The MIMIC-III Waveform Database Matched Subset

The new PhysioNet website is available at: We welcome your feedback.

This is version 1.0 of the MIMIC-III Waveform Database Matched Subset (August 2017; 22,317 waveform records and 22,247 numerics records matched with 10,282 MIMIC-III Clinical Database records.)

The MIMIC-III database is described in

Johnson AEW, Pollard TJ, Shen L, Lehman L, Feng M, Ghassemi M, Moody B, Szolovits P, Celi LA, and Mark RG. MIMIC-III, a freely accessible critical care database. Scientific Data (2016). DOI:10.1038/sdata.2016.35.

Please cite this publication when referencing this material, and also include the standard citation for PhysioNet:

Goldberger AL, Amaral LAN, Glass L, Hausdorff JM, Ivanov PCh, Mark RG, Mietus JE, Moody GB, Peng C-K, Stanley HE. PhysioBank, PhysioToolkit, and PhysioNet: Components of a New Research Resource for Complex Physiologic Signals. Circulation 101(23):e215-e220 [Circulation Electronic Pages;]; 2000 (June 13).

The MIMIC-III Waveform Database Matched Subset (mimic3wdb/matched) contains all MIMIC-III Waveform Database records that have been associated with MIMIC-III Clinical Database records. The record matching process is ongoing, and more records may be added to the matched subset in the future.

All data associated with a particular patient have been placed into a single subdirectory, named according to the patient's MIMIC-III Subject_ID. These subdirectories are further divided into ten intermediate-level directories (p00 to p09).

The name of each mimic3wdb/matched waveform record is of the form


where XXNNNN is the matching MIMIC-III Clinical Database Subject_ID, and YYYY, MM, DD, hh, and mm are the surrogate year, month (01-12), and day (01-31), and the real hour (00-23) and minute (00-59), derived from the starting date and time of day of the mimic3wdb/matched record. The surrogate dates match those of the corresponding MIMIC-III Clinical Database records.

In most cases, the waveform record is paired with a numerics record (see the description of numerics records on the MIMIC-III Waveform Database home page). The numerics record names are of the form


i.e., identical to the name of the associated waveform record except for the appended n.

Frequently there are multiple waveform and numerics record pairs associated with a given clinical record; all of them will appear in the same subdirectory in such a case, and their names will indicate their chronologic sequence. For example, MIMIC-III Clinical Database record p000079 has been matched with two waveform and numerics record pairs, named

p000079-2175-09-26-01-25    p000079-2175-09-26-01-25n
p000079-2175-09-26-12-28    p000079-2175-09-26-12-28n

Each mimic3wdb/matched record is also an undated mimic3wdb record (i.e., it also belongs to the full MIMIC-III Waveform Database). Only the surrogate-dated mimic3wdb/matched header (.hea) files are unique to the MIMIC-III Waveform Database Matched Subset; the others, with names of the form 3*.hea and 3*.dat, are copies of the like-named files in the full database. (Actually, they are symbolic links to the original files, a distinction that matters only when copying them using rsync, which can be used to download them quickly if the undated records have been downloaded previously.)

See the MIMIC-III Waveform Database home page for additional information about these records.

See the releaseinfo page for information about the releases of the MIMIC-III waveform database and its matched subset.

Downloading the Matched Subset

Most users will not need to download the entire MIMIC-III Waveform Database Matched Subset. For those who think they do, please see Can I download an entire PhysioBank database in one step? in the PhysioNet FAQ.

If you have already downloaded parts 0 through 9 of the MIMIC-III Waveform Database, version 3, and your OS supports symbolic links, use this command (as in the FAQ) to download the Matched Subset:

rsync -Cavz /usr/database/mimic3wdb/matched

The command above will require only a few minutes in most cases, since nearly all of the files in the Matched Subset are shared with files in parts 0 through 9 that have been downloaded already.

If, however, you wish to download the Matched Subset without having previously downloaded the entire MIMIC-III Waveform Database, or if your OS does not support symbolic links, use these commands instead:

mkdir -p /usr/database/mimic3wdb/matched
rsync -CaLvz /usr/database/mimic3wdb/matched

This command will download about 2.4 TB, and will require at least a few hours; if you do not have a high-speed network connection it may require a day or more.

You may also download subsections of the matched subset if 2.4 TB is too much to download at once. For example:

mkdir -p /usr/database/mimic3wdb/matched/p00
rsync -CaLvz /usr/database/mimic3wdb/matched/p00
You can also download this data from a AWS S3 bucket, with the file named
Icon  Name                            Last modified      Size  Description
[PARENTDIR] Parent Directory - [   ] DOI 2017-08-07 18:16 20 [   ] RECORDS 2017-08-04 18:01 131K list of record names [   ] RECORDS-numerics 2017-08-04 18:08 826K list of record names [   ] RECORDS-waveforms 2017-08-04 18:08 806K list of record names [DIR] p00/ 2017-07-01 08:00 - [DIR] p01/ 2017-07-01 08:00 - [DIR] p02/ 2017-07-01 08:00 - [DIR] p03/ 2017-07-01 08:00 - [DIR] p04/ 2017-07-01 08:00 - [DIR] p05/ 2017-07-01 08:00 - [DIR] p06/ 2017-07-01 08:00 - [DIR] p07/ 2017-07-01 08:00 - [DIR] p08/ 2017-07-01 08:00 - [DIR] p09/ 2017-07-01 08:00 -

Questions and Comments

If you would like help understanding, using, or downloading content, please see our Frequently Asked Questions.

If you have any comments, feedback, or particular questions regarding this page, please send them to the webmaster.

Comments and issues can also be raised on PhysioNet's GitHub page.

Updated Friday, 28 October 2016 at 16:58 EDT

PhysioNet is supported by the National Institute of General Medical Sciences (NIGMS) and the National Institute of Biomedical Imaging and Bioengineering (NIBIB) under NIH grant number 2R01GM104987-09.