Digital assessment of swollen joints from smartphone photographs can support remote monitoring of patients with inflammatory arthritis, enabling more efficient care delivery. In this study, we propose a method to classify each interphalangeal joint in hand photographs as swollen or not. The classifier is a logistic regression model that uses joint effective width as the main predictor, with covariates including age, sex, body mass index, and the joint type. Effective width measures the thickness of a joint and is derived using computer vision algorithms that segment individual fingers, detect joint landmarks, and calculate distances between points of interest. We validated the model using a dataset collected from 85 individuals with psoriatic arthritis recruited between December 2024 and April 2025 across three countries. For each participant, the dataset includes paired photographs of both hands, captured by the participants themselves, along with demographic and clinical data reported by healthcare professionals. Model validation was performed using k-fold cross-validation, ensuring each validation fold contained exactly one patient with at least one swollen joint. The method achieved an average area under the ROC curve (AUROC) of 0.68 (SD 0.29). These findings demonstrate the potential of smartphone-based joint modeling as a scalable and patient-friendly approach to enhance remote assessment and personalized management of psoriatic arthritis.

Wearable and smartphone-based emotion recognition (WER) remains a challenging setting in affective computing, due to the notorious difficulty and bias associated with in-thewild label collection. The high inter-and intra-subject emotional variability motivates us to explore WER modeling through graph node classification in a limited resources learning scheme powered by Self-Supervised Learning (SSL) graph masking augmentation tasks. We employ a subgraph sampling approach during training, utilizing labeled and unlabeled data, along with supervised, semi-supervised, and SSL mechanisms in a multi-task inductive graph neural network architecture. Our evaluations on K-EmoPhone through leave-one-group-out cross-validation in the binary arousal and valence tasks yield average accuracy gains of 4.3% and 7.8%, compared to the full resource setting, utilizing only 20% and 25% of the labels, respectively. Our model analysis sheds light on the relation of SSL graph augmentations to emotional arousal and valence and justifies the approach of SSL-driven subgraph training for in-the-wild WER.

This study addresses the growing need for noninvasive, secure, and efficient biometric identification methods in Internet of Things (IoT) applications, where traditional biometric systems often face challenges due to privacy concerns, environmental constraints, and practical limitations. To tackle these issues, we introduce a novel biometric identification system that leverages custom-built multiplexed Fiber Bragg Grating (FBG) accelerometers and bispectral feature extraction. By applying bispectral analysis to the acquired gait signals, we extract robust and discriminative features for individual identification. Unsupervised clustering algorithms, namely K-means and DBSCAN, were employed to categorize individuals based on these features, successfully identifying ten distinct clusters corresponding to ten participants and demonstrating the system's effectiveness. The K-means model achieved a Davies-Bouldin Index of 0.79 and a Silhouette Score of 0.45, while DBSCAN yielded a Davies-Bouldin Index of 0.87 and a Silhouette Score of 0.36. Given the proprietary nature of the data and the custom-built FBG accelerometers used in this study, direct comparisons to state-of-the-art methods are not available. However, these results underscore the potential of our unique approach and technology to advance biometric identification, providing a promising non-invasive, scalable, and discreet solution with broad applicability in IoT environments.

ABSTRACT The growing availability of conversational data across multiple platforms has intensified interest in dynamic emotion recognition. Speech plays a pivotal role in shaping the emotional climate (EC) of peer conversations. We propose DeepBispec, the first framework to integrate deep bispectral image analysis with affect dynamics (AD) for speech‐based EC recognition. Bispectrum representations capture nonlinear and non‐Gaussian speech characteristics, while AD descriptors model temporal emotion fluctuations. Evaluated on K‐EmoCon, IEMOCAP and SEWA datasets, DeepBispec consistently improved EC classification performance. For example, on K‐EmoCon, arousal accuracy increased from 79.0% (bispectrum only) to 81.4% (with AD), while valence accuracy improved from 76.8% to 77.5%; similar trends were observed for IEMOCAP and SEWA. DeepBispec outperformed strong CNN, LSTM, and Transformer baselines, demonstrating robust cross‐lingual performance across seven languages. These findings highlight its potential for real‐world applications such as mental health monitoring, affect‐aware learning platforms and empathetic dialogue systems.

Time-varying functional connectivity (TVFC) measured with functional MRI (fMRI) captures dynamic changes in statistical dependencies among regional time series, which can be studied with instantaneous phase synchronization analyses. Phase extraction requires narrow-banded resting-state fMRI (rs-fMRI) data typically extracted with conventional band-pass filtering or advanced mode decomposition techniques. However, filtering methods often struggle to eliminate noise effectively, require prior knowledge of cutoff frequencies, and fail to account for non-stationarity in the data. Likewise, existing mode decomposition techniques strongly depend on input parameters and are less reliable for multivariate analyses. Here, we introduce multivariate swarm decomposition (MSwD), a bio-inspired signal decomposition technique that combines the iterative nature of empirical methods with a robust mathematical foundation. Using synthetic signals and real rs-fMRI data from the Human Connectome Project and the Autism Brain Imaging Data Exchange I, we showed that MSwD-based PS (MSwD-PS) outperforms four state-of-the-art decomposition techniques in several key areas: (1) being more robust to input parameters and better at detecting true synchronizations, showing a 3%–65% lower normalized root mean square error in simulated data and being 15.8-73.1% less prone to identifying short biologically implausible transitions between brain states, and (2) showing a reduced likelihood of false positives, being less affected by spurious synchronizations. Likewise, MSwD-informed functional connectivity analysis improved subject fingerprinting and autism spectrum disorder classification using graph neural networks. Overall, MSwD-PS can reduce the risk of false positives in TVFC, which could be extremely useful for processing rs-fMRI data with unknown ground truth in diverse clinical populations.

