Jose M. Juarez

Temporal similarity measures for querying clinical workflows

OBJECTIVE In this paper, we extend a preliminary proposal and discuss in a deeper and more formal way an approach to evaluate temporal similarity between clinical workflow cases (i.e., executions of clinical processes). More precisely, we focus on (i) the representation of clinical processes by using a temporal conceptual workflow model; (ii) the definition of ad hoc temporal constraint networks to formally represent clinical workflow cases; (iii) the definition of temporal similarity for clinical workflow cases based on the comparison of temporal constraint networks; (iv) the management of the similarity of clinical processes related to the Italian guideline for stroke prevention and management (SPREAD). BACKGROUND Clinical processes are composed by clinical activities to be done by given actors in a given order satisfying given temporal constraints. This description means that clinical processes can be seen as organizational processes, and modeled by workflow schemata. When a workflow schema represents a clinical process, its cases represent different instances derived from dealing with different patients in different situations. With respect to all the cases related to a workflow schema, each clinical case can be different with respect to its structure and to its temporal aspects. Clinical cases can be stored in clinical databases and information retrieval can be done evaluating the similarity between workflow cases. METHODOLOGY We first describe a possible approach to the conceptual modeling of a clinical process, by using a temporally extended workflow model. Then, we define how a workflow case can be represented as a set of activities, and show how to express them through temporal constraint networks. Once we have built temporal constraint networks related to the cases to compare, we propose a similarity function able to evaluate the differences between the considered cases with respect to the order and duration of corresponding activities, and with respect to the presence/absence of some activities. RESULTS In this work, we propose an approach to evaluate temporal similarity between workflow cases. The proposed approach can be used (i) to query clinical databases storing clinical cases representing activities related to the management of different patients in different situations; (ii) to evaluate the quality of the service comparing the similarity between a (possibly synthetic) case, perceived as the good one with respect to a given clinical situation, and the other clinical cases; and (iii) to retrieve a particular class of cases similar to an interesting one.

Temporal similarity by measuring possibilistic uncertainty in CBR

Similarity is an essential concept in case-based reasoning (CBR). In domains in which time plays a relevant role, CBR systems require good temporal similarity measures to compare cases. Temporal cases are traditionally represented by a set of temporal features, defining time series and temporal event sequences. In the particular situation where these features are not homogeneous (i.e. combination of qualitative and quantitative information), systems find difficulties in performing the CBR cycle. Furthermore, temporal similarity measures cannot directly apply the efficient time series techniques, requiring new approaches to deal with these heterogeneous sequences. To this end, recent proposals are focused on direct matching between pairs of features within sequences, mainly based on classical distances. However, three limitations to the traditional approaches have been identified: (1) they do not consider the implicit temporal relations amongst all features of the sequence (ignoring a large amount of temporal information); (2) they ignore the uncertainty produced in any process of analogy; (3) they are designed to compare pairs of sequences, limiting their use to basic aspects of the Retrieval step of CBR (no benefits on other CBR steps). Temporal constraint networks have proved to be useful tools for temporal representation and reasoning, and can be easily extended to manage imprecision and uncertainty. An approach to solve similarity problems could be the transformation of these heterogeneous sequences into uncertain temporal relations, obtaining a temporal constraint network. The overall uncertainty of this network can be considered as an effective indicator of the sequences similarity. Therefore, this paper proposes a non-classical approach to measure temporal similarity of cases which are heterogeneous temporal event sequences. Given two or more sequences, the temporal similarity is measured by describing a unique temporal scenario of possibilistic temporal relations and calculating the uncertainty produced.

Multi-objective evolutionary algorithms for fuzzy classification in survival prediction

