Ebook: Health Data in the Information Society

MIE 2002 is the XVIIth international conference of the European Federation of Medical Informatics. The conference is held twice in every three years since 1978. MIE 2002 is the first MIE conference of the 21^stcentury and of the third millennium. Today, mankind builds up the information society, enabled by the underlying rapid development in computer technology. The significance of the spread of the internet is comparable to the significance of Gutenberg’s invention. On one hand both of them help dissemination of data and knowledge and sharing of ideas. On the other hand the achievements may divide the society, as did non-literacy deprive many people from knowledge throughout centuries. Today millions of people are isolated from an incredibly large amount of information because of “computer non-literacy”, and a new elite mastering the information society has appeared. However, the ease of production and dissemination of information may foster thoughtless communication, and has lead to a flood of information and disinformation. We have to learn how to behave in this new situation, in which the dissemination of information – at an international level – is totally uncontrolled.

In the area of medical or health informatics these questions are more serious. Lack of information, false or inadequate information, as well as improper interpretation of accurate information may seriously harm patients. And the process may go out of the control of the physician, i.e. patients can “treat” themselves just by visiting some health sites on the net. Everybody may throw a message in a bottle in information flood, and everybody may pick up messages at any time. Can we do anything to ensure that all messages are valid? Can we guarantee that our messages reach the intended audience? Can we secure that content has not changed on its way? Do we know that people getting our messages will interpret them correctly? Are we able to understand the intention of a sender, when we get a message totally out of context? These questions build up the framework of MIE2002. The papers are organised into six chapters. The first chapter (“Open the bottle”) deals with the acquisition of information. The second (“Forward the bottle”) focuses on communication and security. The third chapter concentrates on data storage and retrieval (“Keep the bottle”). The fourth chapter collects papers about understanding of messages: data analysis, natural language processing, etc. The fifth chapter concentrates on the ultimate reason of obtaining information in health care: we have to solve problems. Finally the sixth chapter draws our attention to the context: the broader environment, in which we interpret information.

We express our special thanks to all who contributed to this Proceedings; first to the authors of the papers, then to the reviewers and the members of the Scientific Program Committee, and to Gabor P´lvölgyi for the technical editing.

Rolf Engelbrecht

György Surján

Peter McNair

Web Based Training (WBT) is of increasing importance [1][2]. In [5] the main author has presented and discussed an overall approach to WBT for medical image processing. In the present contribution the interactive part of the approach is discussed in detail and its prototypical realisation in the framework of a client server teaching application is presented.

Sonography is a widely used non-invasive diagnostic tool and its main advantage is low cost in comparison with other diagnostic methods such as immunological analyses. In this work it is presented the relation between a sonographic image of thyroid gland and an immunological status of the patients with Hashimoto's lymphocytic thyroiditis (chronic inflammation of the thyroid gland). The results, evaluated on a set of 740 B-mode sonographic images from 37 subjects, show that raw values of individual image pixels in sonogram of thyroid gland with presence and without presence of anti-thyroid antibodies are significantly different (means 31.87 and 44.56; standard deviations 8.6 and 11.82; t=3.4; p=0.0017) and that they can be used for the prediction of presence of anti-thyreoglobulin and anti-thyreoperoxidasis antibodies. This result suggests the possibility to use this method in clinical diagnostic process for reducing the costs. Also the correlation between the image features and the level of antibodies was examined. The highest correlation was found for inverse difference moment and the level of anti-thyreoperoxidasis (coefficient of determination 0.43).

An innovative algorithm for Magnetic Resonance Imaging (MRI) capable of demonstrating the source of various artefacts and driving the hardware and software acquisition process is presented. The algorithm is based on the application of the Bloch equations to the magnetization vector of each point of the simulated object, as requested by the instructions of the MRI pulse sequence. The collected raw data are then used to reconstruct the image of the object. The general structure of the algorithm makes it possible to simulate a great range of imaging situations in order to explain the nature of unwanted artefacts and to study new acquisition techniques. The way the algorithm structures the sequence has also allowed the easy implementation of MRI data acquisition on a commercial general-purpose DSP-based data acquisition board, thus facilitating the comparison between simulated and experimental results.

