Xudong Lu

On mining clinical pathway patterns from medical behaviors

OBJECTIVE Clinical pathway analysis, as a pivotal issue in ensuring specialized, standardized, normalized and sophisticated therapy procedures, is receiving increasing attention in the field of medical informatics. Clinical pathway pattern mining is one of the most important components of clinical pathway analysis and aims to discover which medical behaviors are essential/critical for clinical pathways, and also where temporal orders of these medical behaviors are quantified with numerical bounds. Even though existing clinical pathway pattern mining techniques can tell us which medical behaviors are frequently performed and in which order, they seldom precisely provide quantified temporal order information of critical medical behaviors in clinical pathways. METHODS This study adopts process mining to analyze clinical pathways. The key contribution of the paper is to develop a new process mining approach to find a set of clinical pathway patterns given a specific clinical workflow log and minimum support threshold. The proposed approach not only discovers which critical medical behaviors are performed and in which order, but also provides comprehensive knowledge about quantified temporal orders of medical behaviors in clinical pathways. RESULTS The proposed approach is evaluated via real-world data-sets, which are extracted from Zhejiang Huzhou Central hospital of China with regard to six specific diseases, i.e., bronchial lung cancer, gastric cancer, cerebral hemorrhage, breast cancer, infarction, and colon cancer, in two years (2007.08-2009.09). As compared to the general sequence pattern mining algorithm, the proposed approach consumes less processing time, generates quite a smaller number of clinical pathway patterns, and has a linear scalability in terms of execution time against the increasing size of data sets. CONCLUSION The experimental results indicate the applicability of the proposed approach, based on which it is possible to discover clinical pathway patterns that can cover most frequent medical behaviors that are most regularly encountered in clinical practice. Therefore, it holds significant promise in research efforts related to the analysis of clinical pathways.

Discovery of clinical pathway patterns from event logs using probabilistic topic models

Discovery of clinical pathway (CP) patterns has experienced increased attention over the years due to its importance for revealing the structure, semantics and dynamics of CPs, and to its usefulness for providing clinicians with explicit knowledge which can be directly used to guide treatment activities of individual patients. Generally, discovery of CP patterns is a challenging task as treatment behaviors in CPs often have a large variability depending on factors such as time, location and patient individual. Based on the assumption that CP patterns can be derived from clinical event logs which usually record various treatment activities in CP executions, this study proposes a novel approach to CP pattern discovery by modeling CPs using mixtures of an extension to the Latent Dirichlet Allocation family that jointly models various treatment activities and their occurring time stamps in CPs. Clinical case studies are performed to evaluate the proposed approach via real-world data sets recording typical treatment behaviors in patient careflow. The obtained results demonstrate the suitability of the proposed approach for CP pattern discovery, and indicate the promise in research efforts related to CP analysis and optimization.

Using Recommendation to Support Adaptive Clinical Pathways

Clinical pathways are among the main tools used to manage the quality in health-care concerning the standardization of care processes. This paper deals with a recommendation service to support adaptive clinical pathways. The proposed approach can guide physicians in clinical pathways by providing recommendations on possible next steps based on the measurement of the target patient status and medical knowledge from completed clinical cases. The efficiency and usability of the proposed method is validated by experiments referring to a real data set extracted from Electronic Patient Records. The experimental results indicate that the recommendation service can provide its users with advice rationales that remain consistent even when patient status has changed. This makes adaptive clinical pathways possible.

Integrated Visualization of Multi-Modal Electronic Health Record Data

Due to the rapid development of modern medical information technology, the modality and amount of electronic health record (EHR) data can be incredibly large. Clinicians must rely on the comparison and inter-confirmation among those complex multi-modal EHR data to make accurate clinical diagnoses, treatment plans and preventive measures. Integrated visualization is needed to extract valuable information from EHR data. This paper first studies a unified structure to represent the multi-modal EHR data as various kinds of clinical acts. Then it brings forward a method for the integrated visualization of those data according to the two dimensional act-time relationships. Finally, an integrated viewer is designed and implemented. By this means, we provide the clinicians an overall scene of the patients personal history, present health status and future care plans.

Online Treatment Compliance Checking for Clinical Pathways

Compliance checking for clinical pathways (CPs) is getting increasing attention in health-care organizations due to stricter requirements for cost control and treatment excellence. Many compliance measures have been proposed for treatment behavior inspection in CPs. However, most of them look at aggregated data seen from an external perspective, e.g. length of stay, cost, infection rate, etc., which may provide only a posterior impression of the overall conformance with the established CPs such that in-depth and in near real time checking on the compliance of the essential/critical treatment behaviors of CPs is limited. To provide clinicians real time insights into violations of the established CP specification and support online compliance checking, this article presents a semantic rule-based CP compliance checking system. In detail, we construct a CP ontology (CPO) model to provide a formal grounding of CP compliance checking. Using the proposed CPO, domain treatment constraints are modeled into Semantic Web Rule Language (SWRL) rules to specify the underlying treatment behaviors and their quantified temporal structure in a CP. The established SWRL rules are integrated with the CP workflow such that a series of applicable compliance checking and evaluation can be reminded and recommended during the pathway execution. The proposed approach can, therefore, provides a comprehensive compliance checking service as a paralleling activity to the patient treatment journey of a CP rather than an afterthought. The proposed approach is illustrated with a case study on the unstable angina clinical pathway implemented in the Cardiology Department of a Chinese hospital. The results demonstrate that the approach, as a feasible solution to provide near real time conformance checking of CPs, not only enables clinicians to uncover non-compliant treatment behaviors, but also empowers clinicians with the capability to make informed decisions when dealing with treatment compliance violations in the pathway execution.

