Search

Workstations for health care professionals provide access to distributed healthcare information systems. They are complex systems that have to manage, in a unified framework, the distributed and multimedia information needed for optimal patient care. The objective is to provide the end-user with a virtual and unified device that conceals the complexity of the underlying information system. Integration and reuse of modular software components are important in the development of such a workstation. A reference set of evaluation criteria is proposed that includes functional, technical, organizational, medical, cultural and ethical, economic and industrial components.

The analysis of medical narratives and the generation of natural language expressions are strongly dependent on the existence of an adequate representation language. Such a language has to be expressive enough in order to handle the complexity of human reasoning in the domain. Sowa's Conceptual Graphs (CG) are an answer, and this paper presents a multilingual implementation, using French, English and German. Current developments demonstrate the feasibility of an approach to natural Language Understanding where semantic aspects are dominant, in contrast to syntax driven methods. The basic idea is to aggregate blocks of words according to semantic compatibility rules, following a method called Proximity Processing. The CG representation is gradually built, starting from single words in a semantic lexicon, to finally give a complete representation of the sentence under the form of a single CG. The process is dependent on specific rules of the medical domain, and for this reason is largely controlled by the declarative knowledge of the medical Linguistic Knowledge Base.

This paper describes the approach of the European HELIOS project to integrate image processing tools into ward information systems. The image processing tools are the result of the basic research in image analysis in the Department Medical and Biological Informatics at the German Cancer Research Center. Tools for the analysis of 2-dimensional images and 3-dimensional data volumes with 3-dimensional reconstruction and visualization are part of the Image Related Services of HELIOS.

For hypertension management, it is increasingly recommended to individualize treatment according to the individual patient's characteristics. However, the ability of physicians to predict the patient's risk in the absence of a computer aid is unknown. Our aim was to determine if senior hypertension specialists were able to assess reproducibly cardiovascular disease (cvd) risk in hypertensive patients. The records of 100 patients were selected at random from the ARTEMIS database. The group comprised 53% of males, mean age was 43.7 (15), blood pressure 160 (25) / 99 (12) mmhg and serum cholesterol 5.6 (1.1) mmol/l. The records were summarized on a standardized form. Nineteen items were devoted to general characteristics and medical history of patients, 15 to clinical examination and 7 to biological tests. These items have been chosen because they were used in the framingham cvd risk prediction equation and/or had decisional value in a bayesian program designed for secondary hypertension diagnosis. Six physicians, twice at one month intervals, assigned each patient to one of the three categories of risk (low, moderate, high). The order of cases was randomized. Chance corrected agreement (intra-physician agreement) between the two estimations was only moderate except for one expert (kappa values: 0.658, 0,543, 0.439, 0.363, 0.328 and 0.207 for each expert, respectively). The proportion of patients assigned by each physician to each category ranged from 22.7% to 77.3% for the low risk, and from 4.5% to 43.9% for the high risk category. One physician overestimated risk, one underestimated it and the three others were in keeping with framingham predictions. Bias and lack of reproducibility influence physician judgment in assessing cvd risk of hypertensive patients. Application of computer aids could result in more rational assessments and better decision making.

In the context of health care information systems based on client/server architecture, we address the problem of a common Computer-based Patient Record (CPR). We define it as a collection of faithful observations about patients care, with respect to the free expression of physicians. This CPR model supports several views of the medical data, in order to provide applications with a comprehensive and standardized access to distributed patient data. Finally, we validated our CPR approach as a primary data model server for an application for hypertensive patient management.

Medical language constitutes a large subset of human language. However, it presents special features which have to be taken into consideration for natural language analysis. We have observed that natural language texts dealing with a specific medical area (especially discharge summaries and reports) share a common vocabulary and common habits of word usage. Facing these specificities, we describe in this paper an original system called RECIT, designed for medical text analysis and understanding. In particular, we discuss the principles which guide the analysis of medical jargon using semantic considerations, coupled with syntactic information when needed. As a result of the flexibility of the system, a medical application developed in the context of the AIM project HELIOS-2 is presented.