Psoriatic Arthritis (PsA) is a chronic, inflammatory disease affecting joints, substantially impacting patients' quality of life, with European guidelines for managing PsA emphasizing the importance of assessing hand function. Here, we present a set of novel digital biomarkers (dBMs) derived from a touchscreen-based serious game approach, DaktylAct, intended as a proxy, gamified, objective assessment of hand impairment, with emphasis on fine motor skills, caused by PsA. This is achieved by its design, where the user controls a cannon to aim at and hit targets using two finger pinch-in/out and wrist rotation gestures. In-game metrics (targets hit and score) and statistical features (mean, standard deviation) of gameplay actions (duration of gestures, applied pressure, and wrist rotation angle) produced during gameplay serve as informative dBMs. DaktylAct was tested on a cohort comprising 16 clinically verified PsA patients and nine healthy controls (HC). Correlation analysis demonstrated a positive correlation between average pinch-in duration and disease activity (DA) and a negative correlation between standard deviation of applied pressure during wrist rotation and joint inflammation. Logistic regression models achieved 83% and 91% classification performance discriminating HC from PsA patients with low DA (LDA) and PsA patients with and without joint inflammation, respectively. Results presented here are promising and create a proof-of-concept, paving the way for further validation in larger cohorts.

Understanding human movement patterns and evaluating a range of medical disorders depend heavily on the analysis of gait. In this study, we propose a new method for gait analysis, based on multiplexed fiber Bragg grating (FBG) accelerometers. Our work expands the capabilities of FBG-based accelerometers by extracting gait features through the analysis of output signals. In contrast to traditional wearable sensors, our solution offers scalability and discreet monitoring while integrating smoothly into the current infrastructure. Step duration, cadence, peak acceleration, and gait symmetry are among the critical gait metrics that we calculate using MATLAB-based methods to preprocess the accelerometer data. Experiments show that our method is a good fit for precisely capturing gait dynamics. The study revealed significant differences among individuals (p<0.05) in gait parameters based on height and age groups, indicating variations in step time, and normalized cadence. Our findings have important ramifications for biometric identification, rehabilitation, and healthcare applications.

While machines struggle to cope with acoustical variability and noise, humans show remarkable robustness to recognize speech content under different conditions of environmental noise. The tonotopic organization of the spiral human cochlea has motivated the signal processing community for its superb frequency tuning capabilities. In this work, we design and evaluate a novel spiral cochlear cepstrum space, as a novel, directional feature engineering framework, using a cochlear transform approach, that results in tonotopically organized, orthogonal cochlear modes. Such cochlear modes are then transformed to the spiral cochlear cepstral space, yielding cochlear filterbank cepstral coefficients (CFCCs). As opposed to previous works that define the bio-inspired cepstral features based on Mel-, Equivalent Rectangular Bandwidth (ERB) or linear scales, we define the scaling based on the cochlear spiral geometry that spans from θ = 0° at the base to θ = 990° at the apex. We then compute the log function and the discrete cosine transform of the cochlear modes energy yielding spatially supported cepstral features along the spiral cochlear space, spaced by θ = 45°. We assess the impact of noise on the CFCCs and compare the performance to that of Mel-Frequency Cepstral Coefficients (MFCCs) and Gammatone Filterbank Cepstral Coefficients (GFCCs) using the NOIZEUS dataset. We report, for the first time, that the superiority of the CFCCs noise-robustness stems from the geometrical organization of the cochlea (i.e., its tonotopic map) when evaluated on speech signals contaminated with different noise conditions at different SNRs. The proposed CFCCs constitute a platform for a new class, bio-inspired and noise-robust feature extraction for many applications such as speaker recognition.

Signal decomposition techniques aim to break down non-stationary signals into their oscillatory components, serving as a preliminary step in various practical signal processing applications. This has motivated researchers to explore different strategies, yielding several distinct approaches. A well-known optimization-based method, the Variational Mode Decomposition (VMD), relies on the formulation of an optimization problem utilizing constant-bandwidth Wiener filters. However, this poses limitations in constant bandwidth and the need for constituent count. In this paper, the Dynamic Bandwidth VMD (DB-VMD) is proposed to generalize VMD by addressing the Wiener filter limitations through enhancement of the optimization problem with an additional constraint. Experiments in synthetic signals highlight DB-VMD’s noise robustness and adaptability in comparison to VMD, paving the way for many applications, especially when the analyzed signals are contaminated with noise.

One in every four newborns suffers from congenital heart disease (CHD) that causes defects in the heart structure. The current gold-standard assessment technique, echocardiography, causes delays in the diagnosis owing to the need for experts who vary markedly in their ability to detect and interpret pathological patterns. Moreover, echo is still causing cost difficulties for low- and middle-income countries. Here, we developed a deep learning-based attention transformer model to automate the detection of heart murmurs caused by CHD at an early stage of life using cost-effective and widely available phonocardiography (PCG). PCG recordings were obtained from 942 young patients at four major auscultation locations, including the aortic valve (AV), mitral valve (MV), pulmonary valve (PV), and tricuspid valve (TV), and they were annotated by experts as absent, present, or unknown murmurs. A transformation to wavelet features was performed to reduce the dimensionality before the deep learning stage for inferring the medical condition. The performance was validated through 10-fold cross-validation and yielded an average accuracy and sensitivity of 90.23 % and 72.41 %, respectively. The accuracy of discriminating between murmurs' absence and presence reached 76.10 % when evaluated on unseen data. The model had accuracies of 70 %, 88 %, and 86 % in predicting murmur presence in infants, children, and adolescents, respectively. The interpretation of the model revealed proper discrimination between the learned attributes, and AV channel was found important (score 0.75) for the murmur absence predictions while MV and TV were more important for murmur presence predictions. The findings potentiate deep learning as a powerful front-line tool for inferring CHD status in PCG recordings leveraging early detection of heart anomalies in young people. It is suggested as a tool that can be used independently from high-cost machinery or expert assessment.

