The presentation introduces recent advances in networked embedded computing, namely cyberphysical systems (CPSs). CPS is characterized by a tight integration of embedded computing with physical processes, as well as use of advanced networking technologies. Cyber-physical systems demonstrate an extremely broad potential for new applications in general. Those in medicine can dramatically change numerous healthcare procedures and medical workflows. At the same time, designing of a cyber-physical system comprises many challenges. These challenges are related to required dependability, heterogeneity, multidisciplinary design team, and multirole human-machine interfaces among others. The main design strategies for cyber-physical systems are design modelbased. They need however to be extended with knowledge modeling and support for distributed multidisciplinary collaboration.

Recording and analysis of fetal heart rate (FHR) signal is nowadays the primary method for the biophysical assessment of the fetal state. Since the correct interpretation of crucial FHR characteristics is difficult, methods of automated quantitative signal evaluation are still the subject of the research studies. In the following paper we investigated the possibility of improvement of the fetal state evaluation on the basis of the epsilon-insensitive learning (eIL). We examined two eIL procedures integrated with fuzzy clustering algorithms as well as different methods of logical interpretation of the fuzzy conditional statements. The quality of the FHR signal classification was evaluated using the data collected with the computerized fetal surveillance system. The learning performance was measured with the number of correct classification (CC) and overall quality index (QI) defined as a geometric mean of sensitivity and specificity. The obtained results (CC = 88 % and QI = 87 %) show a high efficiency of the fetal state assessment using the epsilon-insensitive learning based methods.

Telemedical system for fetal home monitoring with smart selection of signal analysis algorithms is presented in this paper. Fetal monitoring signals are provided by a mobile instrumentation consisting of bioelectrical signal recorder and tablet PC which retrieves and processes the data as well as provides wireless data transmission based on Internet. The fetal surveillance system enables analysis, dynamic presentation and archiving of acquired signals and medical data. Novelty of the proposed approach relies on smart fitting of the algorithms for analysis of the abdominal signals in mobile instrumentation, as well as on controlling of the fetal monitoring session from the surveillance center. These actions are performed automatically through continuous analyzing of the signal quality and the reliability of the quantitative parameters determined for the acquired signals. Using that approach the amount and content of data transmitted through remote channels to the surveillance center can be controlled to ensure the most reliable assessment of the fetal well-being.

The most important features indicating appropriate fetal development are the measures of instantaneous variability of a fetal heart rate (FHR), describing fluctuations of the beat-to-beat heart intervals. The most popular method for the FHR acquisition is the Doppler ultrasound technique. However, it is very sensitive to various motion artifacts distorting the signal being acquired. The aim of our work was to evaluate the influence of signal loss episodes on the parameters quantitatively describing the instantaneous variability of the FHR. For this purpose we artificially inserted signal loss episodes to the recordings, in different patterns and percentage, in accordance with the real characteristics of the signal loss segments. We particularly would like to answer the question if the signals with significant amount of signal loss can be reliably evaluated by means of instantaneous variability measures, and which of these measures (numerical indices) are more robust to the missing values.

Analysis of movement activity is important since it enables detection of nonreactive fetal heart rate recordings. The aim of the study was to develop an algorithm for automated detection of the fetal movement activity (actogram), based on analysis of the Doppler ultrasound signal, and to evaluate a reliability of the actogram as a source of information about the fetal movements. Bandpass filtering (20-80 Hz) was used to separate the actogram signal describing the fetal movement activity. Simultaneously there were recorded the markers of fetal movements perceived by mother, being the reference information. For the determination of the binary actogram, the authors proposed an algorithm in which the classification threshold was estimated at the beginning of each recording and was adaptively modified during its duration. The algorithm ensured detection of up to 89% of movement episodes corresponding to movements perceived by mother. At the same time almost as high number of episodes not related to the reference information was recognized. Obtained results revealed that the automated analysis of fetal movements is characterized by much higher sensitivity of movement episode detection compared to the maternal perception.

Cardiotocographic (CTG) monitoring is a method of assessing fetal state. Since visual analysis of CTG signal is difficult, methods of automated qualitative fetal state evaluation on the basis of the quantitative description of the signal are applied. The appropriate selection of learning data influences the quality of the fetal state assessment with computational intelligence methods. In the presented work we examined three different feature selection procedures based on: principal components analysis, receiver operating characteristics and guidelines of International Federation of Gynecology and Obstetrics. To investigate their influence on the fetal state assessment quality the benchmark SisPortor dataset and the Lagrangian support vector machine were used.