OBJECTIVE This paper presents a novel rule-based fuzzy classification methodology for survival/mortality prediction in severe burnt patients. Due to the ethical aspects involved in this medical scenario, physicians tend not to accept a computer-based evaluation unless they understand why and how such a recommendation is given. Therefore, any fuzzy classifier model must be both accurate and interpretable. METHODS AND MATERIALS The proposed methodology is a three-step process: (1) multi-objective constrained optimization of a patients data set, using Pareto-based elitist multi-objective evolutionary algorithms to maximize accuracy and minimize the complexity (number of rules) of classifiers, subject to interpretability constraints; this step produces a set of alternative (Pareto) classifiers; (2) linguistic labeling, which assigns a linguistic label to each fuzzy set of the classifiers; this step is essential to the interpretability of the classifiers; (3) decision making, whereby a classifier is chosen, if it is satisfactory, according to the preferences of the decision maker. If no classifier is satisfactory for the decision maker, the process starts again in step (1) with a different input parameter set. RESULTS The performance of three multi-objective evolutionary algorithms, niched pre-selection multi-objective algorithm, elitist Pareto-based multi-objective evolutionary algorithm for diversity reinforcement (ENORA) and the non-dominated sorting genetic algorithm (NSGA-II), was tested using a patients data set from an intensive care burn unit and a standard machine learning data set from an standard machine learning repository. The results are compared using the hypervolume multi-objective metric. Besides, the results have been compared with other non-evolutionary techniques and validated with a multi-objective cross-validation technique. Our proposal improves the classification rate obtained by other non-evolutionary techniques (decision trees, artificial neural networks, Naive Bayes, and case-based reasoning) obtaining with ENORA a classification rate of 0.9298, specificity of 0.9385, and sensitivity of 0.9364, with 14.2 interpretable fuzzy rules on average. CONCLUSIONS Our proposal improves the accuracy and interpretability of the classifiers, compared with other non-evolutionary techniques. We also conclude that ENORA outperforms niched pre-selection and NSGA-II algorithms. Moreover, given that our multi-objective evolutionary methodology is non-combinational based on real parameter optimization, the time cost is significantly reduced compared with other evolutionary approaches existing in literature based on combinational optimization.

Medical knowledge management for specific hospital departments

Medical knowledge representation and management is concerned with how to organise the often vague clinical experience of medical staff required for computable models. However, few knowledge management and acquisition tools have entered routine use, since such tools are not perceived by physicians as part of the clinical information process. An attempt to partially solve this problem, we identify two key aspects of knowledge representation and management tasks. The first is to adopt a medical knowledge standardisation to provide a consistent terminology control and to simplify the integration between knowledge management tools and the health information system. The second is to establish an effective knowledge acquisition process in specific medical fields by adapting knowledge acquisition tools. Therefore, the main goal of this work is to define computational models and to design mechanisms for the effective acquisition and management of medical knowledge in real-life hospital departments. To this end, we analyse the representation of medical knowledge (based on deep-causal models) and the development of knowledge management tools (based on ontologies), integrated within the information processing activities of the clinical user. Finally, we illustrate its applicability in the Intensive Care Unit and Pediatry scenarios.

Computing context-dependent temporal diagnosis in complex domains

Over the years, many Artificial Intelligence (AI) approaches have dealt with the diagnosis problem and its application in complex environments such as medical domains. Model-Based Reasoning (MBR) is one of the approaches that traditionally have tried to solve this problem thanks to its capacity for modelling and reasoning. The consideration of the temporal dimension in these domains is a challenging topic in MBR, especially if temporal imprecision is taken into account. Unfortunately, despite there being many successful MBR systems, there are still two fundamental problems in their development at the aforementioned domains: (1) the degree of dependency between the model used and the domain; and (2) the reutilization of the systems when the domain changes. First this paper proposes a set of basic requirements for the design of Knowledge-Based Systems that will help to solve the problem of temporal diagnosis for environments of high conceptual complexity. From these principles and through a deep analysis of the various approaches present in AI we establish a generic framework that addresses both goals by integrating MBR and ontologies for domain knowledge representation in order to describe a intermediate model representation to facilitate the low dependency between the model and the application domain. Finally, this paper demonstrates the use of the framework by developing a diagnosis system within a real medical environment (Intensive Care Unit) with a step-by-step description of the process, from the architecture through to implementation.

A Case-Based Architecture for Temporal Abstraction Configuration and Processing

In this work we propose a case-based architecture tackling the problem of configuring and processing temporal abstractions (trends and qualitative states) produced from raw time series data. The parameter configuration is a critical problem in many temporal abstraction processes; in several application domains (especially in medical ones), contextual knowledge plays a fundamental role in the time series interpretation. Since defining the right configuration for each possible contextual situation may be impractical, we propose to adopt a case-based approach, where the suitable configuration can be obtained by looking at the most similar already configured case, with respect to the current situation. Configured cases are indexed by means of contextual information. The obtained configuration can then be used as input to a temporal abstraction module, providing a set of qualitative states, trends and suitable combination of both as a result. Cases can then be exploited in the processing of such results as well, by providing an evaluation of the whole abstraction processing, possibly leading to the revision of the case base. The approach is illustrated by means of an example taken from a medical application, concerning the monitoring and evaluation of patients undergoing hemodialysis treatment

Length of stay prediction for clinical treatment process using temporal similarity