Hypertension is a serious medical condition that affects over a billion people worldwide. The proper management of disease progression requires an extended knowledge of the overall functional and structural changes in the whole body in response to the hypertension. Here, we propose HyperScore, an integrative and unified measure of hypertension progression relative to multi-organ and multi-modality clinical measurements and based on a semi-supervised machine learning (ML) approach. We developed the measure based on a large participating cohort from the UK Biobank database (n=27,099) with over 500 imaging and clinical variables from multiple modalities. The semi-supervised approach was developed based on the contrastive trajectory inference mechanism to provide a score that reflects the proximity of a participant to the disease state (range: 0–1). Modelling revealed that majority of hypertensive participants had scores above 0.25, whereas normotensives had scores below this threshold. The sensitivity and specificity were above 89%, with an area under the receiver operating characteristics of 96.4%. The modelling showed a stable performance when evaluating hidden testing sets on a 10-fold cross-validation scheme with nearly 0.1 error. There was a strong association (r2>0.6) between HyperScore and organs’ phenotypic patterns, especially for variables such as white matter hyperintensity and body mass index. This study is the first to potentiate ML-based modelling of hypertension progression from a multi-organ perspective, which could significantly aid in clinical decision making to save lives.

Signal multiscale decomposition (SMD) is an effective analysis for the identification of modal information in time-domain signals. So far, various SMD approaches, such as the Multiresolution Wavelet Transform (MWT), the Empirical Mode Decomposition (EMD), and the Variational Mode Decomosition (VMD) have been proposed. However, issues, such as mode mixing for signals with closelyspaced modes, have been identified. To confront such problems, we propose here a novel spatial auditory decomposition framework for non-stationary signals, namely the Cochlear Decomposition (CD). CD is inspired by the biological rules of the spiral human cochlea and it is built upon the concept of ‘place-pitch’ or the tonotopic organization of the spiral cochlea. The new insight here is to formulate a set of basis functions, namely cochlear wavelets, whose dilation factors are determined by their angular position on the cochlear spiral (i.e., coupled rotation and dilation). Under proper parameterization, iterative application of spatial filters eventually results in signal’s mono-components, each arising from specific angular positions along the spiral cochlea, with high time and frequency localization. The performance of the proposed CD is validated via synthetic acoustic signals and real non-stationary speech signals analysis. The analysis results show that CD outperforms the performance of MWT, EMD, and VMD, exhibiting, at the same time, high noise robustness. We also show that the CD disentangles lowfrequency temporal modulations of sounds, supporting perceptual phenomena in both speech and music. Clearly, CD paves the way for higher-level speech perception deep learning models and also for efficient cochlear implants design. Future work to incorporate the nonlinearity of the cochlea into a binaural hearing framework is underway.

Fetal cardiac monitoring is very helpful in the early detection of the potential risk of fetal cardiac abnormalities, which enables prompt preventative care and ensures safe births. As a result, it is crucial to regularly check on the embryonic heart. Methods of non-invasively fetal ECG extraction from maternal abdominal ECG signal are thoroughly discussed. Although fetal signals are generally obscured by maternal ECG signals and noise, extracting a clean fetal ECG is a significant difficulty. The majority of techniques for fetal ECG extraction include many extraction steps. We describe a unique method for splitting a single-channel maternal abdominal ECG into maternal and fetus ECG employing two parallel U-nets with transformer encoding, which we refer to as W-NEt TRansformers (W-NETR). Due to its enhanced capacity to simulate remote interactions and capture global context, the suggested pipeline utilizes the self-attention mechanism of the transformer. We tested the proposed pipeline on synthetic and real datasets and outperformed the current state-of-the-art deep learning models. The proposed model achieved the best results on both datasets for QRS detection precision, recall, and F1 scores. More specifically, it achieved F1 score of 99.88% and 98.9% on the real ADFECGDB and PCDB datasets, respectively. These encouraging results highlight the suggested W-NETR's effectiveness in precisely extracting the fetal ECG, which was achieved with high SSIM and PSNR values in the results. This provides the bed set for long-term maternal and fetal monitoring via portable devices as the proposed system performs real-time execution.

The automated recognition of human emotions plays an important role in developing machines with emotional intelligence. Major research efforts are dedicated to the development of emotion recognition methods. However, most of the affective computing models are based on images, audio, videos and brain signals. Literature lacks works that focus on utilizing only peripheral signals for emotion recognition (ER), which can be ideally implemented in daily life settings. Therefore, this paper present a framework for ER on the arousal and valence space, based on using multi-modal peripheral signals. The data used in this work were collected during a debate between two people using wearable devices. The emotions of the participants were rated by multiple raters and converted into classes in correspondence to the arousal and valence space. The use of a dynamic threshold for ratings conversion was investigated. An ER model is proposed that uses a Long Short-Term Memory (LSTM)-based architecture for classification. The model uses heart rate (HR), temperature (T), and electrodermal activity (EDA) signals as its inputs with emotional cues. Additionally, a post-processing prediction mechanism is introduced to enhance the recognition performance. The model is implemented to study the use of individual and different combinations of the peripheral signals, as well as utilizing annotations from different ratings. Additionally, it is employed for classification of valence and arousal in an independent and combined fashion, under subject dependent and independent scenarios. The experimental results have justified the efficient performance of the proposed framework, achieving classification accuracy 96% and 93% for the independent and combined classification scenarios, accordingly. The comparison of the achieved performance against the baseline methods shows the superiority of the proposed framework and the ability to recognize arousal-valance levels with high accuracy from peripheral signals, in real-life scenarios.

