Peter J. F. Lucas

A probabilistic and decision-theoretic approach to the management of infectious disease at the ICU

The medical community is presently in a state of transition from a situation dominated by the paper medical record to a future situation where all patient data will be available on-line by an electronic clinical information system. In data-intensive clinical environments, such as intensive care units (ICUs), clinical patient data are already fully managed by such systems in a number of hospitals. However, providing facilities for storing and retrieving patient data to clinicians is not enough; clinical information systems should also offer facilities to assist clinicians in dealing with hard clinical problems. Extending an information systems capabilities by integrating it with a decision-support system may be a solution. In this paper, we describe the development of a probabilistic and decision-theoretic system that aims to assist clinicians in diagnosing and treating patients with pneumonia in the intensive-care unit. Its underlying probabilistic-network model includes temporal knowledge to diagnose pneumonia on the basis of the likelihood of laryngotracheobronchial-tree colonisation by pathogens, and symptoms and signs actually present in the patient. Optimal antimicrobial therapy is selected by balancing the expected efficacy of treatment, which is related to the likelihood of particular pathogens causing the infection, against the spectrum of antimicrobial treatment. The models were built on the basis of expert knowledge. The patient data that were available were of limited value in the initial construction of the models because of problems of incompleteness. In particular, detailed temporal information was missing. By means of a number of different techniques, among others from the theory of linear programming, these data have been used to check the probabilistic information elicited from infectious-disease experts. The results of an evaluation of a number of slightly different models using retrospective patient data are discussed as well.

Bayesian analysis, pattern analysis, and data mining in health care

Purpose of reviewTo discuss the current role of data mining and Bayesian methods in biomedicine and heath care, in particular critical care. Recent findingsBayesian networks and other probabilistic graphical models are beginning to emerge as methods for discovering patterns in biomedical data and also as a basis for the representation of the uncertainties underlying clinical decision-making. At the same time, techniques from machine learning are being used to solve biomedical and health-care problems. SummaryWith the increasing availability of biomedical and health-care data with a wide range of characteristics there is an increasing need to use methods which allow modeling the uncertainties that come with the problem, are capable of dealing with missing data, allow integrating data from various sources, explicitly indicate statistical dependence and independence, and allow integrating biomedical and clinical background knowledge. These requirements have given rise to an influx of new methods into the field of data analysis in health care, in particular from the fields of machine learning and probabilistic graphical models.

Dynamic Bayesian networks as prognostic models for clinical patient management

Prognostic models in medicine are usually been built using simple decision rules, proportional hazards models, or Markov models. Dynamic Bayesian networks (DBNs) offer an approach that allows for the incorporation of the causal and temporal nature of medical domain knowledge as elicited from domain experts, thereby allowing for detailed prognostic predictions. The aim of this paper is to describe the considerations that must be taken into account when constructing a DBN for complex medical domains and to demonstrate their usefulness in practice. To this end, we focus on the construction of a DBN for prognosis of carcinoid patients, compare performance with that of a proportional hazards model, and describe predictions for three individual patients. We show that the DBN can make detailed predictions, about not only patient survival, but also other variables of interest, such as disease progression, the effect of treatment, and the development of complications. Strengths and limitations of our approach are discussed and compared with those offered by traditional methods.

Prognostic methods in medicine

Prognosis--the prediction of the course and outcome of disease processes--plays an important role in patient management tasks like diagnosis and treatment planning. As a result, prognostic models form an integral part of a number of systems supporting these tasks. Furthermore, prognostic models constitute instruments to evaluate the quality of health care and the consequences of health care policies by comparing predictions according to care norms with actual results. Approaches to developing prognostic models vary from using traditional probabilistic techniques, originating from the field of statistics, to more qualitative and model-based techniques, originating from the field of artificial intelligence (AI). In this paper, various approaches to constructing prognostic models, with emphasis on methods from the field of AI, are described and compared.

Computer-assisted decision support for the diagnosis and treatment of infectious diseases in intensive care units

Diagnosing nosocomial infections in critically ill patients admitted to intensive care units (ICUs) is a challenge because signs and symptoms are usually non-specific for a particular infection. In addition, the choice of treatment, or the decision not to treat, can be difficult. Models and computer-based decision-support systems have been developed to assist ICU physicians in the management of infectious diseases. We discuss the historical development, possibilities, and limitations of various computer-based decision-support models for infectious diseases, with special emphasis on Bayesian approaches. Although Bayesian decision-support systems are potentially useful for medical decision making in infectious disease management, clinical experience with them is limited and prospective evaluation is needed to determine whether their use can improve the quality of patient care.

