Co-occurrence based meta-analysis of scientific texts: retrieving biological relationships between genes.

MOTIVATION: The advent of high-throughput experiments in molecular biology creates a need for methods to efficiently extract and use information for large numbers of genes. Recently, the associative concept space (ACS) has been developed for the representation of information extracted from biomedical literature. The ACS is a Euclidean space in which thesaurus concepts are positioned and the distances between concepts indicates their relatedness. The ACS uses co-occurrence of concepts as a source of information. In this paper we evaluate how well the system can retrieve functionally related genes and we compare its performance with a simple gene co-occurrence method. RESULTS: To assess the performance of the ACS we composed a test set of five groups of functionally related genes. With the ACS good scores were obtained for four of the five groups. When compared to the gene co-occurrence method, the ACS is capable of revealing more functional biological relations and can achieve results with less literature available per gene. Hierarchical clustering was performed on the ACS output, as a potential aid to users, and was found to provide useful clusters. Our results suggest that the algorithm can be of value for researchers studying large numbers of genes. AVAILABILITY: The ACS program is available upon request from the authors.

Distribution of information in biomedical abstracts and full-text publications.

MOTIVATION: Full-text documents potentially hold more information than their abstracts, but require more resources for processing. We investigated the added value of full text over abstracts in terms of information content and occurrences of gene symbol--gene name combinations that can resolve gene-symbol ambiguity. RESULTS: We analyzed a set of 3902 biomedical full-text articles. Different keyword measures indicate that information density is highest in abstracts, but that the information coverage in full texts is much greater than in abstracts. Analysis of five different standard sections of articles shows that the highest information coverage is located in the results section. Still, 30-40% of the information mentioned in each section is unique to that section. Only 30% of the gene symbols in the abstract are accompanied by their corresponding names, and a further 8% of the gene names are found in the full text. In the full text, only 18% of the gene symbols are accompanied by their gene names.

Extensions to the HISA standard--the SynEx computing environment.

OBJECTIVES: This paper reports on the integration work required to support the delivery of healthcare. Specifically, four elements are included being: the DHE (a standards based distributed healthcare environment), Synapses which is a patient record server, XML as a technology to deliver records, and finally two client applications that facilitate structured data entry (SDE) and the remote booking of specialist services. METHODS: A general motivation for the work is presented and augmented on hand of a clinical scenario. Additionally, the adoption of a middleware approach is reviewed. The role of computerised patient records is described followed by an account of a federated record server. The approach favoured by standards bodies in utilising XML is covered and the tailoring to suit the needs of the integration is explained. The main practical challenges in achieving integration are presented, one of these being the mapping between the DHE data model and the Synapses server record architecture. RESULTS: The described environment has been demonstrated to provide the functionality that is required and in addition it has been shown that the engineering challenges can be met in a controlled and orderly manner. CONCLUSIONS: The role of the DHE middleware component acting as an 'anchor' has been shown to be a valid one onto which other specialised components can be added to provide a richer service environment. Additionally, it has been demonstrated that XML is a good candidate technology that facilitates connectivity to client applications over an extranet. The demands inherent in both the clinical scenario and the computerised patient record can be met by the computing environment described in the paper.

Facilitating networks of information.

In this paper we describe an approach to respond to a request for information with the identification and location of the appropriate person as a source of information for answering the question. The expertise of a person is characterized using a weighted profile that has been derived from a series of documents describing the expert's activities. Having these profiles, requests for information can be matched with these profiles. The best matches correspond with the people that are experts for providing information on the request.

UMLS-based access to CPR data. Unified Medical Language Systems.

This paper describes the results of a project that explores the use the Unified Medical Language System (UMLS) for knowledge-driven tasks, such as browsing a computer-based patient record (CPR). The project consisted of a number of steps: the mapping between CPR terms and UMLS concepts, the development of an algorithm that explores the CPR data using this mapping and the implementation of a first prototype browser that visualizes 'found' data. A second task addressed in this project has been the direct access to online medical literature (MEDLINE) using the UMLS concepts found in the CPR data. In this project, we used a preliminary version of the Open Records for Patient Care (ORCA) CPR that consisted only of the history and physical examination data of patient suffering from heart failure.

ORCA: the versatile CPR.

The introduction of computer-based patient records (CPRs) that fully replace paper records proves especially difficult in specialized care, despite the potential advantages of CPRs for patient care and research. Improved data legibility, availability, sharing of records, and decision support may directly benefit patient care. Barriers to the introduction of CPR applications at institutions may be caused by lack of infrastructure, or by financial or organizational issues. To have clinicians interactively enter data at the point of care is still a big challenge. This paper presents an overview of ORCA (Open Record of CAre): a generic CPR, designed for integration with existing systems, presentation of multi-media patient data, and the collection of structured data, directly by clinicians. ORCA can easily be tailored to the needs of a variety of medical specialists without the need for changes to its data model, functionality, or interface. The paper describes the essence of the architecture of ORCA and the user benefits with emphasis on the support of structured data entry.

