Exploration and the application of pattern and association discovery algorithms in the context of sequence, structure, and microarray data. Reverse engineering of metabolic and gene regulatory networks using dynamical and statistical predictive models that can be validated in an experimental setting. Dissection of complex genetic traits in a whole genome context, using association discovery techniques and their related statistical models.

The goal of the lab is to advance the state-of-art in biomedical imaging through the application of mathematics, physical science, and engineering. Research includes both basic science (e.g.,image acquisition) and clinical applications (e.g. measures of image quality). Methods include the study of image formation, qualitative analysis, evaluation and quantitative measures, at the molecular, cellular, tissue, and organ levels of analysis.

Machine learning and visualization are methods for discovering knowledge in large databases including text databases. This work focuses on testing and extending existing statistical and knowledge-based discovery methods to improve their performance on clinical data and text. Important issues include data representation, data accuracy, data completeness, and training set size.

This laboratory conducts research on the complex interaction of clinicians and information systems, seeking to improve the efficiency of clinical processes, reduce errors and improve quality of care. Key questions involve how clinicians collaborate, how information is carried across transitions of care, how natural language is used to communicate information, how clinical plans are developed, and how decisions are made. The lab explores the use advanced methods for information management, including natural language processing, summarization, collaboration tools, usability evaluation, data mining, decision support, document markup standards, clinical ontologies, and web services architecture.

Based on studies of clinicians' information needs (as they occur in daily workflow) and on models of expert information-retrieval strategies, the CIQR project will explore 1). the degree to which spoken natural-language queries via mobile communication devices can be used to mediate information-seeking behavior and systems, and 2). the use of speech-recognition, syntactic parsing, and semantic interpretation software in modeling or reproducing the methods used by professional librarians as they perform medical literature searches in response to clinicians' queries. The project seeks to improve access for physicians to the ever-enlarging body of current research data. CIQR wiki