An evolutionary-fuzzy DSS for assessing health status in multiple sclerosis disease

Abstract

Assisted Living provides a long-term care option that combines supportive systems and services for monitoring and assessing the health status with activities of daily living and health care. Daily monitoring of the health status in subjects characterized by chronic and/or degenerative conditions is not possible in all those cases where the disease progression has to be evaluated only by a direct interaction between the patients and the healthcare structures on a regular basis, over time and for life. In this respect, this work proposes an evolutionary-fuzzy decision support system (DSS) for assessing the health status of subjects affected by multiple sclerosis (MS) during the disease progression over time. Such a DSS has been defined and implemented exploiting a novel approach devised to facilitate the design of fuzzy DSSs for medical problems. The approach is aimed at: (i) introducing a set of design criteria to encode the medical knowledge elicited from clinical experts in terms of linguistic variables, linguistic values and fuzzy rules with the final aim of granting the interpretability; (ii) defining a fuzzy inference technique to best fit the structure of medical knowledge and the peculiarities of the medical inference; (iii) defining an evolutionary technique to tune the formalized knowledge by optimizing the shapes of the membership functions for each linguistic variable involved in the rules. An experimental session has been carried out for evaluating, first of all, the approach on five medical databases commonly diffused in literature and for comparing it with other systems. After that, the evolutionary-fuzzy DSS for assessing MS patient's health status has been quantitatively evaluated on 120 patients affected by MS and compared with other approaches. The achieved results have shown that our approach is very effective on the five databases, since it provides, on average, the second highest accuracy when compared to eight tools. Furthermore, as far as the classification of multiple sclerosis lesions is considered, the proposed system has turned out to outperform nine popular tools.

Introduction

Recent advances in biomedical science, health care technologies, and public health measures are radically impacting the management and monitoring of all those diseases that, due to the lack of a cure, have led to certain disability and death in the past. Since these illnesses are transformed into chronic and/or degenerative conditions, the primary benefit is the potential for individuals affected by them to live longer and with higher quality of life. In this respect, Assisted Living provides a long-term care option that combines supportive systems and services for monitoring and assessing the health status with activities of daily living and health care, aiming at providing assistance customized to the patients’ needs so as to enrich their lives and promote independence and well-being.

Monitoring the subject's health status in the daily living is becoming a common practice due to the increasing spread in aging population of chronic diseases, such as obstructive pulmonary, cardiovascular diseases, etc., and has led to the development of low-cost, innovative technological systems to daily support and assist patients, to prevent and control pathology ongoingness, to adjust drug therapies, and to avoid hospitalization. Yet, the daily health status monitoring in subjects with chronic and/or degenerative conditions is not always possible as it strongly depends on disease-specific features: an example is that of neurodegenerative diseases, which require specific approaches for assessing the subject's health status, since their progression can be evaluated only by a direct interaction between the patients and the healthcare structures on a regular basis, over time and for life.

Concerning the latter typology of pathology, the most representative example is represented by multiple sclerosis (MS), characterized by multiple demyelinated lesions, involving the brain and spinal cord, that cause damage or destruction of myelin surrounding nerve fibers, and, thus, interrupt communications between the nerves and the rest of the body [1]. It has a strong impact on the patient's health status since it produces neurological dysfunctions, such as numbness, impaired vision, loss of balance, weakness, bladder dysfunction, and psychological changes. Many MS cases evolve over a long period (20–30 years) with remissions and exacerbations, but, in almost half of all cases, it relentlessly progresses to severe disability and premature death [2]. Thus, disease outcome is represented by chronic and/or degenerative conditions, which are highly variable between affected individuals. Unfortunately, there is a lack of prognostic markers: indeed, for an individual patient, the severity outcome and the rate of progression of MS are impossible to predict.

Moreover, there is no way for monitoring the health status of MS patients in their daily living since the only manner to control the disease progression relies on clinical examination supported by laboratory investigations including magnetic resonance imaging (MRI) to visualize lesions both in the course of MS and in the assessment of treatment effects [3], [4]. For assessing the patient health status correctly, demonstration of distribution of lesions in both time and space is necessary, delaying the time for an appropriate follow-up [5]. Indeed, the use of MR images as MS marker requires the expert's knowledge and intervention to identify MS lesions; nevertheless, such a task is very thorny and time-consuming due to the huge amount of MR images to be examined and to the variable number, size and spatial distribution of MS lesions per image.