In clinical treatment processes, inpatient length of stay (LOS) is not only a readily available indicator of hospital activity, but also a reasonable proxy of resource consumption. Accurate inpatient LOS prediction has strong implications for health service delivery. Major techniques proposed (statistical approaches or artificial neuronal networks) consider a priori knowledge, such as demographics or patient physical factors, providing accurate methods to estimate LOS at early stages of the patient (admission). However, unexpected scenarios and variations are commonplaces of clinical treatment processes that have a dramatical impact on the LOS. Therefore, these predictors should deal with adaptability, considering the temporal evolution of the patient. In this paper, we propose an inpatient LOS prediction approach across various stages of clinical treatment processes. This proposal relies on a kind of regularity assumption demanding that patient traces of the specific treatment process with similar medical behaviors have similar LOS. Therefore, this approach follows a Case-based Reasoning methodology since it predicts an inpatient LOS of a partial patient trace by referring to the past traces of clinical treatment processes that have similar medical behaviors with the current one. The proposal is evaluated using 284 patient traces from the pulmonary infection CTPs, extracted from Zhejiang Huzhou Central Hospital of China.

Spatiotemporal data visualisation for homecare monitoring of elderly people

OBJECTIVE Elderly people who live alone can be assisted by home monitoring systems that identify risk scenarios such as falls, fatigue symptoms or burglary. Given that these systems have to manage spatiotemporal data, human intervention is required to validate automatic alarms due to the high number of false positives and the need for context interpretation. The goal of this work was to provide tools to support human action, to identify such potential risk scenarios based on spatiotemporal data visualisation. METHODS AND MATERIALS We propose the MTA (multiple temporal axes) model, a visual representation of temporal information of the activity of a single person at different locations. The main goal of this model is to visualize the behaviour of a person in their home, facilitating the identification of health-risk scenarios and repetitive patterns. We evaluate the models insight capacity compared with other models using a standard evaluation protocol. We also test its practical suitability of the MTA graphical model in a commercial home monitoring system. In particular, we implemented 8VISU, a visualization tool based on MTA. RESULTS MTA proved to be more than 90% accurate in identify non-risk scenarios, independently of the length of the record visualised. When the spatial complexity was increased (e.g. number of rooms) the model provided good accuracy form up to 5 rooms. Therefore, user preferences and user performance seem to be balanced. Moreover, it also gave high sensitivity levels (over 90%) for 5-8 rooms. Fall is the most recurrent incident for elderly people. The MTA model outperformed the other models considered in identifying fall scenarios (66% of correctness) and was the second best for burglary and fatigue scenarios (36% of correctness). Our experiments also confirm the hypothesis that cyclic models are the most suitable for fatigue scenarios, the Spiral and MTA models obtaining most positive identifications. CONCLUSIONS In home monitoring systems, spatiotemporal visualization is a useful tool for identifying risk and preventing home accidents in elderly people living alone. The MTA model helps the visualisation in different stages of the temporal data analysis process. In particular, its explicit representation of space and movement is useful for identifying potential scenarios of risk, while the spiral structure can be used for the identification of recurrent patterns. The results of the experiments and the experience using the visualization tool 8VISU proof the potential of the MTA graphical model to mine temporal data and to support caregivers using home monitoring infrastructures.

Querying Clinical Workflows by Temporal Similarity

The degree of fulfillment of clinical guidelines is considered a key factor when evaluating the quality of a clinical service. Guidelines can be seen as processes describing the sequence of activities to be done. Consequently, workflow formalisms seem to be a valid approach to model the flow of actions in the guideline and their temporal aspects. The application of a guideline to a specific patient (guideline instance) can be modeled by means of a workflow case. The best (worst) application of a guideline, represented as a reference workflow case, can be used to evaluate the quality of the service, by comparing the optimal case with specific patient instances. On the other hand, the correct application of a guideline to a patient involves the fulfillment of the guideline temporal constraints. Thus, the evaluation of the temporal similarity degree between different workflow cases is a key aspect in evaluating health care quality. In this work, we represent a portion of the stroke guideline using a temporal workflow schema and we propose a method to evaluate the temporal similarity between workflow cases. Our proposal, based on temporal constraint networks, consists of a linear combination of functions to differentiate intra-task and inter-task temporal distances.

Development of a clinical decision support system for antibiotic management in a hospital environment

The rise of infections caused by multidrug-resistant bacteria has become a very important issue for health institutions around the world, urging them to a more appropriate use of antibiotics. Clinical decision support systems have a very important role in this area. We propose the development of a clinical decision support system focused on the antibiotic stewardship program implemented in hospitals. The need for a multi-user perspective, both reactive and proactive behaviours, and the use of many heterogeneous knowledge sources are identified as the main requirements that differentiate this clinical scenario from a decision support point of view. We show that a combination of production rules, ontologies, workflow modelling and subgroup discovery techniques could be used to fulfil these requirements. Finally, we describe a platform on which these techniques will be developed and tested.