You are on page 1of 7

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/282288049

The medical project management (MPM) system

Conference Paper · June 2015


DOI: 10.1109/GSCIT.2015.7353336

CITATION READS

1 856

3 authors:

Ahmed Dridi Anis Tissaoui


Institut Supérieur de Gestion de Tunis Faculty of Law, Economics and Management of Jendouba, University of Jendouba
9 PUBLICATIONS   18 CITATIONS    25 PUBLICATIONS   53 CITATIONS   

SEE PROFILE SEE PROFILE

Salma Sassi
Université de Jendouba
37 PUBLICATIONS   47 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Healthcare in the age of the IoT View project

Medical decision support systems View project

All content following this page was uploaded by Anis Tissaoui on 29 September 2015.

The user has requested enhancement of the downloaded file.


The Medical Project Management (MPM) System
A semantic middleware for a unified representation of medical records

Ahmed Dridi1, Anis Tissaoui2 and Salma Sassi3


Faculty of Law, Economics and Management of Jendouba
University of Jendouba, Avenue de l’U.M.A, 8189 Jendouba, Tunisia
1
ahmed-dridi@outlook.com, 2tissaouianis@yahoo.fr, 3sassisalma@yahoo.fr

Abstract—The Medical information is one of the types of systems consist of a barrier for they can communicate and
information that are characterized by the multiplicity and exchange data between them. In fact, the communication and
diversity of its sources. This diversity contributes to the the interaction between different information systems is
emergence of many problems; of which the most important is the essential for the management of projects whose information
difficulty of communication and interaction between systems.
was fragmented and scattered on several different and
This is what is called the problem of interoperability. In this
context, we propose in this paper a new mediator semantic that heterogeneous sites. As the case of medical field, our scope of
we call the Medical Project Management System. Based on application, the need to exchange data between of different
ontological components and an XML data exchange format, this systems of healthcare institutions (hospitals, clinics private,
mediator allows the user to have a unified representation of the laboratories of radiology and medical imaging, biological
information originally distributed on several different and analysis laboratories, etc.) is still fundamental because it
heterogeneous sites and which are related to a specific project. In conditions effective coordination of care and patient
addition, it helps the user to interact with a large number of monitoring. However, from a technical point of view, this
medical projects and to research similar projects and comparing sharing is obstructed by some problems.
them to take advantage of best practices and shared experiences
to improve the quality of treatment.
According to [2] these problems consist primarily of
Keywords—Interoperability; heterogeneity; healthcare three points: the first concerns the diversity of systems that
information system; Project Management; Electronic Health makes them difficult communication because these systems
Record; medical information visualization. may involve the same profession but be built with different
models. The second is being the useful information that is
I. INTRODUCTION generally owned by its creator with all the risks that this
may entail of which the immediate consequence of these
In the era of globalization and computerization which its
risks is the loss of information held by the field professional.
first features are appearing in the middle of the last century, the
Finally, the last point, which is a large variation in the nature,
digital information has become a crucial element in al- most all
syntax and format of data. Consequently, access and retrieval
areas and aspects of modern life. Many areas are impacted by
of relevant information in an environment characterized by
computerization, which in turn she had a great impact on the
the three main problems is made very difficult because the
social and cultural consequences of society, and contributes
channels of communication and coordination between the
with the globalization to the emergence of the so- called
different sources of information are almost nonexistent. In fact,
information society and the knowledge economy. In the
these issues described a famous problem that mainly consists
information society, the creation, distribution, use, integration
of the interoperability problem. In the following, we would like
and manipulation of information is a major activity and the
firstly to define the concept of interoperability and these
primary objective of this society is always to get a competitive
dimensions, before that,we try to develop a state of the art for
advantage internationally, by the use of information technology
the proposed solutions to deal with this problem by focusing
(IT) in a creative and productive way [1].
particularly on mediation approaches.
Information systems as a discipline emerged initially as a
Finally, and before concluding this paper, we present our
result of computerization and the development of Information
proposition that consisting mainly of a new semantic mediator
Technology and Communication (ICT) and as a response
designed for viewing, analysis, exchange and processing of
to the need for a means simple, efficient and economical to
medical data that are typically highly distributed and
store, organize, retrieve and restore large volumes of scalable
heterogeneous.
data and information. The information systems are
experiencing a constant growth since inception. Their use II. THE PROBLEM OF INTEROPERABILITY
allows new opportunities to access and process information.
However, the growing number and diversity of information The Institute of Electrical and Electronics Engineers
(IEEE) defines interoperability as the ability of two or more
systems or components to exchange information and to use the MIX [11] (Mediation of Information using XML) is an
information exchanged [3]. Interoperability is considered very XML-based mediator provides the user an integrated view of
important indeed critical in many fields, including computer different sources developed in the framework of a project
science, medical field in the broad sense, rail activities, electro- within the “University of California at San Diego Database
technology, aerospace, military and general industry [4]. Laboratory”. This view is an interface defined using the
Generally, three types of interoperability can be distinguished XMAS Language (XML Matching and Structuring Language).
[5]: XMAS is a declarative language based on the XML- QL query
 Technical interoperability: which aims data transport language used as a definition language of view, and to query
and transport security. This type of interoperability the interface of the system.
appealed to the infrastructure, protocols, and services
that enable, among other things, the communication of PICSEL [12] is a mediator serving as an interface between
data between computers [5]. Include, for example, the users and data sources to query. The architecture of PICSEL