UMLS-based access to CPR data.

This paper describes the results of a project that explores the use the Unified Medical Language System (UMLS) for knowledge-driven tasks, such as browsing a computer-based patient record (CPR). The project consisted of a number of steps: the mapping between CPR terms and UMLS concepts, the development of an algorithm that explores the CPR data using this mapping, and the implementation of a first prototype browser that visualizes "found" data. A second task addressed in this project has been the direct access to online medical literature (MEDLINE) using the UMLS concepts found in the CPR data. In this project, we used a preliminary version of the Open Records for Patient Care (ORCA) CPR that consisted only of the history and physical examination data of patient suffering from heart failure.

Clinical data entry.

Routine capture of patient data for a computer-based patient record system remains a subject of study. Time constraints that require fast data entry and maximal expression power are in favor of free text data entry. However, using patient data directly for decision support systems, for quality assessment, etc. requires structured data entry, which appears to be more tedious and time consuming. In this paper, a prototype clinical data entry application is described that combines free text and structured data entry in one single application and allows clinicians to smoothly switch between these two different input styles. A knowledge base involving a semantic network of clinical data entry terms and their properties and relationships is used by this application to support structured data entry. From structured data, sentences are generated and shown in a text processor together with the free text. This presentation metaphor allows for easy integrated presentation of structured data and free text.

Can data representation and interface demands be reconciled? Approach in ORCA.

Research in the domain of computer-based patient records had always faced the conflicting demands of efficiency for the practicing physician and suitability of the record contents for data analysis in view of decision support, research, and quality assessment. Interface and contents pose different demands on the data model underlying the record. The challenge is to combine the most suitable model for data representation with the interface that best fits the clinical setting. ORCA (Open Record for CAre) provides a solution by making the distinction between domain dependent and domain independent data and letting domain dependence be decisive for the choice of model. Interactive definition of custom-views provides interface flexibility for domain dependent data. Views on domain independent data need not cope with the limitations of multiple table views in relational DBMSs. A standard set of single table queries can support recording of domain independent data, irrespective of the clinical setting.

Trends in Integrated Clinical Workstations.

During the last decade, several projects aiming at integrated clinical workstations have been described and several prototypes have been demonstrated. In most of these projects, the clinical workstation accesses information and functionality provided by the present proprietary legacy systems of health-care institutions. We discuss trends in integrated clinical workstations from the viewpoints of software engineering and integration, considering that the clinical workstation itself basically consists of three layers: a presentation layer, a data integration layer, and a communication layer. The software engineering view on clinical workstations focuses on the development of basic building blocks from which clinical workstations, specific to a particular medical domain, can be composed. The integration view on clinical workstations addresses methods and techniques to deal with the, in general, intrinsically closed information systems in health-care institutions.

ARIS: integrating multi-source data for research in andrology.

Although the concept of distributed systems for the storage of patient data is more and more commonly accepted, for some considerable time yet most patient data will be stored in centralized rather than departmental systems. An important advantage of storage in a central system is hospital-wide access to much of the patient data. Disadvantages are however that these data cannot be reviewed through one user interface, and that the structure of the data does not lend itself to exploitation for other purposes. We describe the implementation of an Andrology Research Information System in which these data are integrated in a well-structured database facilitating multiple views on the patient data through a graphical user interface, and clinical research, quality control and summary reports. The data can be analyzed directly using the Hermes workstation. In this way the strengths of the centralized system are combined with those of the dedicated ARIS system.

Problems with integrating legacy systems.

The economic and organizational impact of imposing state-of-the-art technology to the large number of proprietary legacy systems operational in most hospitals requires integrated clinical professional workstations to provide flexible encapsulation mechanisms for these systems rather than reengineering these systems to this new technology. In this paper the implications of different input/output and translation models of legacy systems for their integration into a clinical workstation is described. Examples of legacy systems that have been integrated in the HERMES clinical workstation are presented as examples of the range of difficulties one might encounter. The features that an integrated workstation should offer for integrating a broad range of legacy systems are also addressed in this paper.

A flexible approach to client-server computing.

The number of standards for client-server computing almost equals the number of client-server applications. Although there are some standardization efforts, there is hardly any experience with applying these solutions to (medical) practice. The layering approach of the integration architecture HERMES anticipates the inclusion of commercial standards (when available and evaluated); it currently supports already the development of client-server applications. The toolkit provides support for the development of the communication between client and server through a callback mechanism. Stubs created with the HERMES toolkit contain a number of built-in callbacks that manage sessions between clients and services. An important feature of HERMES is the ability to include existing legacy systems as if they are true open services. In this way, the growth path from stand-alone to client-server computing can be shortened and the implementation of the client-server architecture can begin immediately. Moreover, the existing legacy systems remain available as a stand-alone solution for daily practice. Clients and services are connected through the HERMES kernel. This kernel uses a database to find the best server match for a request from a client. Moreover, it completes requests with mandatory data and tries to optimize performance. Special features are included to minimize the memory burden of the session-oriented client-server model. Currently, a system for the outpatient clinic cardiology, occupational health care, and clinical data analysis is available.

