Konferenzschriften

Aims/Purpose: Fluorescence lifetime imaging ophthalmoscopy (FLIO) allows in vivo measurement of autofluorescence intensity decays of endogenous fluorophores in the ocular fundus. So far, only devices from Heidelberg Engineering based on the Spectralis system have been used in FLIO research. Here, we present and characterize a new FLIO device based on the RETImap system from Roland Consult. Methods: The device is based on a confocal scanning laser ophthalmoscope (35˚ field, 512 × 512 px). A ps diode laser (BDL-SMN 473 nm, Becker & Hickl GmbH, Berlin, Germany) excites autofluorescence. The fluorescence photons are split into a short (498-560 nm, SSC) and a long (560-720 nm, LSC) spectral channel (one HPM-100-40 detector [Becker & Hickl GmbH] each) and are detected by time-correlated single photon counting (SPC-160, Becker & Hickl GmbH). We determined the maximum laser power (ILT2400, International Light Technologies, Inc. Peabody, MA, USA) and analysed the instrument's behaviour at three different laser power levels (150 μW, 200 μW and max.) in terms of laser spectrum (CAS140CT, Instrument Systems GmbH, Munich, Germany) and instrument response function (IRF). The IRF was determined using a 25 μM Eosin Y solution, mixed with a 5 M solution of potassium iodide, placed in a flat cuvette (110-OS, Hellma GmbH & Co. KG, Müllheim, Germany) in front of the objective lens of the FLIO device. Fluorescence measurements of approximately 1-min duration were performed three times for all three laser powers. The IRF and the full width at half maximum (FWHM) were calculated using FLIMX software (www.flimx.de). Results: The max. laser power was 280 μW. The peak wavelengths of the laser spectra were 467.6 (150 μW), 467.9 (200 μW) and 468.0 nm (280 μW). IRF FWHM in SSC were 298.6 ± 1.1 ps (150 μW), 341.0 ± 2.5 ps (200 μW) and 347.5 ± 6.0 ps (280 μW). In LSC, the IRF FWHM were 290.4 ± 3.8 ps (150 μW), 344.0 ± 3.4 ps (200 μW) and 358.8 ± 1.3 ps (280 μW). Results are mean ± standard deviation. Conclusions: A new fluorescence lifetime imaging ophthalmoscope has been characterized. The device offers a high laser power for fluorescence excitation, a large field of view, a high spatial resolution, and a sufficiently high time resolution. Thus, it is suitable for fluorescence lifetime studies.

Physiological direct current (DC) potential shifts in electroencephalography (EEG) can be masked by artifacts such as slow electrode drifts. To reduce the influence of these artifacts, linear detrending has been proposed as a pre-processing step. We considered quadratic detrending, which has hardly been addressed for ultralow frequency components in EEG. We compared the performance of linear and quadratic detrending in simultaneously acquired DC-EEG and transcutaneous partial pressure of carbon dioxide during two activation methods: hyperventilation (HV) and apnea (AP). Quadratic detrending performed significantly better than linear detrending in HV, while for AP, our analysis was inconclusive with no statistical significance. We conclude that quadratic detrending should be considered for DC-EEG preprocessing.

The importance of document digitization has increased due to recent technological advancements, including in the medical field. Digitization of medical records plays a vital role in the healthcare sector as it helps expedite emergency treatment. Due to the scarcity of published studies and public German textual resources, a medical records database with German handwriting was collected and digitized. In this study, document digitization was accomplished by implementing deep learning, region of interest (ROI) detection, and optical character recognition (OCR) on a dataset containing medical forms filled with German and English characters. To find the best model for ROI detection, YOLOv5, and SSDResNet50 models were utilized and compared with YOLOv5 producing a better mean average precision (mAP) of 0.91. OCR was then carried out on the output from YOLOv5 with two different methods again for comparison. The Gated-CNN-BLSTM algorithm yielded a character error rate (CER) of 9%, while transformer-based OCR (TrOCR) achieved a CER of 6%. The proposed system could be implemented and further tested in local hospitals, with the OCR dictionary being expandable to include other Roman character-based languages.