Protocols TCP/IP, HTTP, etc. consists of a domain ontology and knowledge bases (BCcomp)
 Syntactic interoperability: concerns the structure of connected to the mediator, describing the content of the
the exchanged messages and is intended to ensure information sources. An ontology is used to represent the
coherence in the manner in which the information knowledge of a field in the form of a hierarchy of concepts and
exchanged are represented. Specified data formats, semantic relationships between these concepts [13]. The
communication protocols and others come into play. In ontology in PICSEL provides all the useful vocabulary for
general, XML or SQL standards provide syntactic users to formulate their requests. It is described by using the
interoperability [6]. language of representation CARIN-ALN, a rich language to
 Semantic interoperability: which is ensuring that the describe in a very fine way of a domain ontology and content
exchanges which occur maintain their meaning, that is of information sources. Other knowledge bases allow the
to say, the communicating parties share a common mediator to determine what sources of information can provide
understanding of the meaning of the data they some answers to a user request. Between each knowledge base
exchange [7]. It involves the ability of a user to and each source is connected to a wrapper. Its disadvantage is
understand the information obtained in order to use it that once shared ontology defined, each source must use the
properly [8]. common vocabulary, which limits the autonomy of local data
sources.
III. S TATE OF THE ART: T HE SEMANTIC MEDIATORS
It is inevitable to find a solution for the interoperability The project MOMIS (Mediator envirOnment for Multiple
problem so that the different systems and information sources Information Sources) [14] concerns the creation of a mediator
can communicate and share information. The end of the 1970s, based on structured and semi-structured data integration
much research has been conducted in this area and have given systems. It was developed in collaboration between the
rise to the federated database. These databases are based on a University of Modena and Reggio Emilia and the University of
pivot model manually created and are used to query different Milan and Brescia. It was designed to allow access to
databases. In the 90s, mediators-based techniques [9] allow the heterogeneous in- formation stored in traditional databases
emergence of new mediation architectures. A mediator is a tool (relational and Object- oriented), the systems files and semi-
that allows users to query various heterogeneous databases. It structured data sources. It uses the OQLi query language to
is based on an ontology or a unified view, allowing the user to interrogate the sources. Another system has developed in the
feel that queries a single database. same project to provide access to heterogeneous information
sources in the field of energy (research centers, statistics, etc.)
The mediation approach is to develop an application which is the EDC system (Energy Data Collection) [15]. It
responsible to act as an interface between databases and takes as input queries expressed in OLAP (Online Analytical
applications used by users. A mediator helps to resolve Processing) which it then sends to the query planner developed
conflicts between the models of data sources; it generates an in the architecture of MOMIS. The latter is responsible to send
execution plan between users and those sources. In addition, it responses to the client by querying databases may contain the
simplifies and describes the distributed data. desired information.
IV. T HE M EDICAL P ROJECT M ANAGEMENT (MPM)
E-XMLMEDIA [10] is a mediator system of federated S YSTEM
databases using XML as a federated model. Its purpose is to
provide access to multiple sources of data on the Internet or Always aiming to solve the problem of heterogeneity and
Intranet using XML protocols and tools based on Java ensure interoperability between different medical systems, we
components. It allows to federate heterogeneous data sources, propose a new mediation approach that we call the Medical
by presenting a unified view of all of the sources. The query Project Management (MPM) System. The MPM System is a
language is based on XQuery, and the results are in XML semantic mediator between the so-called information producers
format. and users of information. Based on a powerful semantic
expression provided by ontologies, our mediator allows
applying a sort of reconstruction and conceptualized Querying and Visualization Module: This module is
reformulation, annotated and temporalized information to pro- attached to the Human-Computer Interaction in our system. It
vide a unified and shareable representation of information that includes a set of graphical interfaces that allow the user to
are complex, heterogeneous and already exploded on different interact and communicate with the system.
and highly distributed sites. In addition to its primary purpose
for which the mediator is developed, which consists mainly to 1) Data and Knowledge Module:
provide a shared semantics between users heterogeneous The MPM system consists in a first step to transform the
medical information systems. The MPM system is also de- information contained in the usual information files to concepts
signed to help clinicians and researchers in the medical field to semantically annotated and contextualized in the so called a
explore, visualize, and analyze a set of medical data large and “Unified Virtual Folder” (UVF). We define a UVF as being a
heterogeneous. Practically, our system allows these users to set of concepts linked together by semantic relationships.
view and interact with a large number of Electronic Health Contrary to the information contained in the traditional
Record, and to search for similar EHR and comparing them to information folders that are not described and not referenced,
take advantage of best practices and shared experiences to the concepts of the UVF are semantically described by
improve the quality of treatment. annotations and metadata are contextualized. These concepts
are also referenced by icons, which are attributed to
A. The MPM System architecture
documents. Any Information Folder relating to a patient is
As we have already said, our system is a semantic mediator converted into UVF. The concepts of the DUV obey a Domain
between the so-called information producers and users of Meta-Ontology and Task Meta-Ontology. All created UVF are
information. It takes as input data streams and information stored in a relational database. In this database, each record has
distributed to provide a unified representation of them. All data an identifier and a set of metadata. In fact, the metadata
coming from different sources of information entering into the consists of a summary of knowledge extracted from the file
mediator must be converted to cognitive information itself. The main objective of this metadata is to facilitate
graphically represented, passing through three main modules research especially in large amounts of data as our case.
that are the basis of the architecture of our middleware:

