Ebook: Advances in Medical Informatics

This volume contains an introduction to the AIM Exploratory Action and contributions by the forty-two projects with brief descriptions of their results. Contributions have been included on the three areas of general interest in which AIM took an active part, namely data protection and confidentiality, patient data cards, and standardization.

The enthusiasm of the project partners, the interest of outsiders in the results, and the support by our colleagues inspired us to produce this volume. It is meant to be a summing up and a guide to bring the reader to those experts in whose work he might be interested.

Annex 1 lists the authors; annex 2 the projects with their managers, addresses, project partners, and main documents on scientific and technical work performed; annex 3 is a subjects index; and annex 4 lists the public documents produced during the AIM Exploratory Action.

We would like to express our gratitute to the contributors for their efforts. Working with them made the production of this volume pleasant as well as instructive. As usual the task took far more time than anticipated.

Jaap Noothoven van Goor

Jens Pihlkjær Christensen

Brussels, November 29th, 1991

A brief survey is given of the development of the AIM Programme and the AIM Exploratory Action. The action included 42 projects and accompagnying activities, which are described in the contributions to this volume, and which are introduced here.

During the AIM Exploratory Action the ADAM -ADvanced Architecture in Medicine- project, undertook to design a workstation architecture for General Practitioners, which would be integrated with the services supplied by the National Health Care Systems. This architecture is presented globally. Other results, as contributions to standardisation, conditions for a functional reference model for the information processing environment in general practice, and grants for security are discussed. Validation of the model was achieved by the design and implementation of basic modules, namely the Medical Data Base and the Integration Tools. Account is taken of the exploitation of the results, and further plans are outlined.

Images represent one of the most important sources of medical information. There are, in principle, great benefits to be obtained by using intelligent computer systems to provide support for clinicians involved in making image-based decisions. AEMI has developed a coherent approach to computer-assisted image interpretation, identifying the clinical requirements, developing a formalism for human-computer interaction and proposing a knowledge-based system architecture.

The overall intention of AIDMED was to address the need which medical personnel have for easy access to the wide variety of information and data sources available now, or soon, in computerised form. In the Exploratory Action of the AIM Programme a task was included of providing an interface which served this purpose. An initial analysis of user requirements led quickly to the construction of demonstrator prototypes, which were evaluated with medical personnel on simulated case consultations. Further revision of users’ needs, and the development of interaction models followed. The cycle of information gathering and evaluation led to the building of a hospital department prototype. A functional reference model and an initial top-level design of the system architecture were constructed, from which a detailed system specification could be derived.

This paper describes the development of evaluation techniques to be used in the assessment of information technology applied in the health care sector. The potential benefits and costs of such technology are identified and methods of measuring them explored. An approach to evaluation is presented drawing on experience from projects in the AIM Programme.

The AVICA project was concerned with the processing and manipulation of digitalized medical images. As pilot areas endoscopy and cystoscopy in gyneacology and urology were chosen. Investigation of methods for image enhancement and parallel processing are described. The conditions for an electronic transfer of images, and its possible benefits in medical practise are analyzed. A demonstrator was developed, in this phase including digitization, archiving and analysis techniques.

In this paper we present a synopsis of the results of BIOLAB. The project aims at the elaboration of an integrated biomedical laboratory. This integration is considered around the softaware and hardware specification of a powerfull workstation. The project has studied the hardware and software specificatons of a biomedical laboratory workstation. Through this workastation various heterogeneous madical information systems can be accessed. In the project the harware specification of the workstation has been completed as well as the specification of a language for accessing the various heterogeneous medical information componenets. Such components studied in the project are a medical image processing module a neural network classifier and a neural network diagnostic system.

A Computer Aided Community Oral Health Information System (CACOHIS) was developed under AIM Exploratory Action. The important features of this system are that the oral health requirements of oral health physicians are defined. A mechanism for collecting critical, relevant and meaningful information is described. This information is collected by direct data entry into computers in the field. These elements then provide readily available and useful information for the planning and evaluation of oral health services by health service personnel at various levels. CACOHIS can be used both in the developed and third worlds, particularly for screening for early oral manifestations of viral diseases. The important features of CACOHIS are that it facilitates a normative planning process and can be successfully utilised in under-developed health systems.