Magnetic Particle Imaging (MPI) is a fast imaging technique for the visualization of the distribution of superparamagnetic iron-oxide nanoparticles (SPIONs) in 3D. For spatial encoding, a field free area is moved rapidly through the field of view (FOV) generating a localized signal. Fast moving samples, e.g., a bolus of SPIONs traveling through the large veins in the human body carried by blood flow with velocities in the order of ˜45 cm/s and higher, cause temporal blurring in MPI measurements using common sequences and reconstruction techniques. This hampers the evaluation of dynamics of rapidly moving samples. In this abstract, initial results of rapidly moving samples in form of SPION boluses visualized within an MPI scanner are shown.

Abstract: The here presented study investigated the influence of magnetic particle-particle interactions within the MPI tracer MNP on the imaging performance of the tracers. To realize a proper separation and to increase the distances between the individual tracer MNP (which results in reduced particle-particle interactions), the MNP were diluted by non-magnetic SiO2 spacers in the nanometer range. The obtained MNP/SiO2 mixtures were characterized and used to build up measurement phantoms by embedding the mixture into a long-term stable polymer matrix. In MPS and MPI measurements it was found that reduction of the magnetic interactions encompassed by increasing the MNP distances leads for the tested tracer system to a weaker decrease of higher harmonics in the MPS spectra after immobilization of the particles and thereby, a higher spatial MPI resolution can be achieved.

Self-care and improving one’s well-being has been growing rapidly in recent years for manifold reasons (e.g. higher workload, corona pandemic). Consumer-grade noninvasive stimulation devices are therefore on the rise to counteract the occurrence of mood disorders and burn-out symptoms. Here, we aim at investigating the impact of dynamically varying auditory-visual stimulation patterns on neural entrainment patterns and resonance phenomena. Twenty-two healthy volunteers (11 female, 25.4 ± 5.1 years, one dropout, seven in control group) participated in the study. EEG data (64 channel; equidistant layout) were acquired preand during stimulation for each volunteer. Visual and auditory stimuli were presented via a headset (ATUM, Neuro- Bright; https://www.neurobright.co.uk/). Presentation patterns (frequency, intensity, spatial distribution) varied within a presentation session but were kept constant across all volunteers. Stimulus intensity was adjusted to individual comfort levels. Individual alpha peak frequencies (iAPF) were calculated via the power spectral density with 50% overlapping 10s epochs from pre-stimulation segments. For both, the study and the control group, a time-frequency representation was calculated for the pre- and during-stimulation segments. From this, power values were determined for different frequency-bands (iAPF, stimulation frequencies and second harmonics of the latter). Statistical analyses focused on contrasting the power values between pre- and during stimulation. Mean iAPF values were 10.25 ± 0.99Hz for the study and 10.63 ± 1.21Hz for the control group respectively. Both, power values at the stimulation frequencies and their second harmonics differed significantly between pre- and during stimulation (p stim =0.001; p harm =0.001) in the study group. No such difference was found for the control group (pstim=0.352; pharm=0.237). Further, neither the study nor the control group showed significant iAPF power differences (p study =0.035; p control =0.352; alpha*=0.008). Our results suggest that lightweight, portable auditory-visual presentation devices represent an effective tool for generating entrainment and resonance effects at home. Further analyses will focus on the investigation of individual differences driving such modulatory effects.

Electroencephalography (EEG) is widely used in clinical applications and basic research. Dry EEG opened the application area to new fields like self-application during gaming and neurofeedback. While recording, the signals are always affected by artefacts. Manual detection of bad channels is the gold standard in both gel-based and dry EEG but is timeconsuming. We propose a simple and robust method for automatic bad channel detection in EEG. Our method is based on the iterative calculation of standard deviations for each channel. Statistical measures of these standard deviations serve as indications for bad channel detection. We compare the new method to the results obtained from the manually identified bad channels for EEG recordings. We analysed EEG signals during resting state with eyes closed and datasets with head movement. The results showed an accuracy of 99.69 % for both gel-based and dry EEG for resting state EEG. The accuracy of our new method is 99.38 % for datasets with the head movement for both setups. There was no significant difference between the manual gold standard of bad channel identification and our iterative standard deviation method. Therefore, the proposed iterative standard deviation method can be used for bad channel detection in resting state and movement EEG recordings.