The Architecture of Enterprise Hospital Information System

Because of the complexity of the hospital environment, there exist a lot of medical information systems from different vendors with incompatible structures. In order to establish an enterprise hospital information system, the integration among these heterogeneous systems must be considered. Complete integration should cover three aspects: data integration, function integration and workflow integration. However most of the previous design of architecture did not accomplish such a complete integration. This article offers an architecture design of the enterprise hospital information system based on the concept of digital neural network system in hospital. It covers all three aspects of integration, and eventually achieves the target of one virtual data center with Enterprise Viewer for users of different roles. The initial implementation of the architecture in the 5-year Digital Hospital Project in Huzhou Central hospital of Zhejiang Province is also described

Integration of Medical Information Systems Based on Virtual Database and Web Services

In the healthcare domain, there exists a large number of heterogeneous medical information systems on account of the complexity of the medical environment. To support easier information exchange between these systems, it becomes essential to integrate separated medical information and allow it to be exchanged and retrieved through Internet. In this paper, on data layer, via XML Schema mapping we build a virtual database for medical data sharing, which implements to update, insert and delete sharing medical data based on querying by XQuery; On application layer, we build web-based application integration, which allows medical information systems uniformly exchange information easily by Web Services.

A Term-Mapping Framework for Data Mining in Heterogeneous Medical Data Sources

Medical informatics developed dramatically in recent years, and the resulting large volumes of data may contain useful information. Data mining is used to gain more value from data sources, but it would not work well using the same way in different medical information systems. It is one of the reasons that the data descriptions are various in different systems. This paper presents Term-mapping framework which can be applied in data mining. The framework connects data sources and standard terminology, and it would improve the efficiency of data mining remarkably.

An openEHR based approach to improve the semantic interoperability of clinical data registry

BackgroundClinical data registry is designed to collect and manage information about the practices and outcomes of a patient population for improving the quality and safety of care and facilitating novel researches. Semantic interoperability is a challenge when integrating the data from more than one clinical data registry. The openEHR approach can represent the information and knowledge semantics by multi-level modeling, and it advocates the use of collaborative modeling to facilitate reusing existing archetypes with consistent semantics so as to be a potential solution to improve the semantic interoperability.MethodsThis paper proposed an openEHR based approach to improve the semantic interoperability of clinical data registry. The approach consists of five steps: clinical data registry meta-information collection, data element definition, archetype modeling, template editing, and implementation. Through collaborative modeling and maximum reusing of existing archetype at the archetype modeling step, the approach can improve semantic interoperability. To verify the feasibility of the approach, this paper conducted a case study of building a Coronary Computed Tomography Angiography (CCTA) registry that can interoperate with an existing Electronic Health Record (EHR) system.ResultsThe CCTA registry includes 183 data elements, which involves 20 archetypes. A total number of 45 CCTA data elements and EHR data elements have semantic overlap. Among them, 38 (84%) CCTA data elements can be found in the 10 reused EHR archetypes. These corresponding clinical data can be collected from the EHR system directly without transformation. The other 7 (16%) CCTA data elements correspond to one coarse-grained EHR data elements, and these clinical data can be collected with mapping rules. The results show that the approach can improve semantic interoperability of clinical data registry.ConclusionsUsing an openEHR based approach to develop clinical data registry can improve the semantic interoperability. Meanwhile, some challenges for broader semantic interoperability are identified, including domain experts’ involvement, archetype sharing and reusing, and archetype semantic mapping. Collaborative modeling, easy-to-use tools, and semantic relationship establishment are potential solutions for these challenges. This study provides some experience and insight about clinical modeling and clinical data registry development.

A general-purpose representation of biosignal data based on MFER and CDA

Electric mechanical or chemical signals of biological origin delivered by human bodies can always be of interest for diagnosis, patient monitoring, and biomedical research. Such biomedical signals, namely biosignals, are usually presented by flat file format or proprietary format when digitized for acquisition, storage, transfer and analysis. Recent advances in the ubiquitous computing technology make it possible to measure biosignal data anywhere and anytime. But a big challenge resides in the realm of ubiquitous computing, which is the uniform integration and communication of data. This problem also exists in the development of the electronic health record and telemedicine. Standardizing the data presentation is the best solution to resolve these problems. But a general-purpose, effective, user-friendly, and universally accepted data format for biosignal is not available at this moment. This paper proposes a simple and complete representation of biosignal data based on MFER and CDA. It adopts frame structure of MFER to encode the waveform data of biosignal and uses hierarchical structure of CDA to describe its non-waveform data. This composite format not only inherits the advantages of MFER and CDA, but also facilitates the implementation of interpretation of biosignal.