An Expert System for the Evaluation of EDSS
in Multiple Sclerosis
Mauro Gaspari, Gianluigi Roveda, Cinzia Scandellari, Sergio Stecchi
Technical Report UBLCS-2001-05
February 2001
Department of Computer Science
University of Bologna
Mura Anteo Zamboni 7
40127 Bologna (Italy)
The University of Bologna Department of Computer Science Research Technical Reports are available in
gzipped PostScript format via anonymous FTP from the area ftp.cs.unibo.it:/pub/TR/UBLCS or via
WWW at URL http://www.cs.unibo.it/. Plain-text abstracts organized by year are available in the
directory ABSTRACTS. All local authors can be reached via e-mail at the address last-name@cs.unibo.it.
Questions and comments should be addressed to tr-admin@cs.unibo.it.
Recent Titles from the UBLCS Technical Report Series
99-21 Real-Time Traffic Transmission Over the Internet, Furini, M., Towsley, D., November 1999.
99-22 On the Expressiveness of Event Notification in Data-Driven Coordination Languages, Busi, N., Zavattaro,
G., December 1999.
2000-1 Compositional Asymmetric Cooperations for Process Algebras with Probabilities, Priorities, and Time,
Bravetti, M., Bernardo, M., January 2000 (Revised February 2000).
2000-2 Compact Net Semantics for Process Algebras, Bernardo, M., Busi, N., Ribaudo, M., March 2000 (Revised
December 2000).
2000-3 An Asynchronous Calculus for Generative-Reactive Probabilistic Systems, Aldini, A., Bravetti, M., May
2000 (Revised September 2000).
2000-4 On Securing Real-Time Speech Transmission over the Internet, Aldini, Bragadini, Gorrieri, Roccetti, May
2000.
2000-5 On the Expressiveness of Distributed Leasing in Linda-like Coordination Languages, Busi, N., Gorrieri, R.,
Zavattaro, G., May 2000.
2000-6 A Type System for JVM Threads, Bigliardi, G., Laneve, C., June 2000.
2000-7 Client-centered Load Distribution: a Mechanism for Constructing Responsive Web Services, Ghini, V.,
Panzieri, F., Roccetti, M., June 2000.
2000-8 Design and Analysis of RT-Ring: a Protocol for Supporting Real-time Communications, Conti, M., Do-
natiello, L., Furini, M., June 2000.
2000-9 Performance Evaluation of Data Locality Exploitation (PhD Thesis), D’Alberto, P., July 2000.
2000-10 System Support for Programming Object-Oriented Dependable Applications in Partitionable Systems (PhD
Thesis), Montresor, A., July 2000.
2000-11 Coordination: An Enabling Technology for the Internet (PhD Thesis), Rossi, D., July 2000.
2000-12 Coordination Models and Languages: Semantics and Expressiveness (PhD Thesis), Zavattaro, G., July
2000.
2000-13 Jgroup Tutorial and Programmer’s Manual, Montresor, A., October 2000.
2000-14 A Declarative Language for Parallel Programming, Gaspari, M., October 2000.
2000-15 An Adaptive Mechanism for Securing Real-time Speech Transmission over the Internet, Aldini, A., Gorri-
eri, R., Roccetti, M., November 2000.
2000-16 Enhancing Jini with Group Communication, Montresor, A., Babaoglu, O., Davoli, R., December 2000
(Revised January 2001).
2000-17 Online Reconfiguration in Replicated Databases Based on Group Communication, Bartoli, A., Kemme, B.
Babaoglu, O., December 2000.
2001-1 Design and Analysis of Protocols and Resources Allocation Mechanisms for Real-Time Applications (Ph.D.
Thesis), Furini, M., January 2001.
2001-2 Formalization, Analysis and Prototyping of Mobile Code Systems (Ph.D. Thesis), Mascolo, C., Janaury
2001.
2001-3 Nature-Inspired Search Techniques for Combinatorial Optimization Problems (Ph.D. Thesis), Rossi, C.,
Janaury 2001.
2001-4 Desktop 3d Interfaces for Internet Users: Efficiency and Usability Issues (Ph.D. Thesis), Pittarello, F., Jan-
uary 2001.
2001-5 An Expert System for the Evaluation of EDSS in Multiple Sclerosis, Gaspari, M., Roveda, G., Scandellari,
C., Stecchi, S., February 2001.