The efficient estimation of an approximate model order is essential for applications with multidimensional data if the observed low-rank data is corrupted by additive noise. Certain signal processing applications such as biomedical studies, where the data are collected simultaneously through heterogeneous sensors, share some common features, i.e., coupled factors among multiple tensors. The exploitation of this coupling can lead to a better model order estimation, especially in case of low SNRs. In this paper, we extend the rank estimation techniques, designed for a single tensor, to noise-corrupted coupled low-rank tensors that share one of their factor matrices. To this end, we consider the joint effect of the global eigenvalues (calculated from the coupled HOSVD) and exploit the exponential behavior of the resulting coupled global eigenvalues. We show that the proposed method outperforms the classical criteria and can be successfully applied to EEG, MEG Magnetometer, and Gradiometer measurements. Our real data simulation results show that the estimated rank is highly reliable in terms of dominant components extraction.

Recent research has shown that the joint processing of simultaneously recorded EEG-MEG signals can be beneficial compared to the separate analysis of data if the corresponding tensors have one of their factor matrices in common. In this study, we perform a joint CP decomposition of simultaneously recorded EEG and MEG magnetometer as well as gradiometer measurements via a new extension of the coupled SEmi-Algebraic framework for the approximate CP decomposition via SImultaneous matrix diagonalization (SECSI) to jointly factorize several tensors with at least one factor matrix in common. In case of the biomedical data, the four measured tensors have the factor matrix in the frequency domain in common. The developed coupled SECSI algorithm allows extracting the signal sources even in ill-conditioned scenarios. In the final processing step, coupled SECSI chooses the best out of a variety of possible factor matrix estimates, i.e., a combination that leads to the smallest value of the reconstruction error. However, for data tensors that do not exhibit significant coupling, the uncoupled solution can also be selected. The results are evaluated on measured as well as on synthetic data and compared with other state-of-the-art approaches.

The design of a multistatic ultra-wideband (UWB) microwave radar imaging system for breast cancer detection is presented in this paper. Based on pseudo-noise code (M-sequence) technology, this multiple-input-multiple-output (MIMO) system includes 24 antennas and records data from 12S channels. Furthermore, it realizes complete rotation of the antenna array around the breast, which allows several options of clutter removal and differential imaging, respectively. The system underwent a first feasibility study on volunteers at University Hospital Jena. The procedure of examination and first experiences are described in this paper.

Dry electrodes enable a shorter preparation time for infant EEG. Since infant skin is more sensitive than adult skin, soft electrodes are required to reduce the mechanical stress for this sensitive skin. Thus, soft electrodes are crucial for eventual repetitive and long-term use like in neonatal intensive care units. A biocompatible polyurethane (PU) can be produced in low hardness resulting in a soft and flexible electrode substrate. Silver/silver chloride (Ag/AgCl) electroless plating provides a conductive, electrochemically stable coating but the process may alter the mechanical properties of the electrode substrate. In this study, we assess the hardness of PU material before and after Ag/AgCl plating. The test sample design for Shore hardness measurement is based on ISO 7619-1:2010. Sample production consists of a 3D print master model, silicone molding, PU casting, and finally electroless plating. UPX 8400-1 (Sika AG, Switzerland) is used for the sample substrates. Test samples are produced with 7 different Shore hardness (range A40-A95) and 14 samples (each hardness: 1 uncoated and 1 coated). The hardness measurements are carried out with a lever-operated test stand Shore hardness tester model with a digital hardness tester (TI-AC with HDA 100-1, KERN & SOHN GmbH, Germany). It is shown that there is a hardness increase (Shore A) due to Ag/AgCl coating with a grand average of 1.1±0.7 (p<0.05). The largest increase of 2.1±0.2 is seen on the initial lowest Shore hardness sample (Shore hardness: 43.4±0.1). The absolute increase of hardness due to the Ag/AgCl coating decreases with increasing substrate hardness. It is concluded that there is no strong hardness increase of PU substrates due to Ag/AgCl plating. Therefore, the material is suitable as a soft electrode for repetitive and long-term use in infant applications.