The aim of the study was to assess the potential use of a model of the colour content in retinal fundus images to screen for sight threatening retinopathy in diabetes. Diabetic retinopathy is the most frequent cause of blindness in the population of working age in industrialised countries, but efficient therapies do exist, and accurate and early diagnosis, and correct treatment can prevent blindness in more than 50% of all cases. However, up to 50% of cases of type 2 diabetes, which comprises 85-90% of all patients, are undiagnosed, with an average delay of 10 years between the onset of the condition and diagnosis. In an other study we have described how there is a linear relation between age and the colour composition of retinal images from non-diabetic subjects. In the present study this relation was compared to the colour composition of retinal images from diabetes patients. We found that for the patients in the present study there is a significant difference in the colour composition between normal subjects and diabetic subjects with retinopathy. Although the number of patients in our study is too small to allow any conclusion, we suggest that this difference potentially may be used as the basis for a simple screening method for sight threatening retinopathy in unrecognised diabetes, or potentially may help estimating the risk of developing diabetic late complications in newly diagnosed type 2 diabetes.

In this paper we propose a new algorithm for the detection of clustered microcalcifications using mathematical morphology and artificial neural networks. Considering each mammogram as a topographic representation, each microcalcification appears as elevation constituting a regional maxima. Morphological filters are applied, in order to remove noise and regional maxima that doesn't correspond to calcifications. Each suspicious object is marked using a binary image and finally a feed forward neural network classifies every object achieving a rate of 90% true positive detections with 0.11 false positives per image.

Mammography associated with clinical breast examination is the only effective method for mass breast screening. Microcalcifications are one of the primary signs for early detection of breast cancer. In this paper 1 we propose a new kernel method for classification of difficult-to-diagnose regions in mammographic images. It consists of a novel class of Markov Random Fields, using techniques developed within the context of statistical mechanics. This method is used for the classification of positive Region of Interest (ROI's) containing clustered microcalcifications and negative ROI's containing normal tissue. The obtained results show that the proposed approach can be successfully employed for detection of microcalcifications.

Possibility to obtain information about local changes of heart repolarization from body surface potentials was studied on a model. Activation-recovery intervals (ARI) in surface ECG leads were tested as indicators of changed repolarization in the underlying myocardium. ECG signals corresponding to activation of myocardium with normal and changed action potential duration were simulated on the surface of a realistic inhomogeneous torso. ARI intervals were derived from all ECG signals and displayed as surface ARI maps. Results suggest that shortening and prolongation of action potentials in anterior myocardial regions can be visible in corresponding areas on surface ARI maps while only prolongation in postero-lateral regions can be observed. Reproducibility of ARI maps was checked on real measurements using 63 and 192 surface ECG leads. Obtained ARI maps exhibited acceptable reproducibility with correlation of 0.73 to 0.87. Based on the model and experimental results it is hoped that ARI maps can give some insight into the myocardium repolarization and help to recognize tissue with changed properties, primarily in heart regions underlying the anterior chest.

One of the main difficulties in using body surface potential mapping (BSPM) techniques is the need of complicated multi-channel measuring system. In this paper practical portable ECG mapping system is introduced. The system consists of a notebook computer and a data acquisition system box connected to the computer by fast IEEE 1284 parallel interface working in ECP mode. Concept of the device enables to extend the basic 134-channel high-resolution multi-channel ECG amplifying unit up to 256 channels. Application software includes measurement and real time monitoring of ECG signals, computation and display of several types of body surface potential maps. System can be connected to hospital information networks and supply them with measured ECG data for advanced processing or central archiving.