The paper introduces the problem of designing a telemedical system for pregnancy monitoring at home. It focuses on design challenges concerning embedded computing and networking, requirements modelling, and presents the architecture and solutions when based on new class Medical CyberPhysical Systems (MCPS). The proposed system consists of a Body Area Network (BAN) of advanced sensors that are interconnected on a body of a pregnant woman, a Personal Area Network (PAN) that is responsible for embedded processing of physical signals, smart alarms, data transmission and communication with the Surveillance Centre located in hospital. It is expected that this dependable telemedical system will provide a high societal value to women with high-risk pregnancy.

This paper presents a multistep approach for segmentation of micro tomography images. Various images of porous structures were studied. Proper segmentation of that images is necessary to create 3D models of these structures. The introduced algorithm concerns finding the proper way of image filtering before the use of Canny-Deriche edge detection to obtain the best possible segmentation.

Cell number, density and volume of white and gray matter in brain structures are not constant values. Cellular alterations in brain areas might coincide with neurological and psychiatric pathologies as well as with changes in brain functionality during selection experiments, pharmacological treatment or aging. Several softwares were created to facilitate quantitative analysis of brain tissues, however results obtained from these softwares require multiple manual settings making the computing process complex and time-consuming. This study attempts to establish half automated software for fast, ergonomic and an accurate analysis of cellular density, cell number and cellular surface in morphologically different brain areas: cerebral cortex, pond and cerebellum. Images of brain sections of bank voles stained with standard cresyl-violet technique (Nissl staining), were analyzed in designed software. Results were compared with other commercially available tools regarding number of steps to be done by user and number of parameters possible to measure.

Human face depicts what happens in the soul, therefore correct recognition of emotion on the basis of facial display is of high importance. This work concentrates on the problem of optimal classification technique selection for solving the issue of smiling versus neutral face recognition. There are compared most frequently applied classification techniques: k-nearest neighbourhood, support vector machines, and template matching. Their performance is evaluated on facial images from several image datasets, but with similar image description methods based on local binary patterns. According to the experiments results the linear support vector machine gives the most satisfactory outcomes for all conditions.

The paper presents a method for segmentation of images using region growing, with modification through the use of a correction coefficient based on the variation of intensity (brightness) in the neighborhood of the pixel of the interest. A method for the quantification of variability is based on differences in intensity, as well as the differences in intensity gradients in the surrounding pixels [10]. Evaluation of the gradients were determined by means of numerical differentiation, using the polynomial approximation. The article presents the effects of application of developed methods for segmentation of images of the brain, lungs and heart.

In this paper a new method for the reduction of multiplicative noise in digital images is described. The proposed algorithm is a modification of the Mean-Shift (MS) filter which is based on the concept of the Non-Local Means (NLM) denoising. The proposed algorithm does not focus on single pixels only, as in the case of the mean-shift technique, but also on their neighborhoods. The performance of the novel approach is experimentally verified and the obtained results prove that the new design is superior both to the MS and NLM techniques.

In this paper a method is introduced which enables automatic detection of immunogold markers in transmission electron micrographs. Immunogold markers are used in electron microscopy to determine sub-cellular location of biological relevant macromolecules, such as proteins, lipids, carbohydrates, and nucleic acids. The proposed method combines image segmentation and feature localization approaches to improve accuracy of the immunogold markers detection in low contrast and highly textured image regions. A segmentation algorithm is intended in this study, which applies a flood-fill morphological operation. Accuracy of this method was evaluated by using electron microscopy images of human colorectal carcinoma cells. The experimental results show that the introduced method enables detection of immunogold markers with low false positive and false negative rates.

The paper concerns behavioral biometrics, specifically issues related to the verification of the identity of computer systems users based on user profiling. The profiling method for creating a behavioral profile based on multiple aspects of user activity in a computer system is presented. The work is devoted to the analysis of user activity in environments with a graphical user interface GUI. Mouse activity, keyboard and software usage are taken into consideration. Additionally, an attempt to intrusion detection based on the proposed profiling method and statistical measures is performed. Preliminary studies show that the proposed profiling method could be useful in detecting an intruder masquerading as an authorized user of the computer system. This article presents the preliminary research and conclusions.

