The paper concerns methods of representation of uncertainty and imprecision in successful medical support applications. Advantages of the methods are pointed out and some of their drawbacks are explained. A method of simultaneous representation of imprecision of symptoms and uncertainty of diagnostic rules is proposed. The method suggests an extension of the Dempster-Sahfer theory for fuzzy focal elements. An example of the method is given and their links as well differences from previous approaches are discussed. Conclusions about uncertainty and imprecision representation in medical diagnosis support are provided.

The analysis of eyes movements is a crucial part of eyes examination performed by clinicians. One of the characteristic type of eyes movements is a saccade. Its accurate detection is the base for further processing including the estimation of saccade parameters such as velocity, amplitude and duration. This paper presents averaging of optokinetic nystagmus (OKN) cycles that allows comparing and detecting different types of nystagmus phenomena. In order to average the OKN cycles the ENG signal needs to be processed. The saccade detection function is used to find the location of saccades in OKN waveform allowing the ENG signal to be divided into cycles. The resulting cycles are aligned using the Fourier shift method and then averaged providing the OKN cycle model, which can be used for evaluating the eyes at different movement conditions.

The analysis of eye movements is valuable in both clinical work and research. One of the characteristic type of eye movements is saccade. The accurate detection of saccadic eye movements is the base for further processing of saccade parameters such velocity, amplitude and duration. This paper presents an accurate saccade detection method which is supported by the fuzzy clustering. The proposed detection function is computationally efficient and precisely determines the time position of the saccadic eye movement event. The described method is characterized by low sensitivity for any kind of noise and can be applied in the analysis of the congenital nystagmus.

In this paper the methods for selecting of the most important parts of the human eyes are described. On the basis of the real 21 channel multispectral images the model of finding the lens and the spot are defined. These methods are based on the most popular algorithms of image processing. The approach to veins detection is still undefined but in the article the most important channels are pointed out and the channel difference between eyelash and the veins is also mentioned.

The nervous system"s ability to receive light stimuli and its" processing in the brain in order to produce a visual impression is the definition of the sense of sight. The anatomical form of the sense organ of vision is comprised of the eyeball, the eye"s protective apparatus, the eye"s movement apparatus and the retinal nerve connections made to structures in the brain. The shape of the eye"s lens, which gives the eye its refraction ability, depends on the voltage present in Zinn’s ligaments that regulate the ciliary muscle. Sharpness of vision is produced by changing the shape of the lens, a reflexive adjustment. A domed lens causes stronger light inflexion and allows a sharp visual appearance of close objects. A flattening of the lens results in less light refraction and the seeing of more distant objects. The lens consists of a capsule, a cortex and a nucleus and it has two convex surfaces: the front and the rear. If we imagine the lens as a plum fruit, the capsule is its skin, the cortex is its flesh and the nucleus is its stone. Proper functioning of the lens is essential for accurate vision. Exact assessments of the eye"s lens and the ability to monitor the status of associated diseases are extremely important. The lens may be tested using a slit lamp, but in any situation where an eye disease prevents this assessment of the lens this examination will obviously be inadequate. Thanks to today"s imaging techniques, we can now assess the anatomical arrangement and condition of the lens, even in the eyes of those patients for whom the use of an imaging lens slit lamp is not possible.

This article shows the limitation of the usage of dimensionality reduction methods. For this purpose three algorithms were analyzed on the real medical data. This data are multispectral images of human skin labeled as tumor or non-tumor regions. The classification of new data required the special algorithm of new data mapping that is also described in the paper. Unfortunately, the final conclusion is that this kind of local embedding algorithms should not be recommended for this kind of analysis and prediction.

The paper deals with processing of imaging data obtained in the course of different medical examination procedures, for use in applications focused on improving both the diagnostic process of a patient as well as extending the training possibilities for medical personnel. Computed tomography (CT) data serves as the main source of structural and volumetric information. Fusion with left ventricular systolic function map, provides a vital information on the state of heart muscle. In parallel, the same set of computed tomography data is reused for simulation of trans-esophageal echocardiography (TEE). The presented simulation setup is used in the course of medical personnel training.

In this paper the quality and analysis of the computed tomography scan sets are presented in the context of creating a 3D/4D model of a heart for the ultrasonography simulator. Data was collected during regular patients examination, using various equipment and technique, therefore not every set has required quality. CT data can be fast characterized with histogram that can show if the brightness ranges of objects (heart structures) are selective. This makes CT data usable for simulation by applying a transform function on the CT images to produce ultrasonographylike images. The aim is to use a PACS system of Hospital, which is the source of data. Therefore a proper technique and system for analysis is needed.

The paper deals with an image database organization and utilization in computer-aided cytology. To illustrate the idea we take as an example the problem of bladder cancer early detection based on urine cytology. In spite of its diagnostic potential for discovering malignancy associated changes (MAC) at the cell level it seems to be underestimated. There is common view that sensitivity of the method, especially for early cancer stages, is relatively low. We depict here just one but significant direction of our works that aims to support pathologists making the diagnosis more accurate and reliable. The key idea relies on automatic searching for MAC by comparing nuclear chromatin structure of objects in a smear with a collection of sample patterns contained in a pathomorphological image database.

