The aim of this work is to present two different algorithmic pipelines for human activity recognition (HAR) in real time, exploiting inertial measurement unit (IMU) sensors. Various learning classifiers have been developed and tested across different datasets. The experimental results provide a comparative performance analysis based on accuracy and latency during fine-tuning, training and prediction. The overall accuracy of the proposed pipeline reaches 66 % in the publicly available dataset and 90% in the in-house one.

Cognitive workload is a critical feature in related psychology, ergonomics, and human factors for understanding performance. However, it still is difficult to describe and thus, to measure it. Since there is no single sensor that can give a full understanding of workload, extended research has been conducted in order to present robust biomarkers. During the last years, machine learning techniques have been used to predict cognitive workload based on various features. Gaze extracted features, such as pupil size, blink activity and saccadic measures, have been used as predictors. The aim of this study is to use gaze extracted features as the only predictors of cognitive workload. Two factors were investigated: time pressure and multi tasking. The findings of this study showed that eye and gaze features are useful indicators of cognitive workload levels, reaching up to 88% accuracy.

In the last years, many studies have been investigating emotional arousal and valence. Most of them have focused on the use of physiological signals such as EEG or EMG, cardiovascular measures or skin conductance. However, eye related features have proven to be very helpful and easy to use metrics, especially pupil size and blink activity. The aim of this study is to predict emotional arousal and valence levels which are induced during emotionally charged situations from eye related features. For this reason, we performed an experimental study where the participants watched emotion-eliciting videos and self-assessed their emotions, while their eye movements were being recorded. In this work, several classifiers such as KNN, SVM, Naive Bayes, Trees and Ensemble methods were trained and tested. Finally, emotional arousal and valence levels were predicted with 85 and 91% efficiency, respectively.

This paper presents a clustering-based method for the detection of Multiple Sclerosis (MS) lesions, by including anatomical information, brain geometry and lesion features, while volume quantification is performed. The proposed method utilizes Fluid Attenuated Inversion Recovery (FLAIR) images for the delineation of the plaques and brain atrophy estimation. The methodology includes five steps: (i) image preprocessing, (ii) image segmentation utilizing the K-means clustering algorithm, (iii) post processing for elimination of false positives, (iv) delineation and visualization of the MS lesions, and (v) brain atrophy estimation. It is implemented in two different datasets; (a) a dataset of 3D FLAIR MR Images, acquired in 30 MS patients, and (b) a dataset of 15 FLAIR MR Images, provided by the MICCAI Challenge 2016. A sensitivity 73.80%, and 71.52% was achieved for the two datasets, respectively. Brain atrophy was determined only on the first dataset, since follow up scans are available.

The aim of this work is to present the architecture of the ProMiSi tool, a software for the analysis of magnetic resonance imaging and the extraction of information on the progression of multiple sclerosis disease. ProMiSi is based on the automatic processing, segmentation and post-processing of MRI for the automatic labeling, visualization and volumetric quantification of segmentable brain structures from magnetic resonance image. The combination of the above mentioned volumetric results with other type of information (e.g. clinical, demographic etc.), through autonomous learning intelligent techniques, allows the evaluation of the severity and the progress prediction of the multiple sclerosis and consequently the personalized management of the disease. A proof of concept study with 30 patients will take place for the validation of the algorithms, while ProMiSi will be evaluated in terms of functionality, usability, reliability, performance and supportability.

The aim of this work is to present a machine learning based method for the prediction of adverse events (mortality and relapses) in patients with heart failure (HF) by exploiting, for the first time, measurements of breath and saliva biomarkers (Tumor Necrosis Factor Alpha, Cortisol and Acetone). Data from 27 patients are used in the study and the prediction of adverse events is achieved with high accuracy (77%) using the Rotation Forest algorithm. As in the near future, biomarkers can be measured at home, together with other physiological data, the accurate prediction of adverse events on the basis of home based measurements can revolutionize HF management.

