Skip to main content

European Standardised Telematic Tool to Evaluate Electromyographic Knowledge-Based Systems and Methods

Objective

The use of knowledge-based systems within medicine is today still very sparse despite much research effort and many expectations. A fulfilment of these systems' promise depends on two major issues: 1/ the establishment of scientific methods for their evaluation and validation, and 2/ a complete integration of these systems into their different kinds of data/information sources like clinical databases, different kinds of signal processing equipment etc. to make them fully operable. The lack of these requirements are today the major hindrance for exploiting this technology in a large scale.

On the basis of substantial experience derived from several different disciplines the overall goal of ESTEEM is to develop a prototype workstation, which will demonstrate the feasibility of a fully integrated and coherent EMG information environment.
A prototype workstation is being developed which will demonstrate the of a fully integrated and coherent electromyograph (EMG) information environment environment is technically composed of a set of different functional modules (eg telecommunications device, knowledge based systems KBS and evaluation tools), al communicate via a central medical information software bus. This system entitled will have the potential to manage all clinical informatic needs in the EMG labor be also capable to manage different kinds of KBSs concurrently in order to evalu systematically.

The technical line of the project achieved the following:
speci of the EMG-Platform architecture;
development of an EMG-Platform version 1.0 in EASE kernel connects the existing knowledge based EMG system KANDID to a Counter EMG machine;
development of a general EMG case data collection tools;
specificat version of telecommunication protocol;
development of a preprototype of a second knowledge based EMG system based on a causal functional approach;
development of prototype of a system for deriving sensitivity studies of a KBS;
establishment o telecommunication network enabling exchange of EMG case data files and participa consensus procedures for the interpretation of these cases.
This environment is technically composed of a set of different functional modules e.g. databases, telecommunications device, knowledge-based systems and evaluation tools, all of which can, communicate via a central medical information software bus. This system entitled "EMG-Platform" will have the potential to manage all clinical informatic needs in the EMG laboratory, and must be also capable to manage different kinds of knowledge-based systems concurrently in order to evaluate them systematically. On this basis, two main functional aspects of this information system emerges:

1. To empower the international EMG society with an EMG information network in which the nodes are EMG-Platform systems. This network will enable the EMG clinicians to build up a gold standard patient case database. Also, international comparative studies of the complex EMG diagnostic process will be possible in a quantitative scale unseen today. Finally, knowledge-based EMG systems will be able to be applied everywhere in this network for evaluation purposes.
2. To cover the information processing needs seen in the daily routine situation in an EMG laboratory. This will be materialised into an EMG-workstation which can be used as a stand-alone system or a network system.

Technical Approach

To achieve such objectives the ESTEEM project follows two distinct (but highly interactive) action lines: a technical development line and a clinical usage line. The basic technical concept is to develop a system kernel - the "EASE" kernel (Exchange And Storage of EMG information) - which handles all the communication between the different subsystems/modules of the entire system namely Knowledge-based EMG-systems, clinical databases, Tools for Evaluation Purposes, Telecommunication functions and interface to Hospital Information Systems. Two existent KBS-EMG systems (PC-KANDID and NEUROPS/MYOSYS) will be integrated into the EMG-Platform and evaluated by the clinical user group by means of the provided facilities. In addition, this system kernel must have an open architecture which will enable third parties easily to attach their own developed modules (e.g. another knowledge-based system) to it, following the necessary interface specifications. To demonstrate the feasibility of the open reference architecture a prototype of a second generation of knowledge-based EMG-systems based on a causal-functional approach (Demonstrator system) is to be developed and interfaced to the EMG-Platform.

The perspective of such an EMG-Platform points to the Clinical part of the ESTEEM project. Firstly, it must be considered that a fundamental necessity for the communication of knowledge is a consistent and coherent terminology to represent abstract concepts. This will contribute to a substantial clarification of the process of analysis of data from EMG studies. Secondly, the investigations into the current methodologies used throughout the EEC with the purpose of setting standards for the acquisition, analysis and representation of neurophysiological data provide the techno-line with a robust set of procedures for technical implementation. In addition, it is also important to consider the inter-laboratory variation in epidemiological factors (e.g. sources of referral, distribution of diagnosis, etc.) when evaluating the possibilities for such standardisation of the EMG investigation namely in terms of acceptability and dissemination of results. This will reflect both in the development and evaluation of knowledge-based systems in terms of the coverage of such systems and their usability outside the laboratories where they have been developed, hence circumventing the limited capabilities of today KBS. Complementing this study, the user group will develop a clinical database of EMG patient cases containing the necessary representation of different categories of Neuromuscular Disorders and their respective variants and distinct examination techniques. Finally, the emerged data set structure constitutes the basis for the standard communication protocol for exchange of EMG-data between the modules. Moreover, this database will serve as a reference structure for benchmark evaluations and different calibrating procedures for better performance of a KBS.

ESTEEM project will allow a set of EMG-Platforms to be applied at seven clinical EMG centres within Europe interconnected via the telenet. On the basis of this international network of EMG-Platforms, EMG specialists and computer scientists will collaborate primarily in the field of evaluating and validating knowledge-based EMG systems for their diagnostic and for other supportive capabilities to the clinicians. Secondly, different types of today used EMG examination techniques will be discussed and assessed. With the combination of the network of EMG-Platforms and the associated specialists a truly "Medical Audit" environment will be established.

Key Issues

- Multicentre gold-standard EMG case database
- EMG workstation
- Evaluation of knowledge-based EMG systems
- Evaluation tools
- Second generation knowledge-based EMG systems

Expected Impact

In ESTEEM relevant existing standards are being analysed. Also candidates for genuine new standards will be developed and tried established when necessary. This is so for the ESTEEM communication protocol.

Establishing on-line connections between different EMG laboratories in Europe will open up for radical new possibilities for transnational exchange, analysis and discussion about EMG diagnostics on a concrete case-related basis.

Relationship to previous work

Esprit P599: "An expert assistant for electromyography"

Relationships to other projects and actions

KAVAS-2, OPENLABS, DILEMMA, OEDIPE

Testbed and verification

The clinical user group of ESTEEM will guarantee the project clinical relevance, testbed and verification.

Topic(s)

Data not available

Call for proposal

Data not available

Funding Scheme

Data not available

Coordinator

HOSPITAL EGAS MONIZ
Address
Rua Da Jungueira 126
1300 Lisboa
Portugal
 

Participants (3)

GENTOFTE HOSPITAL
Denmark
Address
Niels Andersens Vej 65
2900 Hellerup
 
UNINOVA
Portugal
Address
Campus Da Fct/unl
2825 Monte Da Caparica
 
Université de Grenoble I (Université Joseph Fourier)
France
Address

38706 Grenoble