Fig. 1. The MPM System architecture


Fig. 2. Data and Knowledge Module

Data and Knowledge Module: in which data coming from 2) Analysis and Treatment Module
multiple information sources can be identified, described, This module serves as an intermediary between the user
unified and conceptualized according to two meta-ontologies (a interface and the basis of the UVF (the basic concept of the
Domain Meta-Ontology and Task Meta-Ontology), that they data module and Knowledge). It runs in the background and
solve the problem of heterogeneity and ensure semantic supports the analysis and execution of user requests. These
interoperability. The information generated is stored in what is queries that can be varied between a simple project
called a 'Unified Virtual Folder'. In fact, a UVF for each patient management task such as adding / deleting or an update of any
must be created. project (EHR of a patient), and a complex task such as a task of
search. In fact, our system allows to search the most relevant
Analysis and Treatment Module: who receives the user's possible projects compared to a search query entered by the
query in order to analyze it and perform the treatments required user and calculating the degree of similarity between them.
to respond. This module is as an intermediary between the user
interface and the module that has already presented (Data and Practically, when the user of the system makes a request
Knowledge Module). and launches the search operation, the system explores the
basis of the UVF to find the UVF files relevant to the request.
Great importance is also attributed to the metadata of the UVFs
stored in the database due to their active roles to accelerate the with the object and its url. Based on these Urls, this module
research process. At this stage, a list of UVF files is selected. will be able to retrieve records of medical examinations of each
Thereafter, each UVF is transformed into a Semantic patient. Each recovered document will be at the entrance of a
Mediation File (SMF: Semantic Mediation XML) through an unification process that aims to reformulate the model of
SMF generator and by taking into consideration the user document. The result of this process is a set of files Document
profile. This le contains the metadata detailing the object Medical Unified (DocMU).
context, the information on the source document associated