The aim of this work is to estimate the medication adherence of patients with heart failure through the application of a data mining approach on a dataset including information from saliva and breath biomarkers. The method consists of two stages. In the first stage, a model for the estimation of adherence risk of a patient, exploiting anamnestic and instrumental data, is applied. In the second stage, the output of the model, accompanied with data from saliva and breath biomarkers, is given as input to a classification model for determining if the patient is adherent, in terms of medication. The method is evaluated on a dataset of 29 patients and the achieved accuracy is 96%.

Heart Failure (HF) is a rapidly increasing cardiovascular chronic disease that affects millions of people globally. Lack of proper management of HF patients increases the risk of frailty and other undesirable effects and contributes to loss of independence. The engagement of the HF patient and all actors related to his/her disease management is critical for empowering the patients in achieving sustainable behaviour change, regarding their adherence and compliance. To address this, the concept and the architecture of a mobile co-operative platform are described. The design and development is based on a multi-stakeholder patient centered mHealth ecosystem for HF patients that will facilitate the collaboration of multidisciplinary actors.

Stents are medical devices used in cardiovascular intervention for unblocking the diseased arteries and restoring blood flow. During stent implantation the deformation of the arterial wall as well as the resulted stresses caused in the arterial morphology are studied. In this paper we study the effect of the composition of the atherosclerotic plaque during the stent deployment procedure, using Finite Element modeling. The stenting procedure is simulated for two different cases; in the first the presence of the plaque is ignored whereas in the second a three dimensional (3D) stiff calcified plaque is located in the stenotic area of the artery. Results indicate that in the second case the von Mises stresses in the arterial wall are higher than the stresses occurred in the first case. In addition, the distribution of the arterial von Mises stress depends on the plaque composition.

The aim of this work is to present an automated non supervised method for the detection of occlusal caries based on photographic color images. The proposed method consists of three steps: (a) detection of decalcification areas, (b) detection of occlusal caries areas, and (c) fusion of the results. The detection process includes pre-processing of the images, segmentation and post-processing, where objects not corresponding to areas of interest are eliminated through the utilization of rules expressing the medical knowledge. The preprocessing, segmentation and post-processing are differentiated depending on the areas that have to be detected (decalcification or occlusal areas). The method was evaluated using a set of 60 images where 286 areas of interest were manually segmented by an expert. The obtained sensitivity and precision is 92% and 80%, respectively.

RT3S (Real Time Simulations for Safer vascular Stenting) is a partially EU-funded research project aiming to develop a software tool for supporting physicians during the preplanning of endovascular stenting procedures. The project is expected to improve the way limb-saving, minimally-invasive stenting procedures are currently performed, with positive clinical and economic impact. A hypothetical cohort of patients was modeled and used to simulate the patient's progression through the treatment of Peripheral Artery Disease (PAD). A Markov health state-based model was implemented, based on clinical and economic parameters derived from the literature and clinicians' feedback. The health-economic analysis allowed quantitative estimation of the economic and clinical advantages related to the implementation of the clinical software. A quantitative estimation of the potential health-economic impact was achieved. The model proved to accord well with observed predictions from endovascular experts in the field. It represents an important reference for future assessment of IT-related innovations in the healthcare sector.

The RT3S E-learning environment enables experts in vascular medicine to prepare educational training content for their trainees (i.e., future endovascular surgeons). Influenced by the Learning Design (LD) information model and building-upon LAMS, the learning process is realized by means of training scenarios (i.e., as an interactive sequence of learning steps). In RT3S, creating learning scenarios does not require that the editors of the scenarios are familiar with the underlying LAMS environment. The RT3S authoring environment is easy to use and customized (i.e., it can be adapted to the needs of the tutor and of the scenario) and enables tutors (e.g., expert surgeons) to easily prepare new educational content. Students (future surgeons) are trained on the assessment of real (and realistic) patient data and on decision-making processes for the management and treatment of patients.