The current study aims at creating an algorithm able to detect Coronary Artery Disease (CAD), using ultrashort (duration of 30 seconds) one-lead ECG recordings. The presented method is designed to allow both electrode and noisy recordings (deriving from a smartwatch) as input. This is achieved by using an autoencoder neural network, which inspects the quality of each recording. The algorithm’s core is a Support Vector Machine (SVM) model, which evaluates each patient’s recordings and predicts whether they indicate CAD. Using statistics and combining the models mentioned above, a light, reliable, easy to use predicting system is created, suitable for deployment in a mobile application, which uses a smartwatch as its recording tool.

The variational mode decomposition (VMD) is a widely applied optimization-based method, which analyzes non-stationary signals concurrently. Correspondingly, its recently proposed multivariate extension, i.e., MVMD, has shown great potentials in analyzing multichannel signals. However, the requirement of presetting the number of extracted components K diminishes the analytic property of both VMD and MVMD methods. This work combines MVMD with the noise injection paradigm to propose an efficient alternative for both VMD and MVMD, i.e., the noise-assisted MVMD (NA-MVMD), that aims at relaxing the requirement of presetting K, as well as improving the quality of the resulting decomposition. The noise is injected by adding noise variables/channels to the initial signal to excite the filter bank property of VMD/MVMD on white Gaussian noise. Moreover, an alternative approach of updating center frequencies is proposed, which uses the centroid of the generalized cross-spectrum instead of a simple average of the individual spectral centroids, showing faster convergence. The NA-MVMD is applied to both univariate and multivariate synthetic signals, showing improved analytical ability, noise intolerance, and less sensitivity in selecting the K parameter.

OBJECTIVES: The purpose of this study was to set an optimal fit of the estimated LVEF at hourly intervals from 24-hour ECG recordings and compare it with the fit based on two gold-standard guidelines. METHODS: Support vector regression (SVR) models were applied to estimate LVEF from ECG derived heart rate variability (HRV) data in one-hour intervals from 24-hour ECG recordings of patients with either preserved, mid-range, or reduced LVEF, obtained from the Intercity Digital ECG Alliance (IDEAL) study. A step-wise feature selection approach was used to ensure the best possible estimations of LVEF levels. RESULTS: The experimental results have shown that the lowest Root Mean Square Error (RMSE) between the original and estimated LVEF levels was during 3-4 am, 5-6 am and 6-7 pm. CONCLUSION: The observations suggest these hours as possible times for intervention and optimal treatment outcomes. In addition, LVEF classifications following the ACCF/AHA guidelines leads to a more accurate assessment of mid-range LVEF. SIGNIFICANCE: This study paves the way to explore the use of HRV features in the prediction of LVEF percentages as an indicator of disease progression, which may lead to an automated classification process for CAD patients.

The primary manifestations of Parkinson Disease (PD) concern abnormalities of movement associated with the constant deterioration of motor skills. Such motor impairment affects patients' movement accuracy and coordination, disrupting their daily life. Taking into account recent studies stating that computer-based physical therapy games can be used as a PD rehabilitation option, we propose a novel Exergame, the iPrognosis Warming up Game (http://www.i-prognosis.eu/), as a user-friendly tool that could both serve as a computer-based physical therapy game, as well as a means of accurately and automatically identifying the severity of PD motor symptoms. To this regard, we propose a novel deep learning methodology for motor impairment stage prediction that relies solely on human body motion data extracted from the recorded game sessions. Experimental results using a dataset of both early and advanced PD patients reveal a good classification performance of the proposed methodology, predicting the motor impairment stage of PD patients and paving the way for additional research in the field.

The use of serious games in health care interventions sector has grown rapidly in the last years, however, there is still a gap in the understanding on how these types of interventions are used for the management of the Parkinson Disease (PD), in particular. Targeting intelligent early detection and intervention in PD area, the Personalized Game Suite (PGS) design process approach is presented as part of the H2020 i-PROGNOSIS project that introduces the integration of different serious games in a unified platform (i.e., ExerGames, DietaryGames, EmoGames, and Handwriting/Voice Games). From the methodological point of view, to facilitate the visualization of 14 game-scenarios, the system interface and the PD contexts, the storyboarding technique was adopted here. Overall, the realization of the PGS sets the basis for establishing a holistic framework that could aim at improving motor and non-motor symptoms, in order to inform health care providers and policy makers for its inclusion in routine management for PD.

A plethora of studies has shown that working memory, processing speed and fluid intelligence are diminished with aging. However, emotional processing remains relatively stable even though emotional processing alters through aging. Neurophysiological studies have employed emotional stimuli to investigate age differences through Event Related Potentials (ERPs). The present approach used affective visual word stimuli derived from the Greek language. Healthy young and elderly volunteers passively viewed the stimuli which were divided into pleasant, unpleasant and neutral. The study shows differential processing of emotional stimuli in comparison to the neutral in terms of temporal resolution (latency) and activation of neuronal assembles (amplitude). The age factor interacts with emotional dimension through a complex pattern while laterality differences also occur. Our results suggest a difference in the way emotional stimuli are processed during aging through functional compensation.