In the past different approaches have been proposed, and, in particular, operator-assisted methods based on local thresholding and automated methods assessing multiple parameters have been successfully employed for assessing the health status in MS patients by identifying lesions [6], [7], [8], [9]. Nevertheless, these methods suffer from one important limitation: they rely on mathematical models based on thresholding to classify MS lesions. Hence, they neither take into account the fuzziness of input data nor reproduce the expert's decision-making process applied in a vague-laden domain such as medicine. In fact, the decision-making model every physician has in mind to perform heuristic diagnosis is often pervaded by uncertainty and vagueness. Thus expert knowledge abounds with imprecise formulations that do not depend on rhetorical inability, but are an intrinsic part of expert knowledge acquired through laborious experience [10]. Any formalism disallowing uncertainty, such as crisp mathematical models, is hence inapt to capture this knowledge [10] and can represent an unrealistic oversimplification of reality, leading to possible wrong interpretations when compared to a direct observation.

Fuzzy Logic [11] has widely demonstrated its capability to overcome such critical issues in medical applications, and many decision support systems (DSSs) based on it have been proposed in literature [12], [13], [14]. This is due to the fact that Fuzzy Logic formalism is suitable to deal with the imprecision intrinsic to many medical problems, so as to offer a more realistic interpretation for the medical inference. Although, in theory, a DSS based on Fuzzy Logic could be proficiently used for identifying lesions in the assessment of MS patient's health status, in practice, the design of a Fuzzy DSS is a complex multi-step process in which the most usual way for collecting medical knowledge is asking the expert to write “if-then” rules. Yet, neuroradiologists usually describe their knowledge by means of incomplete rules that typically model pieces of positive evidence only. Moreover, after formalizing the expert's knowledge under the form of rules, the designer has to choose the shape and location of membership functions for all the linguistic values related to all the linguistic variables involved. This requires both medical expertise and technical intervention along with great effort to identify which among the design choices are suited to the given problem.

To face all these issues, this work proposes an evolutionary-fuzzy DSS for assessing the health status of subjects affected by MS during the disease progression over time, which is aimed at supporting clinicians in the identification of MS lesions. Such a system has been defined and implemented by exploiting a novel approach devised to ease the design of fuzzy DSSs so as to involve the medical practitioner in the definition of the domain knowledge only. Such an approach is aimed at: (i) introducing a set of design criteria to encode the high-level, specialized medical knowledge elicited from clinical experts in terms of linguistic variables, linguistic values and rules with the final aim of granting the interpretability; (ii) defining a fuzzy inference technique to best fit the structure of medical knowledge and the peculiarities of the medical inference; (iii) defining an adaptive technique based on an evolutionary algorithm, i.e. differential evolution [15], to tune the formalized knowledge by optimizing the shapes of the membership functions for each linguistic variable involved in the rules.

In order to address the three above reported goals, the proposed approach has been conceived as general purpose and structured and described in terms of three design stages, respectively, knowledge representation, knowledge reasoning and knowledge tuning. Moreover, an experimental session has been carried out for evaluating quantitatively its performance on five medical databases commonly diffused in literature. In particular, five prototypal DSSs have been built in accordance with the approach, each of them being associated to a different database, in order to compare their results against those provided by a set of widely used machine learning methods on the same set of databases. Successively, the approach has been specifically applied to build an evolutionary-fuzzy DSS for assessing MS patient's health status. This latter has been quantitatively evaluated on 120 patients affected by MS and compared with other existing tools for assessing its effectiveness.

To the best of our knowledge, none of existing methods and systems proposed in literature adopts a similar hybrid approach for assessing MS patient's health status. Many approaches in literature [16], [17] use fuzzy logic both to segment MR images and to classify MS lesions, i.e. to cluster single pixels/voxels into homogeneous groups and identify them as belonging to a brain tissue or to an MS lesion. Differently, we are not interested in segmenting MR images, rather our approach is focused on the definition of a DSS that uses fuzzy logic, and, in particular, fuzzy rules provided by doctors, only for the classification of MS lesions by starting from the results of a preliminary segmentation step. Of course, the DSS performance depends on the reliability of the previous segmentation step, and, as a result, the proposed DSS complements existing efforts already carried out for the segmentation of MS lesions. More specifically, if we changed the segmentation algorithm a different input dataset could be created, which would probably contain a different set of possible lesions. Nonetheless, the DSS would act in exactly the same way.