Nicos Maglaveras

ECG pattern recognition and classification using non-linear transformations and neural networks: A review

The most widely used signal in clinical practice is the ECG. ECG conveys information regarding the electrical function of the heart, by altering the shape of its constituent waves, namely the P, QRS, and T waves. Thus, the required tasks of ECG processing are the reliable recognition of these waves, and the accurate measurement of clinically important parameters measured from the temporal distribution of the ECG constituent waves. In this paper, we shall review some current trends on ECG pattern recognition. In particular, we shall review non-linear transformations of the ECG, the use of principal component analysis (linear and non-linear), ways to map the transformed data into n-dimensional spaces, and the use of neural networks (NN) based techniques for ECG pattern recognition and classification. The problems we shall deal with are the QRS/PVC recognition and classification, the recognition of ischemic beats and episodes, and the detection of atrial fibrillation. Finally, a generalised approach to the classification problems in n-dimensional spaces will be presented using among others NN, radial basis function networks (RBFN) and non-linear principal component analysis (NLPCA) techniques. The performance measures of the sensitivity and specificity of these algorithms will also be presented using as training and testing data sets from the MIT-BIH and the European ST-T databases.

An adaptive backpropagation neural network for real-time ischemia episodes detection: development and performance analysis using the European ST-T database

A supervised neural network (NN)-based algorithm was used for automated detection of ischemic episodes resulting from ST segment elevation or depression. The performance of the method was measured using the European ST-T database. In particular, the performance was measured in terms of beat-by-beat ischemia detection and in terms of the detection of ischemic episodes. The algorithm used to train the NN was an adaptive backpropagation (BP) algorithm. This algorithm drastically reduces training time (tenfold decrease in our case) when compared to the classical BP algorithm. The recall phase of the NN is then extremely fast, a fact that makes it appropriate for real-time detection of ischemic episodes. The resulting NN is capable of detecting ischemia independent of the lead used. It was found that the average ischemia episode detection sensitivity is 88.62% while the ischemia duration sensitivity is 72.22%. The results show that NN can be used in electrocardiogram (EGG) processing in cases where fast and reliable detection of ischemic episodes is desired as in the case of critical care units (CCUs).

Model-based morphological segmentation and labeling of coronary angiograms

A method for extraction and labeling of the coronary arterial tree (CAT) using minimal user supervision in single-view angiograms is proposed. The CAT structural description (skeleton and borders) is produced, along with quantitative information for the artery dimensions and assignment of coded labels, based on a given coronary artery model represented by a graph. The stages of the method are: (1) CAT tracking and detection; (2) artery skeleton and border estimation; (3) feature graph creation; and (iv) artery labeling by graph matching. The approximate CAT centerline and borders are extracted by recursive tracking based on circular template analysis. The accurate skeleton and borders of each CAT segment are computed, based on morphological homotopy modification and watershed transform. The approximate centerline and borders are used for constructing the artery segment enclosing area (ASEA), where the defined skeleton and border curves are considered as markers. Using the marked ASEA, an artery gradient image is constructed where all the ASEA pixels (except the skeleton ones) are assigned the gradient magnitude of the original image. The artery gradient image markers are imposed as its unique regional minima by the homotopy modification method, the watershed transform is used for extracting the artery segment borders, and the feature graph is updated. Finally, given the created feature graph and the known model graph, a graph matching algorithm assigns the appropriate labels to the extracted CAT using weighted maximal cliques on the association graph corresponding to the two given graphs. Experimental results using clinical digitized coronary angiograms are presented.