The CAMAC project developed framework systems supporting both clinical and hospital management and laid the foundation for a common language relating resource use and clinical evaluation.

The main conclusion of the project is that Patient Classification Systems PCS can be defined for a specific objective, like resource management, quality assurance or planning. A European approach is proposed to managing clinical care both from the demand and from the supply side.

In the area of medical data the concept of a Minimum Basic Data Set was revised and a European Patient Classification Systems framework defined. In the area of hospital cost accounting a Minimum Standardized Cost Data Set and common and flexible cost systems were defined to be applicable in different countries and in different hospitals. Management applications have been proposed for financial, utilization review/quality assurance, resource allocation, planning and budgeting. A conceptual framework was developed for a Management Support System uniting all the different applications.

CAMARC is an interdisciplinary project aimed at the assessment of Computer Aided Movement Analysis Systems in the Context of the Clinical Rehabilitation of motor disabilities. The total project had originally a time perspective of five years; during the of the AIM Exploratory Action of 18 months, the feasibility aspects undertaken. This part of the project comprised about 10 man years. Accordind to the plans, the results were specifications, technical and normative reports, hardware laboratory prototypes, software packages and some protoype demonstrators, all presented in 12 deliverables. In this paper the rationale of the project and the main achievements will be described. The possible exploitation of the results is mainly scientific. These aspects and the topics related to standardization are also discussed. The results the perspective of building-up a European Network of Clinical and Research laboratories able to share results, software packages and knowledge on the basis of concerted clinical protocols.

This short report presents a summary of the results of the Community Health Information Classification and Coding (CHIC) project. The project, which covered the period from June 1989 to June 1990, had the objectives of identifying the Minimum Basic Data Set for ambulatory care in Europe, producing an extensible structure for the MBDS, conducting a survey of the acceptability of the MBDS, and developing specifications for supporting software. This report also outlines the results of the additional additional achievement of producing a coherent data set that summarises the patient's medical history.

This paper presents the main results achieved by the COVIRA consortium project during a 15-months period under the AIM Exploratory Action programme. The prime result of the project is a specification of a system for knowledge based interpretation of cranial MR images as a means of computer assistance in diagnostic radiology, radiation therapy planning and stereotactic neurosurgery. This specification is based on demonstrated results from prototypical implementations and an analysis of the state-of-the-art and of the clinical requirements.

Image segmentation and interpretation results were obtained for a set of test images of tumor patients selected by medical experts from the above fields. For each patient slice, two MR spin echo images are available.

For image segmentation, results are presented for three schemes: one using a Canny edge detector followed by a smooth patch fitting for region finding, another using a region detection according to Nagao/Matsuyama and a Marr-Hildreth edge detector based on one of the echoes to guide region merging, and a third scheme using a multiscale approach to edge detection.

The image interpretation results presented are based on a case model representation of clinical, anatomical, MR-Physics, and tissue parameter knowledge. Results are presented for two schemes: one using a fuzzy clustering approach providing a fuzzy segmentation and a subsequent fuzzy relational matching step, and another using a blackboard approach based on the classical low-, middle- and high-level of processing, allowing for a dynamic control with feedback and backtracking mechanisms.

The results were comparatively evaluated by medical experts based on criteria of clinical usefulness.

The general objective of EPIAIM was to define a knowledge based environment to be used as a support tool during the different phases of an epidemiologic survey (planning, management, analysis). The main topics approached by EPIAIM were: i) the definition of eliciting modalities concerning the acquisition of knowledge about the conceptual structure of a survey; ii) the implementation of a Design-Management subsystem; iii) the design of a consultation system for the Epidemiologic Data Analysis (EDA); iv) the definition of user models in order to tailor the behaviour of the system to the specific user knowledge level.

The goal of the EUCLIDES project was to provide a European standard for clinical laboratory data exchange between independent and heterogeneous Medical Information Systems. EUCLIDES aims an open standard i.e. freely avialble to all, and unbiased towards any manufacturer's hardware or software. By convergence with other emerging standards, it could develop into a world standard. It could also be extended to other fields of medicine. EUCLIDES addresses the data exchange problem in a three-pronged approach : 1. the message transfer-mechanism, 2. the message syntax, and 3. the medical coding systems used within the message. Moreover, the EUCLIDES system offers through the EUCLIDES BRIDGE a user-friendly software solution for interfacing with existing local systems at both the sender's and receiver's ends.

