MIMIC-III is a corpus has 58,976 hospital admission and 2,083,180 had written notes by medical professionals. This repo contains an automated build process that creates an SQLite database (file) populated with the MIMIC-III corpus using scripts from the [MIMIC Code Repository].
- Download the source MIMIC-III data files as the file
mimic-iii-clinical-database-1.4.zip
to this directory. - Install git, and GNU make.
- Run the automation process:
make all
. This does the following:- Uncompress the
mimic-iii-clinical-database-1.4.zip
of the compressed CSV MIMIC-III data files. - Clones the MIMIC-III code repository, which has the SQLite DB load scripts.
- Loads the database using the MIMIC-III code repository scripts.
- Creates some indexes on some tables such as
NOTEEVENTS
,ADMISSIONS
andPATEINTS
.
- Uncompress the
- Create additional indexes to suit your needs. Use the post configuration script as an example.
- Check for errors and the existence of the
mimic3.sqlite3
SQLite database file. - Optionally clear up disk usage:
make cleanall
. - Optionally reduce the SQLite database file by editing and rerunning the post configuration script.