Aged people, patients with chronic disease, patients at remote location need continuous monitoring under healthcare professionals. Remote health monitor is likely to be an effective approach to provide healthcare service in a simple and cost effective way. However, effective implementation of this type of framework needs consideration of variety of security threats. In this paper, a layer based remote health monitor framework is proposed to analyze health condition of patients from remote places. Beside this, a multi-modal biometric authentication mechanism is proposed here to reduce misuse of health data and biometrics templates in heterogeneous cloud environment. Main focus of the paper is to design semi-continuous authentication mechanism after establishing mutual 1:1 trust relationship among the participants in cloud environment. Behavioral biometrics keystroke analysis is fused with physiological biometrics face recognition to enhance accuracy of authentication. Instead of considering traditional performance evaluation parameters for biometrics, this paper considers a few performance metrics for determining efficiency of semi-continuous verification of the proposed framework.

The authentication of users is of utmost importance in remote applications such as healthcare, banking, stock markets, etc. Key stroke dynamics are popular biometrics tools used for this purpose. Continuous authentication requires free text analysis which has a number of challenges. This paper has proposed a solution to identify the existence of a unique pattern in each individual user’s keystroke dynamics. However, dense zone identification is important factor in forming the intelligent database of user profile for authentication. The authors have categorized basic key stroke features of digraph into 57 groups depending on distance traversed while moving from one key to another. The paper also includes graphical plots of the grouping of time vector which has unveiled some characteristics of overlapping typing style of users. The authors hope to extend this logic for identifying behavioral disorders in users.

In this paper we present a system for offline signature verification using direction-based Shape Contexts. Images of handwritten signatures were thinned using KMM algorithm and then represented by a set of Shape Context descriptors computed separately in 4 directions in pixel’s 8-neighborhood. The distance measure used to compare Shape Contexts was based on L2 norm. The experiments were conducted using signatures from GPDS database.

In this paper a new method of handwritten signatures verification has been proposed. This method, for each signature, creates complex features which are describing this signature. These features are based on dependencies analysis between dynamic features registered by tablets. These complex features are then used to create vectors describing the signature. Elements of these vectors are calculated using measures proposed in this work. The similarity between signatures is assessed by determining the similarity of vectors in the compared signatures. Research, whose results will be presented in the further part of this work, have shown a high efficiency of verification using proposed method.

The algorithms of pattern recognition were used for differentiation between two forms of EmeryDreifuss muscular dystrophy (EDMD), i.e. autosomal-dominant laminopathy (AD-EDMD) and Xlinked emerynopathy (X-EDMD). A set of some matrix metalloproteinases (MMPs) and their tissue inhibitors (TIMPs) in serum of EDMD patients and healthy subjects were treated as features. In concluding MMPs and TIMPs levels are helpful to identifying the EDMD patients and the disease progress.

This paper presents a medical diagnosis support system based on an ensemble of single parameter k–NN classifiers. System was verified on a database containing real blood test results of diagnosed patients with a liver fibrosis. This dataset contains problems typical to a real medical data – especially missing values. Paper also describes the process of selecting a subset of parameters used for further evaluation (feature selection/elimination algorithm). Complete database contains many parameters, but not all are important for diagnosis, thus eliminating them is an important step. A comparison of proposed method of classification and feature selection with methods known from literature has also been presented.

The selection of classifiers is one of the important problems in the creation of ensemble of classifiers. The paper presents the static selection in which a new method of calculating the weights of individual classifiers is used. The obtained weights can be interpreted in the context of the interval logic. It means that the particular weights will not be provided precisely but their lower and upper values will be used. A number of experiments have been carried out on several medical data sets.

This paper addressees the problem of an early diagnosis of Parkinson’s disease by the classification of characteristic features of person’s voice. A new, two-step classification approach is proposed. In the first step, the voice samples are classified using standard state-of-the-art classifiers. In the second step, the classified samples are assigned to patients and the final classification process based on majority criterion is performed. The advantage of using our new approach is the resulting, reliable patientoriented medical diagnose. The proposed two-step method of classification allows also to deal with the variable number of voice samples gathered for every patient. Preliminary experiments revealed quite satisfactory classification accuracy obtained during the performed leave-one-out cross validation.

In the last 45 years nurse scheduling has received considerable attention in the research community. Nurse rostering can be described as a task of finding a duty roster for a set of nurses in such a way that the rosters comply with work regulations and meet the management’s requests. The objective varies from minimizing the costs of float nurses or minimizing under-staffing to maximizing the degree to which the nurses’ requests are met. In logistics, one aspect is optimization of the steady flow of materials through a network of transport links and storage nodes, and the other is, coordination of a sequence of resources, such as staffing and scheduling clinical resources. The period up to 2000 is characterized by using mathematical programming and objective functions to solve nurse rostering problem. In the period after 2000 the focus of researches aimed at solving nurse rostering and scheduling problem becomes implementation of meta-heuristics and multi-objective functions. The aim of this paper is to present the latest researches conducted in last ten years.