The larvicide effects of three endecticide products were studied against house fly larvae. Repeated- measure design was used for the analysis of variance. The observed data followed different distributions. For this reason, instead of the repeated measure ANOVA, three fast statistical methods were applied simultaneously. First, the test for homogeneity was carried out for registered samples by using the Friedman test, and the Wilcoxon-Wilcox test was applied for pairwise comparisons. Secondly, an extended median test was applied for the same problem of test for homogeneity, and the Neményi test was used for pairwise comparisons. Thirdly, a 2×2 contingency table analysis was carried out, exploiting the specific feature of the database and informatical tools. The three simultaneous evaluations enabled us to come to sound, affirmative conclusions. In the course of veterinary-biological interpretation we demonstrated the statistically reliable temporal efficiency of the three products. We also presented the relative efficiency and recognizable features of the products. We highlighted the significantly prompt and long release larvicide effects of the drugs. We discussed and compared the information content of the applied statistical methods.

Considering resistances of supposed or identified pathogens is of vital importance in antibiotics therapy. Additionally local details play a very important role. So the resistance situation can be different on two wards of the same facility. Furthermore it is possible that there is spatial extension of resistances across the wards that the physicians suppose a future resistance development in a specific unit. The idea of RESIS3D is to visualize spatial and temporal development of resistant pathogens and to give the attending physician (as well as the laboratory) important information for using antibiotics. RESIS3D is an addition to our computer-based therapy adviser called ICONS for calculated antibiotics therapy in intensive care medicine.

A simple non expensive device for biosignal acquisition is presented. It mainly meets the requirements for didactic purposes specific in medical informatics laboratory classes. The system has two main types of devices: ‘student unit’- the simplest one, used during lessons on real signals and ‘demo unit’ which can be also used in medical practice or for collecting biological signals. It is able to record: optical pulse, sphygmogram, ECG (1-4 leads) EEG or EMG (1-4 channels). For didactical purposes it has a large scale of recording options: variable sampling rate, gain and filtering. It can also be used in tele-acquisition via Internet.

The way of investigation some fundamental properties of integrative brain activity was worked out. Within the framework of the qEEG system for computer diagnostics of EEG NeuroResearcher^™'2001 (Version 7.5) is created modules for investigation of multivariate neurodynamic systems of the brain on the base of multidimensional spectral EEG and ESCoG analysis. It is intended for doctors and neuroscientists who use mathematical methods (classical correlation and spectral analysis and newest methods of multidimensional analysis of neurodynamic systems of the brain) for quantitative EEG (qEEG) successfully.

We have recently shown, in studies with patients with Type 1 (insulin dependent) diabetes, that alcohol intake at 2100 hrs significantly reduced blood glucose values after 10-12 hours, compared with control studies with no alcohol. We hypothesised that this was due to the following effects of alcohol: 1. Alcohol metabolism increases NADH, leading to a reduction in hepatic gluconeogenesis; and 2. increased glycogen phosphorylase activity depletes hepatic glycogen stores; 3. After the alcohol is metabolised, hepatic insulin sensitivity is increased, leading to the restoration of glycogen stores and reduction in blood glucose levels, and 4. consequently, after several hours, glycogen stores and insulin sensitivity return to normal. A model describing these changes (DiasNet-Alcohol) was implemented into the DiasNet model of human glucose metabolism. Our study suggests that the DiasNet-Alcohol model gives a reasonable approximation of the effect of alcohol on blood glucose concentration observed in our study and supports our hypothesis for the mechanism behind these effects in Type 1 diabetes.

This study was designed to estimate the changes of different activities (going to bed, getting up, visiting washroom) of elderly as an index of mobility. The different values in time and number of the different activities were obtained from 16 elderly suffering from Alzheimer’s disease in a hospital for short period observation adding up 4 months of observation data. Significant agreement in different activities (going to bed, getting up, visiting washroom, wandering) among the two measurements by multisensor system and by nursing staff notation was found. As interesting results, the follow up of a patient is shown for 13 consecutive nights. These studies show the potential usefulness of assessment of night activities as an index of mobility, mirror of motor behaviour. Due to minor error, the precise mobility index could not be observed, motor behaviour could simply be evaluated.