Standardizing research methods for prognostics

Prognostics and health management (PHM) is a maturing system engineering discipline. As with most maturing disciplines, PHM does not yet have a universally accepted research methodology. As a result, most component life estimation efforts are based on ad-hoc experimental methods that lack statistical rigor. In this paper, we provide a critical review of current research methods in PHM and contrast these methods with standard research approaches in a more established discipline (medicine). We summarize the developmental steps required for PHM to reach full maturity and to generate actionable results with true business impact.

Quality checking of medical guidelines through logical abduction

Formal methods have been used in the past for the verification of the correctness of formalised versions of medical guidelines. In this paper a second possible application of the use of formal methods is proposed: checking whether a guideline conforms to global medical quality requirements. It is argued that this allows spotting design errors in medical guidelines, which is seen as a useful application for formal methods in medicine. However, this type of verification may require medical knowledge currently not available within the guidelines, i.e. medical background knowledge. In this paper, we propose a method for checking the quality of a treatment for a disorder, based on the theory of abductive diagnosis. We also examine the medical background knowledge required to be able to quality check a guideline. The method is illustrated by the formal analysis of an actual guideline for the management of diabetes mellitus type 2.

Converting a rule-based expert system into a belief network.

The theory of belief networks offers a relatively new approach for dealing with uncertain information in knowledge-based (expert) systems. In contrast with the heuristic techniques for reasoning with uncertainty employed in many rule-based expert systems, the theory of belief networks is mathematically sound, based on techniques from probability theory. It therefore seems attractive to convert existing rule-based expert systems into belief networks. In this article we discuss the design of a belief network reformulation of the diagnostic rule-based expert system HEPAR. For the purpose of this experiment we have studied several typical pieces of medical knowledge represented in the HEPAR system. It turned out that, due to the differences in the type of knowledge represented and in the formalism used to represent uncertainty, much of the medical knowledge required for building the belief network concerned could not be extracted from HEPAR. As a consequence, significant additional knowledge acquisition was required. However, the objects and attributes defined in the HEPAR system, as well as the conditions in production rules mentioning these objects and attributes, were useful for guiding the selection of the statistical variables for building the belief network. The mapping of objects and attributes in HEPAR to statistical variables is discussed in detail.

Bayesian network modelling through qualitative patterns

In designing a Bayesian network for an actual problem, developers need to bridge the gap between the mathematical abstractions offered by the Bayesian-network formalism and the features of the problem to be modelled. Qualitative probabilistic networks (QPNs) have been put forward as qualitative analogues to Bayesian networks, and allow modelling interactions in terms of qualitative signs. They thus have the advantage that developers can abstract from the numerical detail, and therefore the gap may not be as wide as for their quantitative counterparts. A notion that has been suggested in the literature to facilitate Bayesian-network development is causal independence. It allows exploiting compact representations of probabilistic interactions among variables in a network. In the paper, we deploy both causal independence and QPNs in developing and analysing a collection of qualitative, causal interaction patterns, called QC patterns. These are endowed with a fixed qualitative semantics, and are intended to offer developers a high-level starting point when developing Bayesian networks.

Analysis of notions of diagnosis

Various formal theories have been proposed in the literature to capture the notions of diagnosis underlying diagnostic programs. Examples of such notions are: heuristic classification, which is used in systems incorporating empirical knowledge, and model-based diagnosis, which is used in diagnostic systems based on detailed domain models. Typically, such domain models include knowledge of causal, structural, and functional interactions among modelled objects. In this paper, a new set-theoretical framework for the analysis of diagnosis is presented. Basically, the framework distinguishes between ‘evidence functions’, which characterize the net impact of knowledge bases for purposes of diagnosis, and ‘notions of diagnosis’, which define how evidence functions are to be used to map findings observed for a problem case to diagnostic solutions. This set-theoretical framework offers a simple, yet powerful tool for comparing existing notions of diagnosis, as well as for proposing new notions of diagnosis. A theory of flexible notions of diagnosis, called refinement diagnosis, is proposed and defined in terms of this framework. Relationships with notions of diagnosis known from the literature are investigated.