This paper describes a Development Environment for medical applications developers coupling Reuse and Object-Oriented technologies. It presents the two main phases of the reuse process, the reusable component creation and the reusable component use, and stresses on the very nature of reusable development components. We provide both a classification model for those reusable components that is a support for an efficient and simple retrieval mechanism, and a reusable component model that considers the same way components of large and low-granularity levels (e.g., application frameworks and methods of a given class). The reusable component model is based on the underlying idea that development components are not operational components since they do not have the same purpose. This model first represents each component as an aggregation of other sub-components, together with all information needed for its retrieval (e.g., classification and facets), its understanding (e.g., informal functional description), and its dependencies with other development components, in particular the applications in which it occurs, since applications are considered here as abstract development components.

In the medical domain, new developments commonly rely on client/server architectures. But face to distributed environments, the software developers encounter a tremendously increasing complexity when building integrated applications. This paper presents the HELIOS Unification Bus (HUB), a communication integration framework for the HELIOS medical software engineering environment that allows the exchange of data between components that can be hosted on heterogeneous machines linked by a network. The HUB is developed as a C++ toolbox over UNIX and TCP/IP. It includes a message routing entity called router and a generic application programming interface (API), implemented as a C++ library, that allows to build easily software components compliant with the standardised HELIOS language. Messages conveyed by the bus are composite objects that are serialized to be transmitted over the bus using the ASN.1 ISO presentation protocol. The article describes the use of the bus to ease the development and execution of distributed medical applications and its role from the communication integration standpoint.

This paper presents the design and implementation of the HELIOS software component that deals with integration of medical applications in health information networks. The problem of interoperability between health information systems based on different data exchange syntaxes is first discussed. A meta-model, relying on CEN TC251 recommendations, is then presented as a possible solution to this problem and a message description language including these recommendations is proposed. Using this meta-model, the Medical Connection Services that comprises a generic message processing automaton, a resource manager and a mapper is able either to interpret messages expressed in a given syntax (e.g., EDIFACT, ASTM) and map them to the application objects or to automate the translation of the messages in another syntax. Special focus is given on the position of the Medical Connection Services within the HELIOS integration strategy (i.e., through data, presentation and communication). The problem of semantic heterogeneity is then discussed.

Information and knowledge processing technologies have contributed to the development of Computer assisted instruction (CAI) applications. Multimedia technologies will increase the possibilities of these programmes in the future. However the evaluation of the software as well as the other technologies is necessary. The authors have tried to quantify the contributions of CAI analysing the literature referenced in MEDLINE from 1989 to 1992. On the 470 papers indexed with the key-word "Computer Assisted Instruction", 56 studies take evaluation into account. Only 26, that is 5.5%, presented a clear and explicit methodology of evaluation. On the 26 studies, 16 were considered with positive result and among these, 6 presented a statistical significance. The conditions and the criteria of the evaluation studies do not permit a meta analysis. Is is necessary to develop standardized evaluation criteria to estimate the quantitative and qualitative impact of the CAI technics on the educational process and their own pedagogical contribution.

Interoperability is a key issue in distributed health information systems. This paper presents the design and implementation of the Connection Service (CS) component of the HELIOS medical software engineering environment. An exchange message meta-model based on CEN TC251 recommendations and a message description language are proposed. Using this meta-model, the CS is able either to interpret messages expressed in a given syntax (e.g., EDIFACT, ASTM) and map them to the application objects or to automate the translation of the messages in another syntax. It is concluded that this meta-model approach contributes to the management of semantic heterogeneity.

After a brief description of the main domains of application of medical informatics, a typology of questions to be assessed, and of the involved actors, is presented. Dimensions of technology assessment include techniques, medical and health efficacy, economics, sociology, and law and ethics. Barriers to evaluation are analysed. They include barriers related to the domain complexity, human motivation and methodological barriers. Finally categories of solutions and research areas are suggested.