Robust point cloud registration in real-time is an important prerequisite for many mapping and localization algorithms. Traditional methods like ICP tend to fail without good initialization, insufficient overlap or in the presence of dynamic objects. Modern deep learning based registration approaches present much better results, but suffer from a heavy runtime. We overcome these drawbacks by introducing StickyPillars, a fast, accurate and extremely robust deep middle-end 3D feature matching method on point clouds. It uses graph neural networks and performs context aggregation on sparse 3D key-points with the aid of transformer based multi-head self and cross-attention. The network output is used as the cost for an optimal transport problem whose solution yields the final matching probabilities. The system does not rely on hand crafted feature descriptors or heuristic matching strategies. We present state-of-art art accuracy results on the registration problem demonstrated on the KITTI dataset while being four times faster then leading deep methods. Furthermore, we integrate our matching system into a LiDAR odometry pipeline yielding most accurate results on the KITTI odometry dataset. Finally, we demonstrate robustness on KITTI odometry. Our method remains stable in accuracy where state-of-the-art procedures fail on frame drops and higher speeds.

With a limited number of workers or staff in the hospital, it is not possible to manually monitor all of the medical device in the hospital. Many medical devices were lost by mistake, many assets went unused because they were not well-stocked, and many assets were destroyed without recognizing it. This will undoubtedly be very negative to hospitals in terms of resources, which are often expensive, and will, of course, diminish the effectiveness and efficiency of medical services. The necessity for hospitals to modernize their technology is apparent. The rapid advancement of technology allows us to overcome these issues, in fact, IoT-based technology is now so advanced that paper-based technology must be gradually phased out. The technology is a real-time location system (RTLS), there are many different ways to implement this technology, one of them is to use Ultra-Wide band (UWB) technology, with this solution, hospitals can track the location of their medical device, as well as other information - this including the Transesophageal Echo (TEE) Probe. DWM1001 is one of the UWB modules that researchers can develop, but its deployment in hospitals still need more research and reliability. This study will address techniques for improving the reliability of anchor mapping and hybrid Wi-Fi solutions as backup solutions.

State-of-the-art self-supervised learning approaches for monocular depth estimation usually suffer from scale ambiguity. They do not generalize well when applied on distance estimation for complex projection models such as in fisheye and omnidirectional cameras. This paper introduces a novel multi-task learning strategy to improve self-supervised monocular distance estimation on fisheye and pinhole camera images. Our contribution to this work is threefold: Firstly, we introduce a novel distance estimation network architecture using a self-attention based encoder coupled with robust semantic feature guidance to the decoder that can be trained in a one-stage fashion. Secondly, we integrate a generalized robust loss function, which improves performance significantly while removing the need for hyperparameter tuning with the reprojection loss. Finally, we reduce the artifacts caused by dynamic objects violating static world assumptions using a semantic masking strategy. We significantly improve upon the RMSE of previous work on fisheye by 25% reduction in RMSE. As there is little work on fisheye cameras, we evaluated the proposed method on KITTI using a pinhole model. We achieved state-of-the-art performance among self-supervised methods without requiring an external scale estimation.

The temperature dependence of the dielectric parameters of tissues and tissue-mimicking phantoms is very important for non-invasive temperature measurement in medical applications using microwaves. We performed measurements of this dependence in the temperature range of 25-50&ring;C using distilled water as a reference liquid commonly used in dielectric property studies. The results were compared with the literature model in the frequency range of 150-3000 MHz. Using this method, the temperature dependence of dielectric parameters of a new muscle tissue-mimicking phantom based on agar, polyethylene powder, and polysaccharide material TX-151 was measured in the temperature range of 25-50&ring;C. The temperature dependence of the dielectric properties of this new muscle phantom was fitted to that of the two-pole Cole-Cole model and the deviation of the results between measured and modeled data was quantified.

Monitoring brain activity at home using electroencephalography (EEG) is an increasing trend for both medical and non-medical applications. Gel-based electrodes are not suitable due to the gel application requiring extensive preparation and cleaning support for the patient or user. Dry electrodes can be applied without prior preparation by the patient or user. We investigate and compare two dry electrode headbands for EEG acquisition: a novel hybrid dual-textile headband comprising multipin and multiwave electrodes and a neoprene-based headband comprising hydrogel and spidershaped electrodes. We compare the headbands and electrodes in terms of electrode-skin impedance, comfort, electrode offset potential and EEG signal quality. We did not observe considerable differences in the power spectral density of EEG recordings. However, the hydrogel electrodes showed considerably increased impedances and offset potentials, limiting their compatibility with many EEG amplifiers. The hydrogel and spider-shaped electrodes required increased adduction, resulting in a lower wearing comfort throughout the application time compared to the novel headband comprising multipin and multiwave electrodes.

