Search

This paper is based on a panel discussion held at the Artificial Intelligence in Medicine Europe (AIME) conference in Amsterdam, The Netherlands, in July 2007. It had been more than 15 years since Edward Shortliffe gave a talk at AIME in which he characterized artificial intelligence (AI) in medicine as being in its "adolescence" (Shortliffe EH. The adolescence of AI in medicine: will the field come of age in the '90s? Artificial Intelligence in Medicine 1993;5:93-106). In this article, the discussants reflect on medical AI research during the subsequent years and characterize the maturity and influence that has been achieved to date. Participants focus on their personal areas of expertise, ranging from clinical decision-making, reasoning under uncertainty, and knowledge representation to systems integration, translational bioinformatics, and cognitive issues in both the modeling of expertise and the creation of acceptable systems.

Objective: The purpose of this study was to assess the performance of a real-time ("open-end") version of the dynamic time warping (DTW) algorithm for the recognition of motor exercises. Given a possibly incomplete input stream of data and a reference time series, the open-end DTW algorithm computes both the size of the prefix of reference which is best matched by the input, and the dissimilarity between the matched portions. The algorithm was used to provide real-time feedback to neurological patients undergoing motor rehabilitation., Methods and Materials: We acquired a dataset of multivariate time series from a sensorized long-sleeve shirt which contains 29 strain sensors distributed on the upper limb. Seven typical rehabilitation exercises were recorded in several variations, both correctly and incorrectly executed, and at various speeds, totaling a data set of 840 time series. Nearest-neighbour classifiers were built according to the outputs of open-end DTW alignments and their global counterparts on exercise pairs. The classifiers were also tested on well-known public datasets from heterogeneous domains., Results: Nonparametric tests show that (1) on full time series the two algorithms achieve the same classification accuracy (p-value =0.32); (2) on partial time series, classifiers based on open-end DTW have a far higher accuracy (kappa=0.898 versus kappa=0.447;p<10(-5)); and (3) the prediction of the matched fraction follows closely the ground truth (root mean square <10%). The results hold for the motor rehabilitation and the other datasets tested, as well., Conclusions: The open-end variant of the DTW algorithm is suitable for the classification of truncated quantitative time series, even in the presence of noise. Early recognition and accurate class prediction can be achieved, provided that enough variance is available over the time span of the reference. Therefore, the proposed technique expands the use of DTW to a wider range of applications, such as real-time biofeedback systems.

The increasing pressure on Health Care Organizations (HCOs) to ensure efficiency and cost-effectiveness, balancing quality of care and cost containment, will drive them towards a more effective management of medical knowledge derived from research findings. The relation between science and health services has until recently been too casual. The primary job of medical research has been to understand the mechanisms of disease and produce new treatments, not to worry about the effectiveness of the new treatments or their implementation. As a result many new treatments have taken years to become part of routine practice, ineffective treatments have been widely used, and medicine has been opinion rather than evidence based. This results in suboptimal care for patients. Knowledge management technology may provide effective approaches in speeding up the diffusion of innovative medical procedures whose clinical effectiveness have been proved: the most interesting one is represented by computer-based utilization of evidence-based clinical guidelines. As researchers in Artificial Intelligence in Medicine (AIM), we are committed to foster the strategic transition from opinion to evidence-based decision making. Reviews of the effectiveness of various methods of guideline dissemination show that the most predictable impact is achieved when the guideline is made accessible through computer-based and patient specific reminders that are integrated into the clinician's workflow. However, the traditional single doctor-patient relationship is being replaced by one in which the patient is managed by a team of health care professionals, each specializing in one aspect of care. Such shared care depends critically on the ability to share patient-specific information and medical knowledge easily among them. Strategically there is a need to take a more clinical process view of health care delivery and to identify the appropriate organizational and information infrastructures to support this process. Thus, the great challenge for AIM researchers is to exploit the astonishing capabilities of new technologies to disseminate their tools to benefit HCOs by assuring the conditions of knowledge management and organizational learning at the fullest extent possible. To achieve such a strategic goal, a guideline can be viewed as a model of the care process. It must be combined with an organization model of the specific HCO to build patient careflow management systems. Artificial intelligence can be extensively used to design innovative tools to support all the development stages of those systems. However, exploiting the knowledge represented in a guideline to build them requires to extend today's workflow technology by solving some challenging problems.