An Expert System for the Evaluation of EDSS
in Multiple Sclerosis
Mauro Gaspari1, Gianluigi Roveda1, Cinzia Scandellari2, Sergio Stecchi2
Technical Report UBLCS-2001-05
February 2001
Abstract
Multiple Sclerosis is still an unsolved disease and although several advances have been achieved in the last
few years some problems such as the ethiology and the prognostic criteria are unknown or do still not have
a satisfying solution. Several experimental trials of therapy in multiple sclerosis are in course in order
to discover a successful treatment. Most of these experiments use a clinical rating scale named EDSS
as evaluation tool for the effects of drugs. This scale is defined by a set of rules written in english which
provide a numerical quantification of the neurological examination. Althougth EDSS has been widely used
for almost 20 years, its application still depends on the interpretation of the neurologist which performs
the neurological examination, and many applications of the scale done by different neurologist on the same
patient can give different results. This is a serious problem for international trials because they lack of a
reliable measure of the effects of drugs. Here we present an expert system for the automatic evaluation of
EDSS in multiple sclerosis which has been developed to overcome this problem. The expert system exploits
an explicit representation of EDSS rules, it is able to explain its conclusions, and it provides a revision tool
to support the user if no satifying solution can be reached. Using this expert system clinical trials based on
EDSS can benefit of a more reliable evaluation tool providing more valuable results.
1. Dipartimento di Scienze dell’Informazione, University of Bologna, Via Mura Anteo Zamboni, 7, 40127 Bologna, Italy.
e-mail: gaspari@cs.unibo.it
2. Multiple Sclerosis Centre of Villa Mazzacorati, Bologna Italy. e-mail:centrosmbo@libero.it
1
1 Introduction
Multiple Sclerosis is still an unsolved disease and although several advances have been achieved
in the last few years some problems such as the ethiology and the prognostic criteria are un-
known or does still not have a satisfying solution. The efforts of contemporary multiple sclerosis
research are oriented to develop complex strategies based on ideas originating from immunology,
neurobiology, brain imaging, and animal modelling. The advent of these increasingly complex
strategiesmakes clinical trials difficult to control, in fact, numerical powerful studies on uncertain
markers are needed involving multicenter international collaboration for reaching a critical mass
of patients. Moreover the ExpandedDisability Status Scale (EDSS)which is the most widely-used
measure for the clinical evaluation of patients involved in trials has several limitations and it not
completely reliable [15], EDSS is a clinical rating scale defined by a set of rules (20) written in
english which provide a numerical quantification of the neurological examination. Since natural
language is intrinsically ambiguous many applications of the scale done by different neurologists
on the same patient can give different results. This is a serious problem because in multicenter
and international trials the EDSS evaluation is always done by different neurologists. Despite
these limitations of EDSS, all the attempt to replace it with new measures failed and presumably
EDSS will be the standard tool for the next decade [18, 15].
In this scenario a fundamental issue becomes to improve EDSS reliability. We address this
issue proposing an interactive tool called AEDSS (Automatic EDSS) able to compute EDSS semi-
automatically asking the user when more information is needed. The tool is based on expert
systems technology which provides mechanisms to encode human knowledge and reasoning
capabilities. EDSS rules have been encoded in AEDSS developing a set of production rules which
is very similar to the original EDSS definition of Kurtzke [9]. Thus AEDSS is not just an algorithm
which implements EDSS evaluation, like MS-cane [2], but it is an attempt to automate the actual
Kurtzke definition. AEDSS is able to explain its conclusions and it provides a revision tool to
support the user when no satifying solution can be reached. The expert system is avaliable over
the internet and thus it can be used world-wide.
The paper is organized as follows: in Section 2 we shortly introduce multiple sclerosis and the
main problems concerning the treatment and the course of this disease; in Section 3 we introduce
the expanded disability status scale which has been automated by AEDSS. In Section 4we present
AEDSS and its modules; and in Section 5 we illustrate the experimental results. Finally, in the
discussion section (Section 6) we compare AEDSS with other tools and we stress the possible
impacts of AEDSS in multicenters international trials; in the last section we describe the future
developments of such a research.