Acoustic or visual warning signals for workers in hazardous situations might fail under loud and/or lowvisibility work situations. A warning system should be developed that uses electrocutaneous stimulation through textile electrodes. Previous work investigated suitable stimulation parameters using TENS electrodes. The aim of this study was to compare TENS and textile cuff electrodes in terms of sensation thresholds, qualitative and spatial sensation. A study on 30 healthy volunteers (f=13, m=17) of mean age of 26.7 years was conducted applying bi-phasic rectangular current pulses to electrodes attached to the upper right arm. The study revealed that perception, attention and intolerance thresholds, qualitative and spatial perception are comparable indicating that future studies with the textile cuff electrodes can be generally based on the previous results with TENS electrodes.

The efficient estimation of an approximate model order is very important for real applications with multi-dimensional data if the observed low rank data is corrupted by additive noise. In this paper, we present a novel robust method for model order estimation of multi-dimensional data based on the LineAr Regression of Global Eigenvalues (LaRGE). The LaRGE method uses the multi-linear singular values obtained from the HOSVD of the measurement tensor to construct global eigenvalues. In contrast to the Modified Exponential Test (EFT) that also exploits the approximate exponential profile of the noise eigenvalues, LaRGE does not require the calculation of the probability of false alarm. Therefore, it is well suited for the analysis of biomedical data. The excellent performance of the LaRGE method is illustrated via simulations and results obtained from EEG as well as MEG recordings.

Many applications in chemistry, biology and medicine use microfluidic devices to separate, detect and analyze samples on a miniaturized size-level. Fluid flows evolving in channels of only several tens to hundreds of micrometers in size are often of a 3D nature, affecting the tailored transport of cells and particles. To analyze flow phenomena and local distributions of particles within those channels, astigmatic particle tracking velocimetry (APTV) has become a valuable tool, on condition that basic requirements like low optical aberrations and particles with a very narrow size distribution are fulfilled. Making use of the progress made in the field of machine vision, deep neural networks may help to overcome these limiting requirements, opening new fields of applications for APTV and allowing them to be used by nonexpert users. To qualify the use of a cascaded deep convolutional neural network (CNN) for particle detection and position regression, a detailed investigation was carried out starting from artificial particle images with known ground truth to real flow measurements inside a microchannel, using particles with uni- and bimodal size distributions. In the case of monodisperse particles, the mean absolute error and standard deviation of particle depth-position of less than and about 1 [my]m were determined, employing the deep neural network and the classical evaluation method based on the minimum Euclidean distance approach. While these values apply to all particle size distributions using the neural network, they continuously increase towards the margins of the measurement volume of about one order of magnitude for the classical method, if nonmonodisperse particles are used. Nevertheless, limiting the depth of measurement volume in between the two focal points of APTV, reliable flow measurements with low uncertainty are also possible with the classical evaluation method and polydisperse tracer particles. The results of the flow measurements presented herein confirm this finding. The source code of the deep neural network used here is available on https://github.com/SECSY-Group/DNN-APTV.