Workflow Management Systems integrate domain and organisational knowledge to support business processes. When applied to the medical environment, they can be termed "Careflow Management Systems", and may be used to manage care delivery by enhancing co-operation among healthcare professionals. This paper focuses on care delivery based on clinical practice guidelines. Healthcare organisations are very different from industrial or commercial companies: their main goal is not profit, but maintaining and improving the health of the public. Therefore, outcomes are difficult to measure. Firstly, physicians, while playing a variety of roles, are quite independent decision-makers; secondly, the object of the process, i.e. the patient, may be involved in choosing treatment options, and may be treated by different institutions. For these reasons, the standard functionality of typical Workflow Management Systems must be strongly enhanced in order to cope with healthcare delivery needs. A major issue is accounting for exceptions. In most non-clinical settings this is not a problem because processes are very well defined and can often be easily controlled by some higher authority. As explained above, this does not happen in healthcare organisations. Responsibilities are widely shared, and health care professionals may be non-compliant with guidelines for a variety of reasons. The paper presents a classification of possible exceptions, and shows how the sequence of tasks described by a guideline may be altered, at the implementation level, in order to meet actual user needs, while maintaining guideline intentions as much as possible. A terminology server is also exploited towards this end. This work illustrates a prototype of a Careflow Management System based on an international guideline for ischemic stroke treatment, developed by the American Heart Association.

This paper describes a methodology for achieving an efficient implementation of clinical practice guidelines. Three main steps are illustrated: knowledge representation, model simulation and implementation within a health care organisation. The resulting system can be classified as a 'guideline-based careflow management system'. It is based on computational formalisms representing both medical and health care organisational knowledge. This aggregation allows the implementation of a guideline, not only as a simple reminder, but also as an 'organiser' that facilitates health care processes. As a matter of fact, the system not only suggests the tasks to be performed, but also the resource allocation. The methodology initially comprehends a graphical editor, that allows an unambiguous representation of the guideline. Then the guideline is translated into a high-level Petri net. The resources, both human and technological necessary for performing guideline-based activities, are also represented by means of an organisational model. This allows the running of the Petri net for simulating the implementation of the guideline in the clinical setting. The purpose of the simulation is to validate the careflow model and to suggest the optimal resource allocation before the careflow system is installed. The final step is the careflow implementation. In this phase, we show that the 'workflow management' technology, widely used in business process automation, may be transferred to the health care setting. This requires augmenting the typical workflow management systems with the flexibility and the uncertainty management, typical of the health care processes. For illustrating the proposed methodology, we consider a guideline for the management of patients with acute ischemic stroke.

This paper describes the application of a method for the intelligent analysis of clinical time series in the diabetes mellitus domain. Such a method is based on temporal abstractions and relies on the following steps: (i) 'pre-processing' of raw data through the application of suitable filtering techniques: (ii) 'extraction' from the pre-processed data of a set of abstract episodes (temporal abstractions); and (iii) 'post-processing' of temporal abstractions; the post-processing phase results in a new set of features that embeds high level information on the patient dynamics. The derived features set is used to obtain new knowledge through the application of machine learning algorithms. The paper describes in detail the application of this methodology and presents some results obtained on simulated data and on a data-set of four diabetic patients monitored for > 1 year.

Exploiting the information technology may have a great impact on improving cooperation and interoperability among the different professionals taking part to the process of delivering health care services. New paradigms are therefore being devised considering software systems as autonomous agents able to help professionals in accomplishing their duties. To this aim those systems should encapsulate the skills for solving a given set of tasks and possess the social ability to cooperate in order to fetch the required information and knowledge. This paper illustrates a methodology facilitating the development of interoperable intelligent software agents for medical applications and proposes a generic computational model for implementing them. That model may be specialized in order to support all the different information and knowledge related requirements of a Hospital Information System. The architecture is being tested for implementing a prototype system able to coordinate the joint efforts of the professionals involved in managing patients affected by Acute Myeloid Leukemia.

The World-Wide Web (WWW) is increasingly being used as a platform to develop distributed applications, particularly in contexts, such as medical ones, where high usability and availability are required. In this paper we propose a methodology for the development of knowledge-based medical applications on the web, based on the use of an explicit domain ontology to automatically generate parts of the system. We describe a development environment, centred on the LISPWEB Common Lisp HTTP server, that supports this methodology, and we show how it facilitates the creation of complex web-based applications, by overcoming the limitations that normally affect the adequacy of the web for this purpose. Finally, we present an outline of a system for the management of diabetic patients built using the LISPWEB environment.

This paper describes a methodology for representing clinical practice guidelines and facilitating their introduction into the medical routine. Since this methodology can be exploited in a www environment, it can represent the basis for sharing clinical guidelines both between different institutions and between human and software agents cooperating within a clinical context. In addition, the proposed guideline formalization is intended to deal with patient and organization preferences. This goal is achieved by augmenting the guideline with decision analytic models and by linking the guideline with an organizational model of the clinical setting. The designed framework allows guideline development, tailoring and implementation, real-time access to the guideline prescriptions and guideline validation.