HERMES: a health care workstation integration architecture.

An architecture is described that facilitates integration of existing databases and applications without modifying them. By means of this architecture, data from different sources dispersed in a network can be combined and directly used in existing applications or applications that have been developed specially for integration. This feature of combining data from different sources into one workstation is viewed as the enabling technology on which computer-based patient records can be built. The abstraction of computer-, network- and application-specific details is completely dealt with by the integration architecture. This integration architecture has been developed with extendibility and flexibility in mind, and allows for a growth-path towards application of the open system paradigm in medicine.

Multiple imputation as a missing data machine.

This paper deals with problems concerning missing data in clinical databases. After signalling some shortcomings of popular solutions to incomplete data problems, we outline the concepts behind multiple imputation. Multiple imputation is a statistically sound method for handling incomplete data. Application of multiple imputation requires a lot of work and not every user is able to do this. A transparent implementation of multiple imputation is necessary. Such an implementation is possible in the HERMES medical workstation. A remaining problem is to find proper imputations.

A cooperative model for integration in a cardiology outpatient clinic.

With the increasing amount of digitally stored patient information, such as images and findings, the possibility and need arise for a system which is able to both store and display this information in a structured, user-friendly way. In the common situation where different information systems are used within one department, this means that the various information systems have to be integrated. However, integration requires more complex management of data, processes and windows. This management can be handled by a Workspace Manager, which controls inter-application tasks, and interaction with the user. Furthermore, this Workspace Manager supports the user with the definition of complex tasks such as collecting problem-related patient information from the various remote systems. In an outpatient clinic for cardiology this architecture is used to achieve cooperative integration of an open Computer Patient Record with a range of information-specific services.

User evaluation of an integrated medical workstation for clinical data analysis.

Results are presented of the user evaluation of an integrated medical workstation for support of clinical research. Twenty-seven users were recruited from medical and scientific staff of the University Hospital Dijkzigt, the Faculty of Medicine of the Erasmus University Rotterdam, and from other Dutch medical institutions; and all were given a written, self-contained tutorial. Subsequently, an experiment was done in which six clinical data analysis problems had to be solved and an evaluation form was filled out. The aim of this user evaluation was to obtain insight in the benefits of integration for support of clinical data analysis for clinicians and biomedical researchers. The problems were divided into two sets, with gradually more complex problems. In the first set users were guided in a stepwise fashion to solve the problems. In the second set each stepwise problem had an open counterpart. During the evaluation, the workstation continuously recorded the user's actions. From these results significant differences became apparent between clinicians and non-clinicians for the correctness (means 54% and 81%, respectively, p = 0.04), completeness (means 64% and 88%, respectively, p = 0.01), and number of problems solved (means 67% and 90%, respectively, p = 0.02). These differences were absent for the stepwise problems. Physicians tend to skip more problems than biomedical researchers. No statistically significant differences were found between users with and without clinical data analysis experience, for correctness (means 74% and 72%, respectively, p = 0.95), and completeness (means 82% and 79%, respectively, p = 0.40).(ABSTRACT TRUNCATED AT 250 WORDS)

A new architecture for integration of heterogeneous software components.

An architecture is described that integrates existing applications in a network-wide system. The architecture follows the new open software paradigm, and defines kernel and application services that collaboratively solve the tasks of end-users and provide them with an intuitive user-interface. This paper describes the message language and the kernel mechanism for addressing application services. The architecture has been developed as much as possible to conform with current standards.

A prototype integrated medical workstation environment.

In this paper the requirements, design, and implementation of a prototype integrated medical workstation environment are outlined. The aim of the workstation is to provide user-friendly, task-oriented support for clinicians, based on existing software and data. The prototype project has been started to investigate the technical possibilities of graphical user-interfaces, network technology, client-server approaches, and software encapsulation. Experience with the prototype encouraged discussion on both the limitations and the essential features for an integrated medical workstation.

Does an integrated medical workstation really help clinicians?--A formal user evaluation.

A formal user evaluation of an integrated medical workstation for support of clinical data analysis is described. Twenty-six users participated in an assessment study that consisted of a self-instruction course, followed by an experiment in which six clinical data analysis problems had to be solved, and the assessment was concluded with an evaluation form. To facilitate an objective and quantitative assessment, the time spent to the course and to the problems, as well as completeness, correctness, number of solving attempts and persistence were measured. Not all problems were solved by all participants. From an analysis of the measurements, the following conclusions were drawn. (1) Clinicians correctly solved about 69% of all problems tackled. However, they skipped 33% of the problems. (2) Non-clinicians correctly solved 87% of the tackled problems and skipped only 10%. (3) The performance of users not experienced in clinical data analysis was raised to the level of those with clinical data analysis experience. (4) An inventory which can be used to direct future developments was made of the errors of interaction. This assessment study has contributed to gaining insight into the conceptual problems and practical bottlenecks clinicians have with clinical data analysis.