Die Computertomographie macht einen relativ geringen Anteil an der Gesamtzahl der Röntgenuntersuchungen aus, trägt aber mittlerweile über 50% zur kollektiven effektiven Dosis der Bevölkerung bei. Dieses Missverhältnis ist aus Sicht des Strahlenschutzes kritisch zu betrachten und macht Maßnahmen zur Dosisreduktion erforderlich. Demgegenüber stehen Entwicklungen in der CT-Gerätetechnik, die theoretisch zu einer starken Dosiserhöhung beitragen. Hier ist vor allem die Zunahme der Anzahl der Detektorzeilen und die gleichzeitige Reduzierung der Schichtdicke an neueren CT-Geräten zu nennen. Vor diesem Hintergrund wurde im Rahmen einer Diplomarbeit (TU Ilmenau) ein Vergleich der Strahlenexposition an älteren und neueren CT-Geräten im Raum Thüringen durchgeführt. Dabei wurde zunächst die Exposition für Untersuchungsprotokolle aus der klinischen Routine betrachtet. Als Dosisgröße wurde der Volumen-CTDI verwendet. Dieser wurde auf drei Arten ermittelt. Zum einen wurde der Volumen-CTDI auf Grundlage von Phantommessungen bestimmt. Außerdem wurde der Volumen-CTDI dem Dosisprotokoll des CT-Systems entnommen. Des Weiteren wurde die Dosis mit Hilfe des Dosisberechnungsprogramms CT-Expo berechnet. Die Ergebnisse der drei genannten Verfahren zur Dosisbestimmung wurden verglichen. Anhand der Ergebnisse wurde schließlich überprüft, ob ein Zusammenhang zwischen der Höhe der Dosis und der Aufnahmeschichtdicke zu erkennen ist. Für einen objektiven Gerätevergleich ist neben der Dosis auch die Bildqualität zu berücksichtigen, die sich unter anderem durch das Bildrauschen beschreiben lässt. Für diesen Vergleich wurde ein Untersuchungsprotokoll verwendet, dessen Parameter sich gerätespezifisch nur geringfügig an den einzelnen CT-Geräten unterscheiden. Für dieses Vergleichsprotokoll wurde die Dosis ermittelt, die an jedem Gerät notwendig war, um ein identisches Rauschniveau an allen sieben betrachteten Geräten zu realisieren. Die Dosiswerte wurden wiederum hinsichtlich eines Zusammenhangs zur Schichtdicke überprüft. Aufgrund der Ergebnisse konnte kein Trend festgestellt werden, dass die Strahlenexposition an neueren CT-Geräten mit dünneren Aufnahmeschichtdicken zunimmt. Dosiswerte, die auf die drei genannten Arten ermittelt wurden, stimmten überwiegend überein.

Inverse problems, such as in bioelectric and biomagnetic source reconstruction procedures, are computationally complex and don’t have unique solutions. Consequently, validation is required not only for newly developed inverse algorithms, but also for new combinations of existing data preprocessing and data analysis procedures. Validation approaches in general include simulations, phantom measurements, and animal measurements. In this talk, electric and magnetic phantom measurements are presented, which aimed at the comparison and the estimation of the procedural limits of the source localization accuracy under real world conditions.

Synchronization in the electroencephalogram, which can be quantified by time-variant coherence measures, reveals a communication between brain regions reflecting different functions (e.g. learning or memory). Aim of this work is to compare three time-variant coherence algorithms in single-trial. The first algorithm is based on the short-time Fourier transform, the second one on the adaptive discrete Fourier transform, whereas the third algorithm is based on a recursive smoothed pseudo Wigner-distribution. Parameters which describe the dynamic properties of the algorithms were calculated and compared on simulated data. Differences were observed in the estimation performance of the algorithms. Based on the obtained results, the algorithms are discussed for their utilization in future applications. - Keywords- Time-variant coherence, short-time Fourier transform, adaptive discrete Fourier transform, pseudo Wigner-distribution

For early detection of eye dysfunctions which are caused by glaucoma, optic neuropathy and other forms of ophthalmic and neurologic pathology a selective cone excitation on biological level is necessary. By the use of a RGB-stimulator which enables a free choice of stimulation-colour and pattern it is possible to isolate the specific cone type responses. The right choice of a practical stimulator is very important for a high efficiency of the cone excitation. In our investigations we used two selection criteria - the spectral properties inclusive the effects on the biological level and the dynamic behaviour at the screen change for a stimulation sequence. For an easy handling with the stimulator and for a simple realisation of stimulation sequences, the tested RGB-systems are commercially available. After the investigation of the spectral and dynamic properties we could find a suitable cone stimulator.

In the health service a comprehensive instruction for each illness, which illness needs which diagnostics and which therapies, exists. The acute coronary syndrome represents such a clinical process. In the available paper this clinical process from medical view is arranged. In order to be able to model afterwards this clinical process, the necessary process structures must be put on such as systems, modules or primitives in using the MLDesigner. Subsequently, a model can be produced from the individual structures. The compiled model is evaluated in the connection according to criteria of modelling. Friendliness, correctness, reusability by model elements or the possibility of structuring are estimated. The developed simulation model can be used after completion for the analysis of the process of the acute coronary syndrome, in order to determine bottlenecks and dependence between individual process steps. Subsequently, it applies to find out by resources increase in the individual bottlenecks, which increase is useful for the minimization of the total waiting period of the patient.