The RT3S training application is designed to enable tutors (expert surgeons) prepare educational content for their students, and the students (future surgeons) to be trained on the assessment of real patient cases on peripheral vascular stenting. It incorporates state-of-art technologies (i.e., LAMS) for system design and adopts the latest e-learning standards (i.e., IMS Learning Design) for the implementation of training scenarios. In this work, we focus on the structure of the clinical scenarios that constitute the core of the RT3S training application. In RT3S, creating or accessing learning scenarios does not require that the author (tutor) or the learner (student) be familiar with any programming environment. The learning scenario allows the learners to follow an educational scenario and interact with their tutor.

Tremor is the most common motor disorder of Parkinson's disease (PD) and consequently its detection plays a crucial role in the management and treatment of PD patients. The current diagnosis procedure is based on subject-dependent clinical assessment, which has a difficulty in capturing subtle tremor features. In this paper, an automated method for both resting and action/postural tremor assessment is proposed using a set of accelerometers mounted on different patient's body segments. The estimation of tremor type (resting/action postural) and severity is based on features extracted from the acquired signals and hidden Markov models. The method is evaluated using data collected from 23 subjects (18 PD patients and 5 control subjects). The obtained results verified that the proposed method successfully: 1) quantifies tremor severity with 87 % accuracy, 2) discriminates resting from postural tremor, and 3) discriminates tremor from other Parkinsonian motor symptoms during daily activities.

The accurate diagnosis of diseases with high prevalence rate, such as Alzheimer, Parkinson, diabetes, breast cancer, and heart diseases, is one of the most important biomedical problems whose administration is imperative. In this paper, we present a new method for the automated diagnosis of diseases based on the improvement of random forests classification algorithm. More specifically, the dynamic determination of the optimum number of base classifiers composing the random forests is addressed. The proposed method is different from most of the methods reported in the literature, which follow an overproduce-and-choose strategy, where the members of the ensemble are selected from a pool of classifiers, which is known a priori. In our case, the number of classifiers is determined during the growing procedure of the forest. Additionally, the proposed method produces an ensemble not only accurate, but also diverse, ensuring the two important properties that should characterize an ensemble classifier. The method is based on an online fitting procedure and it is evaluated using eight biomedical datasets and five versions of the random forests algorithm (40 cases). The method decided correctly the number of trees in 90% of the test cases.

In this paper, the Bayesian framework is used for the analysis of functional MRI (fMRI) data. Two algorithms are proposed to deal with the nonstationarity of the noise. The first algorithm is based on the temporal analysis of the data, while the second algorithm is based on the spatiotemporal analysis. Both algorithms estimate the variance of the noise across the images and the voxels. The first algorithm is based on the generalized linear model (GLM), while the second algorithm is based on a spatiotemporal version of it. In the GLM, an extended design matrix is used to deal with the presence of the drift in the fMRI time series. To estimate the regression parameters of the GLM as well as the variance components of the noise, the variational Bayesian (VB) methodology is employed. The use of the VB methodology results in an iterative algorithm, where the estimation of the regression coefficients and the estimation of variance components of the noise, across images and voxels, are interchanged in an elegant and fully automated way. The performance of the proposed algorithms (under the assumption of different noise models) is compared with the weighted least-squares (WLSs) method. Results using simulated and real data indicate the superiority of the proposed approach compared to the WLS method, thus taking into account the complex noise structure of the fMRI time series.

The aim of this work is to propose a new approach for the determination of the design matrix in fMRI experiments. The design matrix embodies all available knowledge about experimentally controlled factors and potential confounds. This knowledge is expressed through the regressors of the design matrix. However, in a particular fMRI time series some of those regressors may not be present. In order to take into account this prior information a Bayesian approach based on hierarchical prior, which expresses the sparsity of the design matrix, is used over the parameters of the generalized linear model. The proposed method automatically prunes the columns of the design matrix which are irrelevant to the generation of data. The evaluation of the proposed approach on simulated and real experiments have shown higher performance compared to the conventional t-test approach.