Duan Huilong

An Integration Approach of Healthcare Information System

Healthcare organizations today are still struggling with the integration challenges of heterogeneous information systems. Those system interfaces are not always compatible out of box, when the number of involved systems increases, complexity increases in an unmanageable way. Decentralized proprietary format data largely restricts the value of healthcare information. In this article, we present a healthcare information system integration approach based on an integration engine to address the integration problems. All the involved systems are connected to the integration engine through different interface adaptors. The integration engine takes the inbound message as the trigger event, generates outbound messages by workflow driver according to the workflow configuration file. The integration engine is also responsible for understanding the message context and converting the healthcare information into a unified format, and then storing it in the clinical data repository. Thus, it can facilitate the utilization and sharing of healthcare information.

Automatic segmentation and classification of human brain images based on TT atlas

It is difficult to automatically segment and classify tomographic images of an actual patients brain. Therefore, many interactive operations are performed. It is very time consuming and its precision is user-dependent. Here, the authors combined a brain atlas and 3D fuzzy image segmentation into the image matching. It can not only find out the precise boundary of an anatomic structure, but also save time in the interactive operation. At first, the anatomic information of the atlas is mapped onto tomograph images of an actual brain with an image matching method. Then, based on the mapping result, a 3D fuzzy structure mask is calculated. With the fuzzy information of anatomic structure, a new method of fuzzy clustering is used to segment and classify the real brain image. There is only a minimum requirement of interaction in the whole process, including removing the skull and selecting some intrinsic point pairs.

A Clinical Document Repository for CDA Documents

To harvest the potential benefits of electronic health record (EHR), the health record data must be in a standard, structured and computer understandable form. The Health Level 7 Clinical Document Architecture (CDA) is an American National Standards Institute (ANSI)-approved standard that specifies the hierarchical structure and semantics of clinical documents for the purpose of exchange. As CDA is used more widely, the efficient management, storage and retrieval of CDA documents become an issue. This paper describes a clinical repository for CDA documents based on Oracle XML database. In the repository, clinical documents exist in the form of enhanced CDA with digital signature. This repository supports both traditional database features with statistical analysis ability and standard protocol for common file management. We implemented this clinical document repository in an EHR system.

Enhanced CDA with Electronic Signature

One of the key technologies in electronic medical record (EMR) is electronic signature which allows healthcare practitioners to sign off on electronic healthcare records. In order to implement electronic signature in clinical environment, two main factors should be considered: how to preserve the electronic signature data and the related contextual information in the EMR documents; how to introduce the tasks of EMR document electronic signature and validation into routing clinical workflow without disarranging well accepted working processes. This paper presents an enhanced CDA structure which extends HL7s CDA with the signature data presentation based on XML Digital Signature standard. In addition, the optimized workflow of clinical document with electronic signature has been designed. The enhanced CDA and related workflow has been used in the EMR system and implemented in the actual clinical environment. The result shows it can well solve the above problems and satisfy the applicable demands.

Multimodality brain atlas based on the visible human project dataset

A multimodality brain atlas is implemented based on the visible human dataset. Firstly, the VHDs images of different modalities are interpolated and registered into a common cubic space. A global rotation and translation is then used to transform the registered dataset into the standard Talairach coordinate. The Talairach atlas is mapped onto the VHDs images with a piecewise linear scaling. A texture segmentation method is developed to label the different tissue types according to the knowledge provided by the Talairach atlas. Finally, the 3D rendering results are provided.

An Application of Ant Colony Optimization Clustering Approach for Primary Headache Diagnosis

In this study, an Ant Colony Optimization (ACO) clustering approach is proposed for medical diagnosis to assign patients into different primary headache groups with Visual C++ program. By experiments, 375 patients were classified into Migraine, Tension-Type Headache (TTH) and Trigeminal Autonomic Cephalalgias (TACs). Results show that the clustering algorithm with ant colony can be important supportive for the medical experts in diagnostic.

The Process Data Driven Workflow Modeling Approach and its' Implementation on Medical Infection Control Process

Workflow technology is gaining attention in healthcare domain because it supports and assures fluent execution of complex healthcare business process. In order to implement workflow technology in healthcare environment, data-flow perspective issues should be considered. This paper analyzes basic characters of healthcare data from the point of workflow view, then presents an approach of process data driven workflow modeling method: the transformation of data fires events, which drive the transfer among tasks that have executable sequence in the workflow process. XML based workflow model description language is proposed as well, which can be mapped into executable software system framework conveniently and improve components capability of reusing and systemic agility. Furthermore, an actual healthcare process modeling example, medical infection control process model, is presented, which illustrate our approach is feasible.