From Physiomimi

Jump to: navigation, search

Physio-MIMI - Multi-Modality Multi-Resource Physiological and Clinical Informatics Infrastructure

Physio-MIMI (Multi-Modality, Multi-Resource Information Integration environment) is an NCRR-funded, multi-CTSA site project designed to develop novel, flexible informatics methodologies, tools and infrastructure to facilitate the collection, management, and analysis of clinical and physiological data.


Physio-MIMI supports collaborative clinical and translational research using diverse data types across institutions. It builds upon, integrates, and extends two existing informatics initiatives: MIMI (CWRU) and Honest Broker (University of Michigan). MIMI is a web-based data and research management system that consists of three conceptually independent but interconnected servers: Meta-Server, Application Server and Data Resource Server.

Led by Case Western Reserve University, in collaboration with the University of Michigan, Marshfield Clinic and the University of Wisconsin, Physio-MIMI sports an expandable Sleep Domain Ontology; fine-grained interface for role-based data-source level access control; plug-and-play Honest-Broker adaptor to mediate data access services; and data schema to Sleep Domain Ontology mapper that in an instance, transforms your local database into a data resource that can be queried using the powerful federated query interface VISAGE on the web.


Physio-MIMI Components and their Features

  • VISAGE query interface incorporating Query Builder, Query Manager and Query Explorer
  • PhysioMap interface supporting the mapping of distributed data sources to a common set of queryable terms
  • Federated query across data sources, including sleep records and other complex data
  • Case-control cohort identification tool with flexible matching and box-plot visualization on web browser
  • EDF Viewer to visualize sleep signals and annotations
  • EDF Translator to convert score labels to standardized format
  • EDF Editor to edit and normalize signal attributes in header to read studies from multiple platforms


Architecture

Physio-MIMI consists of three conceptually independent but interconnected servers:

  1. Meta-Server providing administrative functionality of three types, for:
    1. managing the research project process (organizing clinical data, summarizing scientific workflow, etc.)
    2. role-based access control of data and functionalities of a variety of end-users
    3. cross-linking the various data servers and databases and translation of queries from a variety of end-users to a unified front-end, and then to relevant data resources (where the retrieved results from participating sites’ databases will be de-identified or watermarked)
  2. Application Server, which provides a set of tools (the ToolBox) for processing, analyzing, annotating and manipulating physiological data in support of
    1. data curation for sharing and cross-linking
    2. hypothesis-driven and discovery-driven research (i.e. data mining) studies
  3. Data Resource Server, encompassing the distributed databases residing on each participating site, embedded in a common query interface. The physiological databases will be populated using existing data and a specified framework which will allow standardization of queries and harmonization of results. Databases include:
    1. CWRU: the Sleep Heart Health Study, MrOS, Cleveland Family Study, Cleveland Children’s Health and Sleep Cohort; University Hospitals of Cleveland Sleep Center; Oncore Trials Clinical Trials Management System (ONCORE-ORM) and ONCORE-Biospecimen Management System (ONCORE-BSM)
    2. UM: University of Michigan Sleep Databases; Care Web EHR; Velos Clinical Trial Database
    3. UW/MC: Marshfield Clinic Data Warehouse (MCDW); Marshfield Personalized Medicine Project; MC MESA; Marshfield and Wisconsin Sleep Databases

Retrieved from "http://physiomimi.case.edu/physiomimi/index.php/Physiomimi:About"