As a decision support tool, a hybrid modelling can offer the ability to better understand the dynamics of a particular ecosystem. This paper proposes a hybrid approach that may serve as a means to synthesize/represent knowledge obtained from the data, in order to explore online learning environment (OLE) states, based on different scenarios. The potentiality of the quality of collaboration (QoC) within an Internet-based computer-supported collaborative learning environment and the quality of interaction (QoI) with a learning management system (LMS), both involving fuzzy logic-based modeling, as vehicles to improve the personalization and intelligence of an OLE is explored. In this approach, a novel framework could be established, when bridging the fields of blended- and collaborative-learning into an enhanced educational environment. The combined measures (i.e., QoC, QoI) can form the basis for a more realistic approach of OLEs within the concept of semantic Web and the associated Web 3.0 features, as they effectively capture the behaviour of the stakeholders involved in the context of Higher Education. Finally, a potential case study of the examined hybrid modelling (FuzzyQoC/I), referring to the “i-Treasures” European FP7 Programme, is discussed, to explore its functionality/applicability under pragmatic learning scenarios, serving as a proof of concept.

This paper focuses on detecting and localizing anomalous events in videos of crowded scenes, i.e., divergences from a dominant pattern. Both motion and appearance information are considered, so as to robustly distinguish different kinds of anomalies, for a wide range of scenarios. A newly introduced concept based on swarm theory, histograms of oriented swarms (HOS), is applied to capture the dynamics of crowded environments. HOS, together with the well-known histograms of oriented gradients, are combined to build a descriptor that effectively characterizes each scene. These appearance and motion features are only extracted within spatiotemporal volumes of moving pixels to ensure robustness to local noise, increase accuracy in the detection of local, nondominant anomalies, and achieve a lower computational cost. Experiments on benchmark data sets containing various situations with human crowds, as well as on traffic data, led to results that surpassed the current state of the art (SoA), confirming the method's efficacy and generality. Finally, the experiments show that our approach achieves significantly higher accuracy, especially for pixel-level event detection compared to SoA methods, at a low computational cost.

An efficient heart sound segmentation (HSS) method that automatically detects the location of first ( S1) and second ( S2) heart sound and extracts them from heart auscultatory raw data is presented here. The heart phonocardiogram is analyzed by employing ensemble empirical mode decomposition (EEMD) combined with kurtosis features to locate the presence of S1, S2, and extract them from the recorded data, forming the proposed HSS scheme, namely HSS-EEMD/K. Its performance is evaluated on an experimental dataset of 43 heart sound recordings performed in a real clinical environment, drawn from 11 normal subjects, 16 patients with aortic stenosis, and 16 ones with mitral regurgitation of different degrees of severity, producing 2608 S1 and S2 sequences without and with murmurs, respectively. Experimental results have shown that, overall, the HSS-EEMD/K approach determines the heart sound locations in a percentage of 94.56% and segments heart cycles correctly for the 83.05% of the cases. Moreover, results from a noise stress test with additive Gaussian noise and respiratory noises justify the noise robustness of the HSS-EEMD/K. When compared with four other efficient methods that mainly employ wavelet transform, energy, simplicity, and frequency measures, respectively, using the same experimental database, the HSS-EEMD/K scheme exhibits increased accuracy and prediction power over all others at the level of 7-19% and 4-9%, respectively, both in controls and pathological cases. The promising performance of the HSS-EEMD/K paves the way for further exploitation of the diagnostic value of heart sounds in everyday clinical practice.

A tomographic method that efficiently reconstructs three-dimensional fields, despite the ill-posedness of recovering a vector field from line integrals, is presented in this paper. The analysis takes into consideration the methodology set forth in [1] for 2-D reconstruction and demonstrates that with the analogous discretization of the 3-D space and scanning lines, data redundancy is achieved and the solution is obtained from a linear equations system solution, using only information from finite boundary measurements. The adequacy of the method is illustrated by means of simulations on electrostatic fields. The motivation behind this work lies in its potential to bring forward an alternative brain mapping model from EEG recordings.

Wireless capsule endoscopy (WCE) is a revolutionary, patient-friendly imaging technique that enables non-invasive visual inspection of the patient's digestive tract and, especially, small intestine. However, reviewing the endoscopic data is time consuming and requires intense labor of highly experienced physicians. These limitations were the motive to propose a novel strategy for automatic discrimination of WCE images related to ulcer, the most common finding of digestive tract. Towards this direction, WCE data are color-rotated in order to boost the chromatic attributes of ulcer regions. Then, texture information is extracted by utilizing the local binary pattern operator that analyses the spatial structure of the images at a very local level. Experimental results demonstrated promising classification accuracy (91.1%) exhibiting high potential towards a complete computer-aided diagnosis system that will not only reduce WCE data reviewing time, but also serve as an assisting tool for the training of inexperienced physicians.

Wireless Capsule Endoscopy (WCE) is a fairly new technology that offers a low-risk, non invasive visual inspection of the patient's digestive tract, especially the small bowel, that was previously unreachable using the traditional endoscopic methods. However, the large amount of images produced by WCE requires a highly trained physician to manually inspect them; a procedure that is time consuming and prone to human error. This was the rationale to propose a novel strategy for automatic detection of WCE images related to ulcer, one of the most common findings of the digestive tract. This paper introduces a new texture extraction method based on the Discrete Curvelet Transform (DCT), a recent multi-resolution analysis tool. Textural information is acquired by calculating the lacunarity index of DCT subbands of the WCE images. The classification step is performed by a Support Vector Machine (SVM), demonstrating promising classification accuracy (86.5%) and pointing towards further research in this field.