EURODIABETA was one of the largest (16 partners, 3 sub-contractors, from 7 EC member states) AIM funded projects bringing European experts together to perform a feasibility study on the production of a Computer assisted Chronic Health Care Environment to support diabetes care as a paradigm of all chronic disease care. Good project management has demonstrated the viability of such an ambitious goal. Existing prototypes have been evaluated as components for an integrated system, and conceptual modelling has been identified as a means for providing the logical framework within which the components will fit.

The project is concerned with the specification, implementation and testing of a system to be connected as an add-on device to an echographic scanner. The system will enable preprocessing, analysis and processing of echographic data and images, in order to achieve the employment of the full information available from echograms. The hardware (VME-bus based) is installed at the two institutions. A special 40 MHz (8 bits) ADC-board was developed in addition. The data acquisition software was defined and implemented enabling the single-flash storage of a full scan in RF-format (2 MBytes), as well as a storage of the original video images. The preprocessing consists of corrections for the beam diffraction, the tissue inhomogeneities and the equipment performance settings (gain, TGC). Acoustospectrographic tissue parameters and image texture parameters are estimated for tissue characterization. Multi-center clinical testing and the development and assessment of image processing methods are scheduled in the next phase of the project.

A suitable General Architecture for Medical Expert Systems (GAMES) has been defined in terms of generic tasks (diagnosis, therapy planning and monitoring) to be executed for developing medical reasoning in a patient's management. Their epistemological structure has been investigated. The GAMES project aims at designing a general architecture for knowledge based systems in medicine. The focus of the project is on the comparison and validation of different approaches in order to implement the generate and test cycle, which represents the basic feature of the epistemological model of medical reasoning we developed.

The objective of the HELIOS project was to create a Software Engineering Environment to facilitate the development of medical applications. HELIOS comprises a set of Software Components, communicating through an abstract channel called Software Bus, providing services to the application designer, and finally to the application end-user. The kernel comprises three main Software Components, the Information System, the Documentation Facilities, and a component that provides a convenient user interface, the Interface Manager. The Services consist of a collection of toolkits providing the necessary services to medical applications. Up to now, they are represented by an Image Processing Toolbox.

The object oriented paradigm is used both as a methodology for modelling the entities and procedures found in an application and as the basic structure for building the Software Components. Development standards include UNIX as operating system and X Window/MOTIF as windowing environment. The target application for the prototype is the development of a hospital Ward Information System with its Ward Image Manipulation Subsystem.

Reasons are given for the little clinical successes the first generation of PACS had known. It has become clear that the next generation of PACS should be integrated with the HIS or RIS. The enormous amounts of data to be transported put specific requirements on network management. Distributed data-bases will be essential. Adaptive user interface will facilitate the use by and the communication between several types of professional users. The HIPACS project studied the aspects and concluded that we are moving from a development driven by technology towards a approach based on the embedding of medical information in the systems.

Observation at the microscope is basic and crucial to patient care and the most time consuming of medical image examinations tasks. There is a clear and urgent need to make these examinations faster, more reproducible and objective. The HOME microscope workstation presented here proposed to provide the pathologists, cytotechnicians and the laboratory manager with the necessary tools to : facilitate the conventional tasks at the microscope, interface cell and tissue image processing facilities in daily routine work, and integrate the microscopes in a network compatible with the computer infrastructure of clinical pathology laboratories. The innovation is to project the computer images, graphics and alpha-numeric instructions into the microscope optical image itself, which then becomes an addressable bitmap system. Neither the alterations to the microscope nor the PC housed under the workstation are apparent and the keyboard has been replaced by a mouse. Thus, a computer innocent cytotechnician or pathologist can use it to : drive the microscope stage, select the desired application program, select the type of measurement to be performed without having to look up from the microscope. Several prototypes have already been developped by Carl ZEISS (Germany) and LEICA (Germany). Typical application software packages for cervical cancer screening and assistance to myopathies diagnostic are now available.

A European approach towards resource management and strategic planning was implemented in the HOSCOM project of AIM by defining information standards needed across countries, as well as a methodology to measure resources and costs at institutional and inter-institutional levels.