Fig. 3. Analysis and Treatment Module

The result of this process is a set of Document Medical The result of this comparator is depicted in a file called the
Unified (DocMU) files associated with a patient, which file of similarity in the form of a table, whose columns present
represents the concepts of medical documents in a structured medical episodes (diseases) relating to a patient given (in the
way and which are all with a unique model and a uniform case of basic calculation ) and parameters medical tests in case
structure. These DocMUs are subsequently passed to a second of advanced calculation. And the lines present a list of patients
process called ’the grouping process’ which in turn make (patient identifiers) that must be compared with the patient
together those in a new file called the Document Medical target.
unified Patient (DocMUP) file. Therefore, this file is the
concatenation of different DocMUs of the same patient and 3) Querying and Visualization Module
used for comparison with other DocMUPs other patients. This module rep- resents the HMI part of our mediator. It
The resulting DocMUP with the corresponding SMF files provides a set of graphical interfaces that allow the user to
are passed to the last process in this module, the process interact and communicate with the system. Through these
of calculating similarity that aims to calculate the similarity interfaces, the user is able to consult a database containing a
between the different projects. massive amount of Electronic Health Records, enter queries
The result of calculation will be in the form of a file called and view the results.
File of Similarity (FoS). The similarity calculation process
includes two steps of calculation: The data from the sources and heterogeneous systems that
are recovered and described in the format of a UVF are
 Basic calculation (default): where the role of represented graphically in the form of a card that we call it
comparator is to count the common medical episodes ’Iconic Information Card’. The latter represents a synoptic
(diseases which present in the SMF files) between the history of the project of a patient based on an iconic
two files to compare. representation of events and an axis of time on which these
 Advanced calculation: the role of comparator in this events are placed. In fact, the IC can be viewed according to
level is not calculate diseases Commons, hands farther different modes of display:
than that, he is interested in the tests, medical  The Mono Project Visualization: This mode presents
examinations, and their parameters (values) that are in the simplest case, when the user want to view a
the file DocMUP. medical record of a single patient. In this mode of
visualization, the EHR is displayed in the form of an display the result of research in the form of a list. This
iconic information card, which is used to represent a mode of visualization presents to the user a synthetic
synoptic history of the patient from his birth until his overview on similar to a health record given folders. In
death. this mode, the projects are displayed in a table, or each
 The Bi-Project visualization: allows the user to line consists of a medical record.
visualize two health records of two distinct patients,  The Metaphorical visualization: this mode allows to
which thus gives him the opportunity to compare these display on a card of a human body the different
two files. diseases of a patient, which allows the user to have an
 The Multi-projects visualization: follow-up to an overview of the patient profile of more quickly way
operation to search for similar projects, the and more expressive.
visualization multi-project sub-module intervenes and

Fig. 4. Querying and Visualization Module

opinions or suggestions about the degrees of usefulness and


usability of systems.
V. EVALUATION OF THE MPM SYSTEM To start a test, we described to the user project and the
The evaluation of the efficiency of an information system is purpose of the test. We asked him then some information on
one of the key issues in research on this field. Also, researchers his profile. We realize a brief demonstration of the tool, in
and practitioners in the field of health informatics recognize the order to give the user the discovery of the features. Thereafter,
importance of the evaluation of Healthcare Information the user is invited to complete the questionnaire and he could
Systems (HIS). Ammenwerth et al [16] defined the evaluation also record his suggestions and propose improvements.
of SIS as “the act of measuring or exploring properties of a The results of evaluations obtained are rather encouraging.
health information system (in planning, development, Clinicians have found all the features useful and it's very easy
implementation, or operation), the result of which informs a to handle the interface. The main criticisms collected focused
decision to be made concerning that system in a specific on the lack of the vocal aspect interfaces. Healthcare
context”. We present in this section the tests carried out with professionals would find it very relevant to integrate the notion
different healthcare professionals to assess the MPM prototype. of speech in interfaces to make them save time.
We are only interested in the usefulness and usability of Although 70% of clinicians said they have not used this
visualization capabilities, graphic objects filtering, research type of tools, they found the general use of the application
similar cases and comparison between different health records. rather very simple to 80%. Only 10% of clinicians said they
Our evaluation approach is qualitative. It involves had really search features (the others have found more easily)
collecting a set of qualitative data through a questionnaire and and only 20% of clinicians have confessed to handling errors.
analyze these data. A questionnaire is the most common Manipulation, iconic visualization and comparison of projects
method for evaluating an HMI (Human Machine Interface). as a whole seems to be rather very easy to understand. What
This technique is to provide a questionnaire for users to was most encouraging is that clinicians have all said that this
evaluate and rate the HMI through a series of questions. Our prototype helps immensely to develop plans for the proper care
choice fell on the questionnaire because it allows to gather and communicate more easily with other professionals of
health through the functionality of MPM system.
Fig. 5. The GUI of the MPM system