A multiagent system enhancing home-care health services for chronic disease management

In this paper, a multiagent system (MAS) is presented, aiming to enhance monitoring, surveillance, and educational services of a generic medical contact center (MCC) for chronic disease management. In such a home-care scenario, a persistent need arises for efficiently monitoring the patient contacts and the MCCs functionality, in order to effectively manage and interpret the large volume of medical data collected during the patient sessions with the system, and to assess the use of MCC resources. Software agents were adopted to provide the means to accomplish such real-time information-processing tasks, due to their autonomous, reactive and/or proactive nature, and their effectiveness in dynamic environments by incorporating coordination strategies. Specifically, the objective of the MAS is to monitor the MCC environment, detect important cases, and inform the healthcare and administrative personnel via alert messages, notifications, recommendations, and reports, prompting them for actions. The main aim of this paper is to present the overall design and implementation of a proposed MAS, emphasizing its functional model and architecture, as well as on the agent interactions and the knowledge-sharing mechanism incorporated, in the context of a generic MCC.

The citizen health system (CHS): a Modular medical contact center providing quality telemedicine services

In the context of the Citizen Health System (CHS) project, a modular Medical Contact Center (MCC) was developed, which can be used in the monitoring, treatment, and management of chronically ill patients at home, such as diabetic or congestive heart failure patients. The virtue of the CHS contact center is that, using any type of communication and telematics technology, it is able to provide timely and preventive prompting to the patients, thus, achieving better disease management. In this paper, we present the structure of the CHS system, describing the modules that enable its flexible and extensible architecture. It is shown, through specific examples, how quality of healthcare delivery can be increased by using such a system.

Human epigenome data reveal increased CpG methylation in alternatively spliced sites and putative exonic splicing enhancers.

The role of gene body methylation, which represents a major part of methylation in DNA, remains mostly unknown. Evidence based on the CpG distribution associates its presence with nucleosome positioning and alternative splicing. Recently, it was also shown that cytosine methylation influences splicing. However, to date, there is no methylation-based data on the association of methylation with alternative splicing and the distribution in exonic splicing enhancers (ESEs). We presently report that, based on the computational analysis of the Human Epigenome Project data, CpG hypermethylation (>80%) is frequent in alternatively spliced sites (particularly in noncanonical) but not in alternate promoters. The methylation frequency increases in sequences containing multiple putative ESEs. However, significant differences in the extent of methylation are observed among different ESEs. Specifically, moderate levels of methylation, ranging from 20% to 80%, are frequent in SRp55-binding elements, which are associated with response to extracellular conditions, but not in SF2/ASF, primarily responsible for alternative splicing, or in CpG islands. Finally, methylation is more frequent in the presence of AT repeats and CpGs separated by 10 nucleotides and lower in adjacent CpGs, probably indicating its dependence on helical formations and on the presence of nucleosome positioning-related sequences. In conclusion, our results show the regulation of methylation in ESEs and support its involvement in alternative splicing.

A Personalized Framework for Medication Treatment Management in Chronic Care

The ongoing efforts toward continuity of care and the recent advances in information and communication technologies have led to a number of successful personal health systems for the management of chronic care. These systems are mostly focused on monitoring efficiently the patients medical status at home. This paper aims at extending home care services delivery by introducing a novel framework for monitoring the patients condition and safety with respect to the medication treatment administered. For this purpose, considering a body area network (BAN) with advanced sensors and a mobile base unit as the central communication hub from the one side, and the clinical environment from the other side, an architecture was developed, offering monitoring patterns definition for the detection of possible adverse drug events and the assessment of medication response, supported by mechanisms enabling bidirectional communication between the BAN and the clinical site. Particular emphasis was given on communication and information flow aspects that have been addressed by defining/adopting appropriate formal information structures as well as the service-oriented architecture paradigm. The proposed framework is illustrated via an application scenario concerning hypertension management.

A Pervasive Health System Integrating Patient Monitoring, Status Logging, and Social Sharing

In this paper, we present the design and development of a pervasive health system enabling self-management of chronic patients during their everyday activities. The proposed system integrates patient health monitoring, status logging for capturing various problems or symptoms met, and social sharing of the recorded information within the patients community, aiming to facilitate disease management. A prototype is implemented on a mobile device illustrating the feasibility and applicability of the presented work by adopting unobtrusive vital signs monitoring through a wearable multisensing device, a service-oriented architecture for handling communication issues, and popular microblogging services. Furthermore, a study has been conducted with 16 hypertensive patients, in order to investigate the user acceptance, the usefulness, and the virtue of the proposed system. The results show that the system is welcome by the chronic patients who are especially willing to share healthcare information, and is easy to learn and use, while its features have been overall regarded by the patients as helpful for their disease management and treatment.