Purpose: Recent surveys show that nearly 50% of the epileptic patients are likely to be aware of problems with cognitive function (International Bureau for Epilepsy. Patients in Mind Survey of Cognitive Function, 2004). The current contribution is a part of a study aiming at investigating the correspondence between objective psychological and electrophysiological features characterising such impairments. - Methods: Two different test batteries of diagnostic methods were designed: a combination of neuropsychological tests and a combination of electrophysiological cognitive stimulation paradigms. Both were applied in a group of patients with focal, and focal secondary generalized epilepsy (n=10, age=39.9±11.7) and a group of neurologically healthy controls (n=15, age=37.3±11.4). Here, the statistical parametric (t-Test, (TT)) and nonparametric (Wilcoxon-Test, (WT)) analysis of selected amnestic features from the Berlin Amnesia Test (BAT) and the Diagnostic Cerebral Damage Test (DCS) are focused on. - Results: The comparison between the groups shows that firstly, there is no difference between the features recall unstructured&hardcy; (RU) and recall associative pattern&hardcy; (RA); secondly, the feature recall semantic structure&hardcy; (RS) (=0.05,TT), recall and destructor&hardcy; (RD) (=0.10,TT), short time memoryspan&hardcy;(SM) (=0.05,WT), semantic interference&hardcy; (SI) (=0.05,TT) are significantly different and the DCS index of lability&hardcy; (LI) (=0.05,WT) is significantly higher in the patient group. - Conclusions: The results entail that there are no differences between verbal and figural working memory. The conspicuousness in the RD, ST and SI as well as the higher LI, point out a presumably weak memory trace. The lower value of RS indicates possible deficits in strategies of the storage process.

Purpose: To investigate the changes in the instantaneous heart rate (IHR), as a marker of the activity of autonomous nervous system, during hyperventilation in a group of refractory epilepsy patients and a group of healthy controls. - Methods: The electrocardiogram (ECG) signals were acquired in the patient group (n=6, 4 females, 2 males; focal or multi-focal epilepsy diagnosis, epilepsy history ≥ 21 years) during the clinical electroencephalogram (EEG) measurements of which hyperventilation (3 minutes) was a part. The recording was continued another 3 minutes after the activation method. The procedure was repeated in the control group (n=6, 4 females, 2 males). The IHR values were computed from the pre-processed ECG signals as a function of time. The percentage normalized ratio of the average IHR within the hyperventilation and post-hyperventilation intervals is assigned as an index of changes in the IHR. - Results: None of the patients were observed to have epileptic activity in EEG during the measurements. The results of ECG analysis show that a) the increase and the decrease in the IHR have a non-linear character; b) the index assigned is lower in the epilepsy group (7.5±3.6) than in the control group (22.5±9.4). - Conclusions: In a previous study we had determined differences between patients and controls in the EEG/DC-shifts accompanying hyperventilation [Kirlangic ME et al. Epilepsia 2003; 44: Suppl. 8: 142). Although no epileptic discharges were observed in EEG, a difference was also determined in the changes in the IHR, in this study. These differences can be exploited for epilepsy diagnostics and therapy evaluation.

Vorraussetzung für die computergestützte Analyse von Bildfolgen des Augenhintergrundes ist eine Kompensation der Augenbewegung. Für die Lösung dieser Aufgabe stehen vielfältige Möglichkeiten zur Verfügung. An diese Algorithmen werden hohe Anforderungen, wie Genauigkeit, Schnelligkeit und Robustheit gegenüber schwierigen Bildbedingungen gestellt. Eine Möglichkeit diese Aufgaben zu lösen besteht in der Verwendung von kantenlistenbasierten Templatematchingverfahren. Da die Augenhintergrundbilder verzerrt sind, werden mehrere kleine Templates verwendet, um so Verzerrungen im Bild ausgleichen zu können. Das Ergebnis der Templatematchingalgorithmen hängt jedoch von der Wahl der verwendeten Templates ab. - Im folgenden wird eine Methode zur Extraktion von Templates unter Verwendung eines kantenlistenbasierten Templatematchingalgortihmus vorgestellt.

This paper presents a shift reducing algorithm of fundus images consisting of the following steps: edge detection by a rule based gradient algorithm, algorithm to search significant templates and a template matching algorithm based on edge lists. The result of the shift reducing algorithm depends on the capability of the edge detection algorithm and the detected templates. The shift reducing algorithm works well and effectively.