmental factors on innovation and entrepreneurship affect the estimated efficiency scores. The proposed approach is applied to a set of 23 European countries and their 185 corresponding regions. The results show that there are large differences regarding the efficiency scores of the different stages and levels, implying the existence of significant divergences from the expected norm concerning innovation efficiency. The contribution of the paper lies (i) in the proposed multiobjective model, which is able to model the multiple stages and levels of the innovation process and handle ratio measures without requiring the same set of inputs and outputs at different levels and (ii) in the presented application of the model in the efficiency evaluation of innovation systems, including a meta-analysis of the results based on the framework of the Quadruple Innovation Helix. Such an approach may provide a valuable tool for country/region comparison and policy formulation.

A method for arrhythmic episode classification in ECGs using fuzzy logic and Markov models

A method for arrhythmic episode classification using only the RR-interval signal is presented. The method is based on fuzzy logic and Markov models, while classification is performed for nine categories of cardiac rhythms. A two-stage classifier is applied. In the first stage, a fuzzy system classifies the episode using the mean value and standard deviation of the RR-intervals. In the second, the RR-interval signal is transformed to character sequences, which are classified by Markov models. Two representation techniques are used for the extraction of the character sequences: symbolic dynamics and one based on the RR-interval length. The classification of an episode is achieved combining the outcomes of the two stages. The MIT-BIH arrhythmia database is used for the evaluation of the proposed method. The obtained results indicate high performance (accuracy 73%) in arrhythmic episode classification.

Multiparametric Decision Support System for the Prediction of Oral Cancer Reoccurrence

Oral squamous cell carcinoma (OSCC) constitutes the predominant neoplasm of the head and neck region, featuring particularly aggressive nature, associated with quite unfavorable prognosis. In this paper, we formulate a decision support system that integrates a multitude of heterogeneous data (clinical, imaging, and genomic), thus, framing all manifestations of the disease. Our primary aim is to identify the factors that dictate OSCC progression and subsequently predict potential relapses (local or metastatic) of the disease. The discrimination potential of each source of data is initially explored separately, and afterward the individual predictions are combined to yield a consensus decision achieving complete discrimination between patients with and without a disease relapse.

A multicriteria decision based approach for ischaemia detection in long duration ECGs

A new method based on a multicriteria classification algorithm, which is applied on data extracted from the electrocardiographic signal and the patients history, is described for the detection of ischaemic episodes in long duration electrocardiographic recordings. The classification procedure is embedded into a four-stage system of automated ischaemia diagnosis. Using a task-specific cardiac beat database the multicriteria method was properly adjusted to classify each presented cardiac beat as ischaemic or normal. The method was tested using the cardiac beat database and we achieved good results, equal or even better than other reported methods, demonstrating the efficiency of the proposed methodology.

Multi-level multi-stage efficiency measurement: the case of innovation systems

Efficiency measurement has been receiving significant attention the last years especially after the recent economic crises and the need of efficient use of public money. Although single step efficiency measurement is usually applied, taking into account the internal structure of the system is expected to provide more meaningful and informative results. This could be done in two axes: decomposition of the process into sub-processes and hierarchical modeling among system components. In this framework we extend the Data Envelopment Analysis approach by examining efficiency through multi-level and multi-stage modeling. The proposed modeling approach (1) can give a better insight in sub-processes compared to single-stage ones and (2) can take into account functional/systemic characteristics (e.g., a Decision Making Unit operating as a part of a greater system). Through a ‘soft’ integration approach, not only different stages can be introduced but also hierarchies can be easily accommodated. Analysis of the efficiency of national/regional innovation systems is used as an illustrative example. The innovation process is modeled as a multi-stage process including knowledge production and knowledge commercialization and multi-level one, where regional innovation is achieved within a national innovation system.

Towards building a Dynamic Bayesian Network for monitoring oral cancer progression using time-course gene expression data

In this work we present a methodology for modeling and monitoring the evolvement of oral cancer in remittent patients during the post-treatment follow-up period. Our primary aim is to calculate the probability that a patient will develop a relapse but also to identify the approximate time-frame that this relapse is prone to appear. To this end, we start off by analyzing a broad set of time-course gene expression data in order to identify a set of genes that are mostly differentially expressed between patients with and without relapse and are therefore discriminatory and indicative of a disease reoccurrence evolvement. Next, we employ the maintained genes coupled with a patient-specific risk indicator in order to build upon them a Dynamic Bayesian Network (DBN) able to stratify patients based on their probability for a disease reoccurrence, but also pinpoint an approximate time-frame that the relapse might appear.