User Guide and Documentation for the MIMIC II Database

Gari D. Clifford$^{1,2}$, Daniel J. Scott$^{1,2}$ and Mauricio Villarroel$^{1,2}$

$^1$Harvard-MIT Division of Health Sciences & Technology, Rm E25-505,
45 Carleton St., Cambridge MA 02142, USA
$^2$Massachusetts Institute of Technology Technology, Rm E25-505,
77 Massachusetts Av., Cambridge MA 02139, USA

MIMIC-II database version 2.6.

Rev: 290 LastChangedDate: 2011-09-07 15:23:22 -0400 (Wed, 07 Sep 2011)

Preface

This user guide is intended for clinicians with some knowledge of programming and/or graduate-level researchers with knowledge of biomedical signal processing. The user is expected to have a working knowledge of SQL. Basic knowledge of a statistical (signal processing) package such as Matlab or R is useful. Knowledge of C/C++ and/or Java may also help, but is not essential.

Many of the signal processing algorithms and data sets described in this guide are available from, or described in papers posted at the following URLs:

http://www.physionet.org

http://mimic.physionet.org

Most of the algorithms posted at the above URLs have been written either in C or Matlab. Libraries for reading these databases are also freely available. We hope that through these URLs this database will continue to evolve and add to the growing body of open (repeatable) biomedical research.

Gari Clifford, Li-wei Lehman, Roger Mark, Mohammed Saeed, George Moody, Daniel Scott, Ikaro Silva and Mauricio Villarroel

The Laboratory for Computational Physiology,
Cambridge, MA, USA,
January 2009