A time-windowing feature extraction approach based on time-frequency (TF) analysis is adopted here to investigate the time-course of the discrimination between musical appraisal electroencephalogram (EEG) responses, under the parameter of familiarity. An EEG data set, formed by the responses of nine subjects during music listening, along with self-reported ratings of liking and familiarity, is used. Features are extracted from the beta (13-30 Hz) and gamma (30-49 Hz) EEG bands in time windows of various lengths, by employing three TF distributions (spectrogram, Hilbert-Huang spectrum, and Zhao-Atlas-Marks transform). Subsequently, two classifiers (k-NN and SVM) are used to classify feature vectors in two categories, i.e., "likeâ and "dislike,â under three cases of familiarity, i.e., regardless of familiarity (LD), familiar music (LDF), and unfamiliar music (LDUF). Key findings show that best classification accuracy (CA) is higher and it is achieved earlier in the LDF case {91.02 ± 1.45% (7.5-10.5 s)} as compared to the LDUF case {87.10 ± 1.84% (10-15 s)}. Additionally, best CAs in LDF and LDUF cases are higher as compared to the general LD case {85.28 ± 0.77%}. The latter results, along with neurophysiological correlates, are further discussed in the context of the existing literature on the time-course of music-induced affective responses and the role of familiarity.

Wireless capsule endoscopy (WCE) is a revolutionary, patient-friendly imaging technique that enables non-invasive visual inspection of the patient's digestive tract, especially small intestine. However, the time-consuming task of reviewing the endoscopic data is a burden for the physicians. This limitation was the motive to propose a novel strategy for automatic discrimination of WCE images related to ulcer, the most common finding of digestive tract. Towards this direction, WCE data are processed with Bidimensional Ensemble Empirical Mode Decomposition to reveal their inherent structural components, and also to reconstruct a new refined image. Then, texture information is extracted by analyzing the intrinsic second/higher-order correlation of the original image and by calculating the lacunarity index of the refined image. Experimental results demonstrated promising classification accuracy (97%) exhibiting high potential towards a complete computer-aided diagnosis system.

This paper aims at providing a novel method for evaluating the emotion elicitation procedures in an electroencephalogram (EEG)-based emotion recognition setup. By employing the frontal brain asymmetry theory, an index, namely asymmetry Index (AsI), is introduced, in order to evaluate this asymmetry. This is accomplished by a multidimensional directed information analysis between different EEG sites from the two opposite brain hemispheres. The proposed approach was applied to three-channel (Fp1, Fp2, and F3/F4 10/20 sites) EEG recordings drawn from 16 healthy right-handed subjects. For the evaluation of the efficiency of the AsI, an extensive classification process was conducted using two feature-vector extraction techniques and a SVM classifier for six different classification scenarios in the valence/arousal space. This resulted in classification results up to 62.58% for the user independent case and 94.40% for the user-dependent one, confirming the efficacy of AsI as an index for the emotion elicitation evaluation.

Sign language (SL) forms an important communication canal for the deaf. In this paper, enhanced SL recognition, by relating the individual way of signing with the signer's level of deafness (LoD) through a novel hybrid adaptive weighting (HAW) process applied to surface electromyogram and 3-D accelerometer data, is proposed. Using a LoD-driven genetic algorithm, HAW optimally weights the intrinsic modes of the acquired signals, preparing them for sample entropy (SampEn) estimation that follows. The resulting feature set, namely, weighted intrinsic-mode entropy (IMEn) (wIMEn), aims at increasing the SL-sign-classification accuracy alone or boosted by signer identification and/or signer's LoD-based group identification. The wIMEn was compared with three other feature sets, i.e., time frequency, SampEn, and IMEn, regarding their discrimination ability (both among signers and SL signs). Data from the dominant hand of nine subjects with various LoD were analyzed for the classification of 61 Greek SL (GSL) signs. Experimental results have shown that the introduced wIMEn feature set exhibited higher performance compared to others, both in signer identification and signer's LoD-based group identification and in GSL sign classification. The findings suggest that LoD could be considered in the construction of a signer-independent SL-classification system toward the enhancement of its performance.

This paper aims at providing a new feature extraction method for a user-independent emotion recognition system, namely, HAF-HOC, from electroencephalograms (EEGs). A novel filtering procedure, namely, Hybrid Adaptive Filtering (HAF), for an efficient extraction of the emotion-related EEG-characteristics was developed by applying Genetic Algorithms to the Empirical Mode Decomposition-based representation of EEG signals. In addition, Higher Order Crossings (HOCs) analysis was employed for feature extraction realization from the HAF-filtered signals. The introduced HAF-HOC scheme incorporated four different classification methods to accomplish a robust emotion recognition performance. Through a series of facial-expression image projection, as a Mirror Neuron System-based emotion elicitation process, EEG data related to six basic emotions (happiness, surprise, anger, fear, disgust, and sadness) have been acquired from 16 healthy subjects using three EEG channels. Experimental results from the application of the HAF-HOC to the collected EEG data and comparison with previous approaches have shown that the HAF-HOC scheme clearly surpasses the latter in the field of emotion recognition from brain signals for the discrimination of up to six distinct emotions, providing higher classification rates up to 85.17 percent. The promising performance of the HAF-HOC surfaces the value of EEG signals within the endeavor of realizing more pragmatic, affective human-machine interfaces.