People relying much on vision in the control of posture are known to have an elevated risk of falling. Dependence on visual control is an important parameter in the diagnosis of balance disorders. We have previously shown that virtual reality methods can be used to produce visual stimuli that affect balance, but suitable stimuli need to be found. In this study the effect of six different virtual reality stimuli on the balance of 22 healthy test subjects was evaluated using force platform posturography. According to the tests two of the stimuli have a significant effect on balance.

The present aim was to develop signal analysis methods to recognize muscle action potentials recorded with electromyography (EMG) from two facial muscle sites, corrugator supercilii (the muscle activated in frowning) and zygomaticus major (the muscle activated in smiling). Fourteen subjects produced volitional activations on both muscle sites, first, on a single muscle site activation basis and finally combinatory activations of both muscle sites. Wavelets and neural networks were used to analyse these voluntarily produced bursts of electric signals. Our results showed well over 90% recognition rates for all signal types. It is possible to utilize these communicative signals, for example, for multimodal human- computer interaction.

The computer model of ventricular activation was used to study the effects of eventual differences in the repolarization pattern between the right and the left ventricle, as well as between the apical and the basal parts of the ventricles. All changes in model action potential durations (APDs) were performed in the range corresponding to the APD variability measured in myocytes. The vectorcar-diographical spatial T loop was very sensitive on the changes in the right to left ventricular gradient of APD, while the similar changes in the apico-basal gradient of APD influenced the T loop minimally.

According to previous modeling studies - propagation of depolarizing wave fronts - consists of subintervals, each characterized by a smooth progression of waves through the myocardium. At the onset and end of these intervals, abrupt changes occur in the 3D pattern of activation waves (e.g. at the time of the collision of activation waves with other waves, obstacles or unexcitable myocardium and epicardial breakthrough), which manifest themselves in the surface ECG as “jumps” (high frequency notches and slurs). The timing of jumps provides diagnostic information on bioelectrical tissue properties of the heart. Findings of this study validated previous simulation results. Furthermore, essential signal processing requirements were formulated for a high-resolution body surface potential mapping technology.

In the last few years, Clinical Reasoning Learning has been used during clinical training courses in many universities. The goal of this educational activity is to develop problem solving and diagnosis skills in medicine, using previously acquired medical knowledge. We propose the integration of educational resources and computer systems to help students during the reasoning process, and in order to prepare students to look for medical information. We also describe a Computer-Supported Collaborative Learning Environment that integrates Information and Communication Technologies in order to improve the Clinical Reasoning Learning Sessions.

It is now common to access and to exchange data through computer networks. The major applications used are Web access and e-mail. But while there is no security concern when accessing classical or anonymous data, the exchange of sensitive data (e.g. patient-related information) is subject to legal constraints. This paper summarizes the current legal obligations in Switzerland and presents state of the art techniques for applying the legal rules. A commercial implementation in Switzerland is taken as illustration.

Trustworthy communication and co-operation requires in general static TTP services describing status and relation of communicating principals as well as their corresponding keys and attributes. Additional TTP services are needed to provide trustworthy information about dynamic issues of communication and co-operation such as message identifier, time and location of processes, workflow relations and system behaviour. Certified time stamps are important notary's functions dealing with the services mentioned above. Within the European RESHEN project, the ONCONET regional health network for improving cancer care has been moving from system-related to certified time stamps using a local secure time stamp server or deploying remote time stamp services provided by an accredited CA. After implementing such services at the beginning of 2002, an evaluation of the operation modes will be performed during 2002 to derive recommendations for optimisation of these important services in national and international health networks.