[4] W. Litwin, L. Mark, and N. Roussopoulos, “Interoperability of multiple


autonomous databases,” ACM Computing Surveys (CSUR), vol. 22,
VI. CONCLUSION no. 3, pp. 267–293, 1990.
[5] B. Bizimana, “Interopérabilité des éléments de métadonnées: vers une
In the last years, a considerable number of research efforts approche sémantique,” 2009.
have studied the means and ways to improve the collection and [6] N. Friesen, “Semantic & syntactic interoperability for learning object
metadata,” Metadata in Practice. Chicago: ALA Editions, 2004.
dissemination of medical information useful to the patient’s [7] P. Degoulet, M. Fieschi, and C. Attali, “Les enjeux de l’interopérabilité
health care. These efforts have resulted to the emergence of sémantique dans les systèmes d’information de santé,” Informatique et
many and multiple healthcare information systems, which are gestion médicalisée, vol. 9, pp. 203–212, 1997.
generally heterogeneous and distributed. Often, these systems [8] S. Heiler, “Semantic interoperability,” ACM Computing Surveys
(CSUR), vol. 27, no. 2, pp. 271–273, 1995.
need to exchange information to ensure good coordination of [9] G. Wiederhold, “Mediators in the architecture of future information
care and patient monitoring by obtaining a global overview of a systems,” Computer, vol. 25, no. 3, pp. 38–49, 1992.
patient’s medical record. However, data exchange and [10] G. Gardarin, A. Mensch, and A. Tomasic, “An introduction to the e-
communication between them is very complex and difficult xmlmedia component suite,” EDBT’2002 proceedings to appear, 2002,
pp 297-306.
due to their heterogeneity. To address this problem, we note [11] C. Baru, A. Gupta, B. Ludäscher, R. Marciano, Y. Papakonstantinou,
that there are some platforms, which help to communication, P. Velikhov, and V. Chu, “Xml-based information mediation with mix,”
such as cross-database approach, federated approaches and in ACM SIGMOD Record, vol. 28, no. 2. ACM, 1999, pp. 597–599.
mediator approaches. In this context, we propose in this paper a [12] M.-C. Rousset, A. Bidault, C. Froidevaux, H. Gagliardi, F. Goasdoué,
C. Reynaud, and B. Safar, “Construction de médiateurs pour intégrer
new semantic mediator which allows to unify the des sources dinformation multiples et hétérogènes: le projet picsel,”
representation of the information related to a patient and which Revue I3 (Information-Interaction-Intelligence), 2002.
are initially heterogeneous and distributed in a Virtual Folder [13] T. Gruber, I. L. L. Ontology, and M. T. Özsu, “Encyclopedia of database
Unified and present it in an iconic format and unified systems,” Ontology, 2009.
chronological and that through the Iconic Information Card. [14] D. Beneventano, S. Bergamaschi, F. Guerra, and M. Vincini, “The
momis approach to information integration,” 2001
[15] J. L. Ambite, Y. Arens, L. Gravano, V. Hatzivassiloglou, E. Hovy, J.
REFERENCES Klavans, A. Philpot, U. Ramachandran, J. Sandhaus, A. Singla,
[1] J. Beniger, The control revolution: Technological and economic origins “Simplifying data access: The energy data collection (edc) project,” in
of the information society. Harvard University Press, 2009. Proceedings of the 2000 annual national conference on Digital
[2] S. Sassi, C. Verdier, and A. Flory, “Approche smantique multi-niveaux government research. Digital Government Society of North America,
pour la reprsentation et le partage des informations dans la gestion 2000, pp. 1–11.
de projet,” in International Conference on Information Systems and [16] E. Ammenwerth, J. Brender, P. Nykanen, H.-U. Prokosch, M. Rigby, J.
Economic Intelligence (SIIE’2009), 2009, pp 301-312. Talmon, HIS-EVAL workshop participants. Visions and strategies to
[3] I. C. S. S. C. Committee, IEEE S t a n d a r d C o m p u t e r improve evaluation of health information systems: reflections and
Dict ionar y : A Compilation of IEEE Standard Computer lessons based on the HIS-EVAL workshop in Innsbruck, Int. J. Med. Inf.
Glossaries, 610, ser. ANSI / IEEE Std. IEEE, 1990. 73 (6) (2004) 479–491.

View publication stats

You might also like