Electroencephalogram (EEG)-based emotion recognition is a relatively new field in the affective computing area with challenging issues regarding the induction of the emotional states and the extraction of the features in order to achieve optimum classification performance. In this paper, a novel emotion evocation and EEG-based feature extraction technique is presented. In particular, the mirror neuron system concept was adapted to efficiently foster emotion induction by the process of imitation. In addition, higher order crossings (HOC) analysis was employed for the feature extraction scheme and a robust classification method, namely HOC-emotion classifier (HOC-EC), was implemented testing four different classifiers [quadratic discriminant analysis (QDA), k-nearest neighbor, Mahalanobis distance, and support vector machines (SVMs)], in order to accomplish efficient emotion recognition. Through a series of facial expression image projection, EEG data have been collected by 16 healthy subjects using only 3 EEG channels, namely Fp1, Fp2, and a bipolar channel of F3 and F4 positions according to 10-20 system. Two scenarios were examined using EEG data from a single-channel and from combined-channels, respectively. Compared with other feature extraction methods, HOC-EC appears to outperform them, achieving a 62.3% (using QDA) and 83.33% (using SVM) classification accuracy for the single-channel and combined-channel cases, respectively, differentiating among the six basic emotions, i.e., happiness, surprise, anger, fear, disgust, and sadness. As the emotion class-set reduces its dimension, the HOC-EC converges toward maximum classification rate (100% for five or less emotions), justifying the efficiency of the proposed approach. This could facilitate the integration of HOC-EC in human machine interfaces, such as pervasive healthcare systems, enhancing their affective character and providing information about the user's emotional status (e.g., identifying user's emotion experiences, recurring affective states, time-dependent emotional trends).

Primary music education introduces children to the world of sound and music. For some children, however, auditory process disorders (APDs) act as a barrier to the comprehension of the sound elements, like rhythmic motives, variations, etc. The effect of APD in music perception is often neglected within classroom activities. This paper tries to shed light upon this effect, by focusing on duration comprehension. Experimental results from APD testing in pupils (9-12 yrs) from three Greek primary schools show that age influences the duration pattern sequence perception. Moreover, sensitivity in spatial sound information is increased in the age of 10 yrs. These results initiate some implications for ICT-based approach in music education that would foster adaptation to pupilspsila special needs during the educational process.

An educational software, namely ldquoSEE and SEErdquo, that aims to enhance literacy skills of deaf or hard of hearing children is presented in this paper. The proposed computer-based educational environment, takes into account childrenpsilas visual learning characteristics. The software provides a series of adjustable functionalities to the teacher, so s/he can create visual-kinetic educational information for each pupil. Moreover, pupils have the opportunity to use bilateral presentations of the information content, by evoking sign language video for each text sentence, accompanied with comprehensive diagrams and pictures. A series of test-questions are also incorporated that reflect the contribution of each educational process to the learning curve. The whole activity is logged and archived at a local database that outputs statistic/activity reports and files. ldquoSEE and SEErdquo can prove to be a useful learning object that contributes to the normalization of the educational environment towards the children needs.

A novel educational environment for kids with autistic spectrum disorders (ASDs), namely NOESIS, is presented in this paper. NOESIS takes into account ASD kidspsila individual characteristics (level of autism, source sensitivity, reaction target, etc), their emotional state (stress level, hyper-/hypo-tension) during their educational procedure, and creativity during guided- and self-activity (e.g., gaming). It adapts to each kidpsilas specific characteristics through system adaptation and self-regulation procedures. Moreover, it provides assistance to the educator for preparation, customization and optimization of the educational material for each kid and provision of enhanced evaluation procedures (scores/tools) via well-managed Web services. Parentspsila updating is also provided via reporting material with learning curve descriptions. Overall, NOESIS contributes to the provision of opportunities to all ASD children to be educated by facilitating access and tuning innovative technology to social needs.

Human-computer interaction requires efficient acquisition of the relevant information from the user. This poses high accuracy in the performance of the relevant systems, especially when visual information is used as the basic means to capture the user's information. At the same time, speed and efficiency under a variety of conditions (e.g, resolution, luminance) is required. From this perspective, a new approach in the automated iris and gaze detection problem is presented here, using the chrominance of low resolution webcam data. The algorithms developed, when combined with a face tracker, exhibit high performance, circumventing the variability of the lighting conditions and/or user's facial characteristics.

A complexity-based analysis of the turn-taking sequences produced during peers' computer-mediated collaboration is presented in this paper. The collaborative contributions monitored by the system, namely Lin2k, are mapped to turn-taking sequences, which, in turn, are transformed to symbol-sequences and analyzed for pattern extraction. The use of a normalized complexity measure when applied to collaborative data from the field of environmental engineering reveals peers' tendencies to more complex turn-taking patterns when they receive appropriate feedback from the Lin2k. The efficiency of the proposed complexity analysis in pattern identification and its simplicity in implementation makes it a useful tool for effective tracking of temporal changes in collaborative activity.

The proposed paper aims to model the non-verbal interactions that occur during a computer-mediated music composition process between pupils in primary school within a collaborative framework. The socio-cultural approach, drawn from Vygotsky's theoretical context, is being further developed and some of its aspects are outlined and used as a framework to further understand the collaborative processes that occur during collaborative learning. The latter is achieved by identifying effective peer's verbal and non-verbal interactions in a collaborative computer-based music composition environment (CCMCE).The present paper aims to explore and model children's patterns of body language signs which contribute to measure peer's active and passive collaborative learning in CCMCE. Experimental results from a case study in a primary state school in Greece promise to yield important cues for peer's thinking, concentration and mental processing that will be useful to anyone engaged in examining children's computer-mediated collaborative interactions.