Information about the level how well arterial blood is saturated with oxygen, i.e. the overall patient saturation, in particular a patient with a ventricular assist device (VAD) inoculated, is very important for efficient and dependable course of blood circulation support. The paper presents the method for measurement of blood saturation with oxygen that is carried out with use of a commercially available pulse oximetric sensor upon necessary modifications and installed directly on a outflow connector of the VAD. Operability of the presented solution has been verified by experiments on a flow control test bench with use of animal blood. The final validation of the measuring circuit was carried out during experiments with an animal. The completed experiments demonstrated that both the blood saturation with oxygen as well as artificial pulse rate can be measured with accuracy. Unfortunately, similar measurements for venous blood proved infeasible.

This work is an attempt to assess the reliability of indirect abdominal electrocardiography as an alternative technique of fetal monitoring. As a reference signal we used the simultaneously acquired direct fetal electrocardiogram. Each recording consisted of four signals acquired from maternal abdomen and the reference signal acquired directly from fetal head. The first stage of our study concerned the signal loss episodes. In order to reduce the influence of incorrectly detected R-waves, some certain validation rules were applied. In the second stage, the corresponding intervals determined on basis of both acquisition methods were matched and the accuracy of fetal heart rate measurement was evaluated. Although the accuracy of abdominal electrocardiography turned out to be slightly lower than reported for ultrasound method, it still has some unique features deciding of its prevalence in a certain circumstances.

This paper presents 15 texture features based on GLCM (Gray-Level Co-occurrence Matrix) and GLRLM (Gray-Level Run-Length Matrix) to be used in an automatic computer system for breast cancer diagnosis. The task of the system is to distinguish benign from malignant tumors based on analysis of fine needle biopsy microscopic images. The features were tested whether they provide important diagnostic information. For this purpose the authors used a set of 550 real case medical images obtained from 50 patients of the Regional Hospital in Zielona Góra. The nuclei were isolated from other objects in the images using a hybrid segmentation method based on adaptive thresholding and kmeans clustering. Described texture features were then extracted and used in the classification procedure. Classification was performed using KNN classifier. Obtained results reaching 90% show that presented features are important and may significantly improve computer-aided breast cancer detection based on FNB images.

Volumetric parameters of operation of the pulsatile ventricular assist device (VAD) are the key clinical indicators for the evaluation of adequacy of the process of mechanical heart assistance and ensuring the patient’s safety. The bioimpedance method has been proposed to evaluate blood volume changes of the VAD. The method is based on volume electric conductivity of the blood. A change in VAD volume involves changes in impedance. Periodic changes in VAD volume during its operation result in a pulsatile impedance wave; its amplitude in the subperiod of filling and ejection correlates with the volume of blood in the VAD. The paper presents the way of acquisition of the bioimpedance signal. It also presents the results of experimental studies, that were carried out with the use of the flow test stand, and an approach to analysing the bioimpedance signal to evaluate the stroke volume.

The paper presents an unsupervised approach to biomedical signal segmentation. The proposed segmentation process consists of several stages. In the first step, a state-space of the signal is reconstructed. In the next step, the dimension of the reconstructed state-space is reduced by projection into principal axes. The final step involves fuzzy clustering method. The clustering process is applied in the kernel-feature space. In the experimental part, the fetal heart rate (FHR) signal is used. The FHR baseline and the acceleration or deceleration patterns are the main signal nonstationarities but also the most clinically important signal features determined and interpreted in computer-aided analysis.

The subject of examination was mainly the determination of safety of available ECG electrodes use in conditions of defibrillation treatment in monitored patients. The electrodes of four European producers used in healthcare have been tested. For reason of periods of electrodes obtaining and their physical condition, the examinations have comprised four selected years and 7 types of electrodes. For testing needs, the electrodes were glued into pairs, in gel-to-gel mode. Testing was based on requirements of globally used American ANSI/AAMI EC 12:2000/(R)2010 standard for disposable ECG electrodes. For the electrodes pairs tested, determined was the influence of fourfold defibrillation release on polarization potential and their impedance measured with the current of 10 Hz frequency. It has been generally found that in result of fourfold exposition to defibrillating pulse the drop of impedance occurred in all tested, electrode pairs. For each electrodes series, mean value of this drop falls into interval of 10,8% to 57%. It was also pointed out that the ECG electrodes examined, while complying to requirements of the standard, allow the operator to perform the defibrillation procedure safely, including the possibility to fast evaluate the defibrillation result basing on ECG waveform recorded. Such general estimation of safety level of the available European ECG electrodes may be used as some reference in today very diversified market as to provenience end quality of products.

Modern medical devices are very complex, they must show many different kind of information, from ordinary digital values to sophisticated graphical data. Medical devices have to do it in an understandable way to avoid dangerous mistakes. There are several different ways to present the information. The most appropriate is to use display, in this case the best choice would be the color graphical display. Such kind of display allows many different data to be presented clearly and makes possibilities to create an very intuitive user friendly interface, avoiding misinformation. In portable devices this kind of data presentation is great advantage but we need to take under consideration low power consumption. Additionally, in medical portable devices very important aspects are power source operation time prediction possibility and power consumption stability. To achieve this we need to use displays which have well defined power characteristics, because in portable applications, color displays are often most power consuming parts.