This paper presents a method for the identification of the dynamics of non-linear systems by learning from data. The key idea which underlies our approach consists of the integration of qualitative modeling techniques with fuzzy logic systems. The resulting hybrid method exploits the a priori structural knowledge on the system to initialize a fuzzy inference procedure which determines, from the available experimental data, a functional approximation of the system dynamics that can be used as a reasonable predictor of the patient's future state. The major advantage which results from such an integrated framework lies in a significant improvement of both efficiency and robustness of identification methods based on fuzzy models which learn an input output relation from data. As a benchmark of our method, we have considered the problem of identifying the response to the insulin therapy from insulin-dependent diabetic patients: the results obtained are presented and discussed in the paper.

Ignorant Belief Networks (IBNs) are a class of Bayesian Belief Networks (BBNs) able to reason on the basis of incomplete probabilistic information and to incrementally refine the precision of the inferred probabilities as more information becomes available. In this paper, we will describe how can be used to develop a system able to forecast blood glucose concentration in patients affected by insulin dependent diabetes mellitus (IDDM). The major difference between our approach and the traditional ones is that probability distributions over the IBN are not provided by some human expert or by the current literature but they are directly extracted from a clinical database of IDDM patients. This choice capitalizes on the large amount of information generated by the daily control of blood glucose and allows the system to improve the accuracy of predictions as more information becomes available. We will show how, even with a very small subset of the information needed to specify a BBN, the IBN is able to carry out predictions about the future blood glucose concentration in a patient by explicitly taking into consideration the level of ignorance embedded in the network.

The project we describe here is aimed at assisting out-patients affected by Insulin Dependent Diabetes Mellitus. Our approach exploits the usual scheme of diabetic patients management, based on (i) a periodic evaluation of patients' metabolic control performed by the physician, and (ii) patient-tailored tables for self-adjustments of insulin dosages. Following this scheme we have defined a system built on a two-level architecture. The High Level Module exploits both medical knowledge and clinical information in order to assess an insulin protocol, defined in terms of insulin timing, type, and total amount. The High Level Module exchanges information with the Low Level Module in order to define the control actions to be taken at the low level, as well as to periodically evaluate protocol adequacy on the basis of patient data. The goal of the Low Level Module, whose characteristics can be adaptively modified by the High Level Module, is to suggest the next insulin dosage, depending on the actual blood glucose measurement and a certain pre-defined insulin delivery protocol. The Level Control Module is based on an adaptive controller, consisting of a Fuzzy Set Controller and an ARX (Autoregressive eXogenous input) Model. The scheme here presented may be conveniently viewed in a telemedicine context, in which the low level controller is implemented on a portable device communicating to the high level controller, implemented on a remote computer. A preliminary assessment has been performed, analyzing a data set of 60 patients provided by the American Association of Artificial Intelligence, Artificial Intelligence in Medicine Subgroup, and the implementation of the system is currently in progress.

The goal of our work is to facilitate the development of medical knowledge-based systems by providing a library of reusable ontologies. The availability of such a library reduces the amount of knowledge acquisition required to create knowledge bases of new applications, and makes it easier to connect a knowledge-based system to existing data bases. This article presents a case study in constructing such a library. The emphasis is on studying the principles that underly the internal structure of the library as well as on the process of constructing and using the library. We envision that, in the future, application ontologies can be constructed by the selection and refinement of generic ontologies and domain ontologies from such a library.

This paper describes the approach we are pursuing for modeling inferential processes in knowledge-based systems. It is aimed at overcoming the lack of generality affecting many of the systems described in the literature. This mainly happens since the problem-solving method adopted by those systems is too closely tied to the particular domain problem over which the method itself has been modeled. We also describe a system called M-KAT (Medical Knowledge Acquisition Tool) which is useful in simplifying the process of acquiring inferential knowledge. M-KAT relies on an epistemological model of medical reasoning which represents a generalization of most of the problem-solving methods adopted in medical knowledge-based systems. The metarules formalism has been adopted as a mean for representing inferential knowledge and making its acquisition easier, thus allowing the computational implementation of the epistemological model of medical reasoning.

This article describes the major projects going on in Europe in the field of Artificial Intelligence in Medicine. The important role of the Commission of the European Communities in providing the needed resources is stressed throughout the paper. Particular attention is given to the methodological and technological issues addressed by the European research teams, since the results which these teams accomplish are fundamental for a more extensive diffusion of knowledge-based systems in real medical settings. The variety of medical problems tackled shows that there is no field of medicine where the potential of advanced informatics technologies has not yet been assessed.