2 Multiple Sclerosis
Multiple sclerosis (MS) is an inflammatory disease characterised by demyelinization of the cen-
tral nervous system (CNS) [17]. It is an immunological disease, and even if its pathogenetic
mechanism is well known, some problems such as the ethiology and the prognostic criteria are
unknown or still have not a satisfying solution [3]. Histopathology of MS is defined by a chronic
inflammation and demyelination. Ongoing disease activity is due to an active inflammatory pro-
cess, mainly mediated by T lymphocytes and macrophages, and is associated with blood-barrier
damage. B lymphocytes and plasma cells are present especially in the lesions that occur during
the late chronic stage of the disease. Although all plaques are characterized by demyelination,
the patterns of oligodendroglia destruction and of damage to other tissue elements, such as axons
and astrocytes, are variable in different cases. Oligodendrocytes are less affected by plaques that
develop during the first bouts of the disease than by those plaques arising after several years of
disease duration.
2.1 Diagnosis and Causes of the Disease
For MS the best set of diagnostic criteria is still that of the Schumacher Panel [14]. This essentially
spells out MS as a white matter disease with evidence of continued or repeated clinical activity
UBLCS-2001-05 2
over time. These criteria have been amended by the Poser committee [13] to include laboratory
evidence, the paraclinical evoked responses, and the cerebrospinal fluid (CSF). However the MS
diagnosis is essentially a clinical process based on spatial and temporal dissemination of symp-
toms and lesions.
Several hypothesis have been done on the possible causes of MS, we summarise here some
of them concerning genetics and virology. There is evidence to support the view that MS is a
complex trait determined by both genetic and environmental factors. The genetic component
is reflected in the higher rate of concordance in monozygotic vs dizygotic twins and in familial
recurrence risk data. There is no evidence that any genetic marker, acting alone or in combination,
protect individual from development of the disease. But with the more recent completion of
full genome screens, new regions of potential genetic interest with respect to multiple sclerosis
susceptibility have emerged.
Infectious agents have been postulated as causes of multiple sclerosis for over a century. The
possible role of a virus or viruses is supported by several data. For example viral infections may
precipitate exacerbations of the disease, and patients with MS have abnormal immune responses
to viruses, but no specific virus has still been individuated as a certain cause of the disease.
2.2 Classification of MS Forms
McAlpine and associates [12] have depicted the variation in the natural history of MS in their
textbook. Sixty percent of MS patients have attacks with quite good recovery andminimal deficit,
particularly at the beginning of their illness. Ultimately MS culminates in a progressive course
with relatively fewer exacerbations, yet gradually worsening disability. Fifteen percent of MS
patients have progressiveMS from onset, with no or relatively few attacks, but develop gradually
worsening disability. Finally, there is another group that constitutes 15 or 20% of all MS patients,
who have benign MS, having relatively few attacks early on, without developing any or very
little permanent disability. In summary, the course of the disease can be characterised by the
following forms [11]:
� benign forms, with few attacks and no disability;
� primary-progressive (PP) with an highly disability degree;
� relapsing-remitting (RR) which is characterised by exacerbations which normally remit and
sometimes evolve in secondary progressive forms.
� secondary progressive (SP) having a progressive course with relatively fewer exacerba-
tions, yet gradually worsening disability.
The disability degree of MS patients is normally expressed by its clinical situation which is deter-
mined by a neurological examination. Note that the different courses are all defined observing
the evolution of the disease over time. Neurologist need to measure any changes in the status of
the patient to understand the course of the disease and to determine the best way to treat each
patient.
Several neurological rating scales have been developed for this purpose to give a numerical
value to the results of neurological examinations [1, 15]. These scalesmeasure symptomsworking
on the assumption that symptoms reflect disease severity and activity. The most widely used
scale is EDSS which is the subject of our research and it is described in details in the next section.
3 Expanded Disability Status Scale
One of the first measure of neurological status of MS patients is the Disability Status Scale (DSS)
originally developed in 1955 by John Kurtzke [8], which has 10 grades of clinical impairment due
to MS starting from 0 (normal health) to 10 (death due to MS).
Although DSS has several limitation concerning its expressive power (it was found to miss
small changes in disease severity) it has been used for about 20 years untill Kurtzke presented the
Expanded Disability Status Scale (EDSS) [9] which is the actual standard. EDSS was expanded to
20 steps by adding half steps between steps one to 10.
UBLCS-2001-05 3