The cardio-respiratory complications of sleep apnea syndrome have been prospectively assessed in 60 patients with massive obesity and free of chronic obstructive lung disease while the associated cardiovascular diseases and the alterations of pulmonary function were taken into account. These cardio-respiratory complications were observed only in patients with a number of apneas per hour of sleep greater than 20. The sleep apneas induced nocturnal hypoxemia that is frequently severe and independently correlated to the apnea index, diurnal hypoxemia and hypercapnia that are usually moderate, and presumably left ventricular hypertrophy that is not related to the development of daytime hypertension. However the nocturnal apneas were not associated with the development of an impairment of right or left ventricular function, or with the occurrence of cardiac arrhythmias or conduction disturbances. The absence of severe cardiac complications in this study may be related to the fact that the patients were relatively young and that the sleep apnea syndrome was diagnosed at an early stage of evolution. The findings of this study could help to define a more rationale approach in several therapeutic indications of sleep apnea syndrome.

This paper describes the integration of an existing American Hospital Information System with a European Image Processing System. Both systems were built independently (with no knowledge of each other), but on open systems standards. The easy integration of these systems demonstrates the major benefit of open standards-based software design.

This paper presents the architecture of the Information System of HELIOS, a medical Software Engineering Environment. It is an object oriented framework for the development of medical applications which puts particular emphasis on tools and techniques favouring reuse of previous work and enhancing collaboration between developers.

The objective of the HELIOS European project is to create a Software Engineering Environment (SEE) to alleviate the development of medical applications. The resulting system can be viewed as a layer above an operating system providing services to the application designer, and finally to the application end-user. HELIOS is constituted of a set of Software Elements communicating through an abstract channel called Software Bus. The SEE kernel is made of three main software elements, the Information System, the Documentation Facilities and a component which provides the application user with a convenient user support interface, the Interface Manager. The Services consist of a collection of tools providing the necessary services to the medical application designers including an image processing toolbox.

The connections and possible interactions between coffee consumption, tobacco consumption, blood pressure (BP), age, and body mass index (BMI) were analysed in a cross-sectional epidemiologic study of 1098 men and 393 women of the Algiers district. Systolic blood pressure (SBP) was negatively and significatively correlated with cigarette smoking. Diastolic blood pressure (DBP) was positively and significantly associated with coffee consumption (P less than 0.001) and negatively and significantly with cigarette consumption (P less than 0.001). These associations remained significant after multivariate analysis including sex, age, BMI, physical exertion at work, athletic activity, rural versus urban residency, tobacco, tea and coffee consumption. Coffee consumption and cigarette smoking were positively associated (P less than 0.001). In the subgroup of men, the association between coffee consumption and DBP was significant only after adjustment for cigarette smoking. In the subgroup of women, containing a low percentage of smokers, this association was significant without such adjustment. This finding suggests that cigarette consumption might partly mask the association between coffee consumption and BP. The contradictory interactions observed between BP, cigarette consumption and coffee consumption might help to explain the conflicting reports concerning the association between coffee consumption and ischemic heart disease.

Artificial intelligence techniques and expert systems are graduating from research laboratories to enter many human domains of activities. However, initial expectations of the seventies have been transformed with time to a more sober reality. Starting from examples in nephrology, this paper tries to give a balanced view of this new technology. Justification, indication and limitation of present expert systems are discussed as regards their possible goals. The respective roles of the experts and knowledge engineers are described. The need to integrate the artificial intelligence approach into present patient database management systems in order to build up expert database systems is evaluated.

The computerized medical specialized records covering the follow-up of hypertension (ARTEMIS), diabetes (MELLITEE), chronic renal failure and the minimum medical record of Nephrology (THESEE) using the temporal database management system LIED, have been developed in several units of University Hospitals. The authors show that these computerized records are usable in a Nephrology Unit in a General Hospital.

An integrated approach is described for the computerized management of a nephrology department. On a medical point of view, the system comprises a minimum medical record for every patient, different specialized records and knowledge bases presently covering hypertension, diabetes and chronic renal failure. From a technical point of view, the methodology used integrates data and knowledge management techniques. Various individual reports facilitate patient management. For hypertensive patients, an expert system is combined with the record system. The results of a preliminary evaluation are reported and future developments considered.