A European Health Data Base (EHDB) was set up in order to test data availability and comparability as well as to validate models through macro-comparisons using case-mix (DRG's, refined grouping, disease staging) and micro-comparisons based on three diseases (cardiac valve replacement, diabetes mellitus and hip fracture). Presently this EHDB contains 274,164 medical record summaries sampled from 7 countries. It allowed to build prototypes (using Clipper, Prolog and SQL) in order to compare uniform aggregates for different countries, with standard software tools for statistical comparisons. It showed the present feasibility to use case-mix based on the European Minimum Basic Data Set (MBDS) and the difficulty to obtain uniform data on resources and costs other than length of stay across countries. Medical data confidentiality was assured but not yet population-based representativity. A number of problems have been identified, given the current state of development of the EHDB, which can be solved by international research and development projects in the near future.

Intensive Care Units (ICU) are information and personnel intensive high technology clinical departments. In the U.S.A. they represent 10 % of hospital beds but 25 % of costs. Together with laboratories and radiology departments, ICUs are the medico - technical pre-requisite for high quality hospital medical / surgical practices.

ICSIC has concentrated on action coordination in ICUs. A lot of operational problems in ICUs result from human interactions in fairly large teams (25-100 people) where responsibilities are shared and actions are distributed.

The structure of actions in ICUs was modelled in terms of scenarios - such as therapy scenarios, laboratory test scenarios... ie recurrent organization of work in which personnel and resources are engaged. Workflows are the building blocs of scenarios : they describe the transfer of tasks and responsibilities from requestor to actor in 4 steps (requesting, setting rules and conditions of satisfaction, ie expected results and planning, performing the planned technical tasks and reporting, evaluating the results). Workflow management is the basis of nursing, care, and support of medical activities. This is not specific for ICUs and has broader application.

User acceptance and system transferability are a must for any successful industrial product. Standards are required also about how people work together, about how they assess changing situations, about how they interact with a computer at the bedside and not only about data communications. In this respect a complex data base structure was designed and an innovative Human Computer Interaction scheme was conceived and implemented ; it is currently being tested in the field. It has application beyond the ICU world.

To support these activities, an innovative technological platform is being tested (UNIX V, PC network, Ethernet,TCP/IP & NFS protocols, X-windows, SQL database). Short term exploitation of the results was achieved through a fast development of spin off systems, as early as begin 1991.

The long-term aim of the INFORM Project is to develop, evaluate and implement a new generation of Information Systems for hospital High Dependency Environments (HDE). Such environments include Intensive Care Units (ICU), Neonatal Units, Burns Units, Operating and Recovery Rooms, and other specialised areas. The INFORM System will integrate Decision Support with on-line, off-line and observed patient data and, in addition, will incorporate and integrate unit management features.

A Functional Specification has been produced in the Exploratory Phase of the Project, and this has been achieved by dividing the Project into the following four main components:

conceptual modelling of the HDE;

evaluation of existing HDE Information Systems;

development of a novel software architecture using a Knowledge-Based Systems (KBS) methodology, and based on a critical review of KBS applied to the HDE;

monitoring of appropriate leading-edge technological developments, and of current and emerging standards

Activities relating to both clinical patient management and unit management have been modelled using a modern structured analysis methodology; process models are described with dataflow diagrams, and data models with entity relationship diagrams.

Key and difficult areas in clinical decision making have been identified; an important aspect of INFORM is the integration of clinical Decision Support for these areas into the Information System through a layered architecture. The lower layers are concerned with monitoring and alarming and the higher levels with patient assessment and therapy planning.

It was apparent from the evaluation of current systems in HDEs that these are not designed in such a developed and integrated way as to yield their full potential. The user interface is also crucial to a system's success. The INFORM System must reflect the needs and working practices of its users, and the highest priority will be given to these aspects throughout its design, development / prototyping, and implementation phases.

The goal of the IRHIS project was to explore a solution to the problem of information overload in an effort to improve health-care delivery. This was performed by implementing a prototype demonstrator that assists the user to easily access on-line medical records. The system is based on predefined user requirements and uses techniques such as graphics, hypertext and artificial intelligence. We cannot claim that we solved the problem of information overload. From the evaluation phase of the project however, it appears that the implemented prototype would allow the user to save time and get better insight into complex medical problems.