The identification problem of P seismic phase onset has been addressed, based on the maximum-kurtosis assumption, /spl kappa/-statistics, and the Chebyshev inequality. Depending on two statistical decision criteria, the proposed approach provides either a single P onset peak or an interval in which the P onset exists, both with a confidence percentage. Results on real seismic data evaluated using a performance index justify the contribution of the proposed method toward an accurate and fully automated P onset identification.

A fuzzy logic-based expert system, namely collaboration/reflection-fuzzy inference system (C/R-FIS,) is presented. By means of interconnected fuzzy inference systems (FIS), it automatically evaluates the collaborative activity, during asynchronous, written, Web-based collaboration. This information is used for the provision of enhanced support during the collaboration. The proposed model extents the evaluation system of a Web-based collaborative tool namely Lin2k, which served as a test-bed for the C/R-FIS experimental use. The results proved the potentiality of the proposed model to significantly contribute to the enhancement of the collaborative activity.

An efficient technique for detecting explosive lung sounds (LS) (fine/coarse crackles and squawks) or bowel sounds (BS) in clinical auscultative recordings is presented. The technique is based on a fractal-dimension (FD) analysis of the recorded LS and BS obtained from controls and patients with pulmonary and bowel pathology, respectively. Experimental results demonstrate the efficiency of the proposed method, since it clearly detects the time location and duration of LS and BS, despite their variation either in time duration and/or amplitude. A noise stress test justifies the noise robustness of the FD-based detector, indicating its potential use in everyday clinical practice.

The automatic and accurate P phase arrival identification is a fundamental problem for seismologists worldwide. Several approaches have been reported in the literature, but most of them only selectively deal with the problem and are severely affected by noise presence. In this paper, a new approach based on higher-order statistics (HOS) is introduced that overcomes the subjectivity of human intervention and eliminates the noise factor. By using skewness and kurtosis, two algorithms have been formed, namely, Phase Arrival Identification-Skewness/Kurtosis (PAI-S/K), and some advantages have been gained over the usual approaches, resulting in the automatic identification of the transition from Gaussianity to non-Gaussianity that coincides with the onset of the seismic event, despite noise presence. Experimental results on real seismic data, gathered by the Seismological Network of the Department of Geophysics of Aristotle University, demonstrate an excellent performance of the PAI-S/K scheme, regarding both accuracy and noise robustness. The simplicity of the proposed method makes it an attractive candidate for huge seismic data assessment in a real-time context.

This paper evaluates the compression performance and characteristics of two wavelet coding compression schemes of electrocardiogram (ECG) signals suitable for real-time telemedical applications. The two proposed methods, namely the optimal zonal wavelet coding (OZWC) method and the wavelet transform higher order statistics-based coding (WHOSC) method, are used to assess the ECG compression issues. The WHOSC method employs higher order statistics (HOS) and uses multirate processing with the autoregressive HOS model technique to provide increasing robustness to the coding scheme. The OZWC algorithm used is based on the optimal wavelet-based zonal coding method developed for the class of discrete "Lipschitizian" signals. Both methodologies were evaluated using the normalized rms error (NRMSE) and the average compression ratio (CR) and bits per sample criteria, applied on abnormal clinical ECG data samples selected from the MIT-BIH database and the Creighton University Cardiac Center database. Simulation results illustrate that both methods can contribute to and enhance the medical data compression performance suitable for a hybrid mobile telemedical system that integrates these algorithmic approaches for real-time ECG data transmission scenarios with high CRs and low NRMSE ratios, especially in low bandwidth mobile systems.

The separation of pathological discontinuous adventitious sounds (DAS) from vesicular sounds (VS) is of great importance to the analysis of lung sounds since DAS are related to certain pulmonary pathologies. An automated way of revealing the diagnostic character of DAS, by isolating them from VS, based on their nonstationarity, is presented in this paper. The proposed algorithm uses two adaptive network-based fuzzy inference systems to compose a generalized fuzzy rule-based stationary-nonstationary filter (GFST-NST). The training procedure of the fuzzy inference systems involves the outputs of the wavelet transform-based stationary-nonstationary filter (WTST-NST), proposed by Hadjileontiadis and Panas [1]. The basic idea of the GFST-NST was initially proposed by the authors with the introduction of the fuzzy rule-based stationary-nonstationary filter (FST-NST) [2], tested with the separation of crackles from VS. The main contribution of this paper is the modification of the structure of the FST-NST filter to a serial-type fuzzy filter that, unlike the parallel operation of the FST-NST filter, sends a predicted stationary signal (VS) into the predictor of the nonstationary (DAS). Applying the GFST-NST filter to fine-coarse crackles and squawks, selected from three lung sound databases, the coherent structure of DAS is revealed and they are separated from VS. The separation performance of the GFST-NST filter was evaluated through quantitative and qualitative indexes that proved its efficiency and superiority against the FST-NST filter. When compared to the WTST-NST filter, the GFST-NST filter performed similarly in accuracy and objectiveness, but in a faster way. Thus, the GFST-NST filter combines the separation accuracy of the WTST-NST filter with the real-time implementation of the FST-NST filter, so it can easily be used in clinical medicine as a module of an integrated intelligent patient evaluation system.