Vasudevan Jagannathan

Assessment of commercial NLP engines for medication information extraction from dictated clinical notes.

PURPOSE We assessed the current state of commercial natural language processing (NLP) engines for their ability to extract medication information from textual clinical documents. METHODS Two thousand de-identified discharge summaries and family practice notes were submitted to four commercial NLP engines with the request to extract all medication information. The four sets of returned results were combined to create a comparison standard which was validated against a manual, physician-derived gold standard created from a subset of 100 reports. Once validated, the individual vendor results for medication names, strengths, route, and frequency were compared against this automated standard with precision, recall, and F measures calculated. RESULTS Compared with the manual, physician-derived gold standard, the automated standard was successful at accurately capturing medication names (F measure=93.2%), but performed less well with strength (85.3%) and route (80.3%), and relatively poorly with dosing frequency (48.3%). Moderate variability was seen in the strengths of the four vendors. The vendors performed better with the structured discharge summaries than with the clinic notes in an analysis comparing the two document types. CONCLUSION Although automated extraction may serve as the foundation for a manual review process, it is not ready to automate medication lists without human intervention.

Natural Language Processing Framework to Assess Clinical Conditions

OBJECTIVE The authors developed a natural language processing (NLP) framework that could be used to extract clinical findings and diagnoses from dictated physician documentation. DESIGN De-identified documentation was made available by i2b2 Bio-informatics research group as a part of their NLP challenge focusing on obesity and its co-morbidities. The authors describe their approach, which used a combination of concept detection, context validation, and the application of a variety of rules to conclude patient diagnoses. RESULTS The framework was successful at correctly identifying diagnoses as judged by NLP challenge organizers when compared with a gold standard of physician annotations. The authors overall kappa values for agreement with the gold standard were 0.92 for explicit textual results and 0.91 for intuited results. The NLP framework compared favorably with those of the other entrants, placing third in textual results and fourth in intuited results in the i2b2 competition. CONCLUSIONS The framework and approach used to detect clinical conditions was reasonably successful at extracting 16 diagnoses related to obesity. The system and methodology merits further development, targeting clinically useful applications.

Collaborative infrastructures using the WWW and CORBA-based environments

The current proliferation of the World Wide Web (WWW) and the growing popularity of the Java language for programming such environments has fuelled a need for Web browsers to provide access to the variety of information sources and services now available on the Internet. In parallel to this revolution, computer vendors have been cooperating to develop an interoperable standard for communication which is based on the Common Object Request Broker Architecture (CORBA). We present the tools we have developed that provide powerful mechanisms for accessing any service based upon the interoperable CORBA standards from the popular WWW browsers. These tools are available in the public domain and have been downloaded and used around the world for building practical, fieldable systems.

Machine learning-based coreference resolution of concepts in clinical documents

OBJECTIVE Coreference resolution of concepts, although a very active area in the natural language processing community, has not yet been widely applied to clinical documents. Accordingly, the 2011 i2b2 competition focusing on this area is a timely and useful challenge. The objective of this research was to collate coreferent chains of concepts from a corpus of clinical documents. These concepts are in the categories of person, problems, treatments, and tests. DESIGN A machine learning approach based on graphical models was employed to cluster coreferent concepts. Features selected were divided into domain independent and domain specific sets. Training was done with the i2b2 provided training set of 489 documents with 6949 chains. Testing was done on 322 documents. RESULTS The learning engine, using the un-weighted average of three different measurement schemes, resulted in an F measure of 0.8423 where no domain specific features were included and 0.8483 where the feature set included both domain independent and domain specific features. CONCLUSION Our machine learning approach is a promising solution for recognizing coreferent concepts, which in turn is useful for practical applications such as the assembly of problem and medication lists from clinical documents.

Model based information access

The engineering data of a large enterprise is typically distributed over a wide area and archived in a variety of databases and file systems. Access to such information is crucial to a team member, particularly in a concurrent engineering setting. However, this is not easy, because (1) a model of the relevant information is not available, and (2) there is no simple way to access the information without being knowledgeable about various computer data formats, file systems, and networks. The authors have developed a system called the Information Sharing System (ISS) to enable access to diverse and distributed information within a corporation. Such data could be stored in different repositories such as databases and file systems including those that contain multiple media. The paper describes the methodology fo the ISS, the details of the implementation nd extensions planned for the future.<<ETX>>

Secure collaboration technology for healthcare enterprises

Healthcare organizations have a legacy of relatively isolated vendor-proprietary departmental systems. The cost of integrating disparate healthcare systems is a significant barrier to collaborative endeavors such as telemedicine. Recent legislative measures for the protection of healthcare information place significant responsibilities on healthcare organizations to ensure that their healthcare systems and information technology practices adequately protect the privacy of healthcare information in their charge. The authors, who are developing secure telemedicine applications, identify some of these hurdles and discuss their approach to enable healthcare organizations to engage in collaborative healthcare activities.

Functional specifications for collaboration services

Describes an attempt at specifying a generic and reusable set of services for computer-supported collaboration among teams engaged in any collaborative process within a distributed organization or group of organizations. The services are cataloged under the headings: information management, group communications, group management, process management, and infrastructure support.<<ETX>>

ARTEMIS: a research testbed for collaborative health care informatics

Patient centred healthcare delivery is an inherently collaborative and information-intensive process. It involves a wide range of individuals and organizations with different roles. The key to cost reduction and quality improvement in health care is effective management of this collaborative process. The Concurrent Engineering Research Center (CERC) developed a number of key technologies to enable collaborative processes. These technologies, integrated into an open collaborative environment, is currently being customized to create a research testbed, ARTEMIS, that addresses all aspects of patient-care life-cycle. ARTEMIS will equip each provider with his own personal assistant-a customized role-oriented workstation connected to his own information world through an information bus. With ARTEMIS, the users will be able to process multimedia patient-care information: look-up, compute, communicate, archive and collaborate.<<ETX>>

Architectural alternatives for community care networks

Patient centered healthcare delivery is an inherently collaborative and information-intensive process. It involves a wide range of individuals and organizations with different roles. The key to cost reduction and quality improvement in healthcare is effective management of this collaborative process. We investigate various alternatives and their ramifications in implementing a community care network. We leverage the existing technology and software developed earlier at our center as well as from the ongoing project on an Advanced Research Testbed for Medical Information System (ARTEMIS).<<ETX>>