Effectiveness of home-centered care through telemedicine applications for overweight and obese patients: a randomized controlled trial.

OBJECTIVE: To determine if home-centered monitoring through telemedicine has an impact on clinical characteristics, metabolic profile and quality of life in overweight and obese patients.DESIGN: Randomized controlled trial, 6-month duration.SETTING: Tertiary care academic hospital.SUBJECTS: A total of 122 patients were eligible to participate as they met the inclusion criteria of increased body mass index (BMI>25 kg/m2), age>18 and <70 y and ability to operate electronic microdevices.INTERVENTIONS: All patients in the control group (n=77) received standard hospital care. Patients in the intervention group (n=45), additionally, measured three times a week, for 6 months, their blood pressure and body weight and transmitted them to an automated call center. These values were not shared with the patients’ physician or dietician.MAIN OUTCOME MEASURES: Clinical (body weight, BMI, blood pressure), laboratory (fasting plasma glucose, triglycerides, HDL-cholesterol, total cholesterol) and quality of life parameters (SF-36®, Visual Analog Scale of European Quality-5 Dimensions, Obesity Assessment Survey). Data were analyzed in an intention-to-treat-way (last observation carried forward).RESULTS: Drop-out rate was similar in the control and intervention groups: 12 vs 11 percent, respectively, P=NS. There were no significant differences at baseline between intervention and control groups in all main outcome parameters. There were significant decreases for patients in the intervention group in body weight (from 101.6±22.4 to 89.2±14.7 kg, P=0.002, P=0.05 vs controls at 6 months), total cholesterol (from 247.6±42.0 to 220.7±42.6 mg/dl, P=0.002, P=0.05 vs controls at 6 months) and triglycerides (from 148.4±35.0 to 122.3±31.4 mg/dl, P=0.001, P=0.01 vs controls at 6 months). Intervention group patients made a total of 1997 phone contacts. The number of phone contacts was correlated positively with Social Functioning (SF), Vitality (VT) and Mental Health (MH) scores of SF-36® at baseline (r=0.48, r=0.41, r=0.41, respectively, P=0.05) but not with weight loss.CONCLUSIONS: Home-centered, intense treatment through the use of telemedicine can be effective in improving short-term obesity outcomes.

Machine Learning and Data Mining Methods in Diabetes Research

The remarkable advances in biotechnology and health sciences have led to a significant production of data, such as high throughput genetic data and clinical information, generated from large Electronic Health Records (EHRs). To this end, application of machine learning and data mining methods in biosciences is presently, more than ever before, vital and indispensable in efforts to transform intelligently all available information into valuable knowledge. Diabetes mellitus (DM) is defined as a group of metabolic disorders exerting significant pressure on human health worldwide. Extensive research in all aspects of diabetes (diagnosis, etiopathophysiology, therapy, etc.) has led to the generation of huge amounts of data. The aim of the present study is to conduct a systematic review of the applications of machine learning, data mining techniques and tools in the field of diabetes research with respect to a) Prediction and Diagnosis, b) Diabetic Complications, c) Genetic Background and Environment, and e) Health Care and Management with the first category appearing to be the most popular. A wide range of machine learning algorithms were employed. In general, 85% of those used were characterized by supervised learning approaches and 15% by unsupervised ones, and more specifically, association rules. Support vector machines (SVM) arise as the most successful and widely used algorithm. Concerning the type of data, clinical datasets were mainly used. The title applications in the selected articles project the usefulness of extracting valuable knowledge leading to new hypotheses targeting deeper understanding and further investigation in DM.