ImageCLEFmed 2007 Data

Last updated - 25 May 2007, William Hersh

The data for the ImageCLEFmed 2007 task is available in the protected area of the Web site as described in the email sent to participants. Note that usage of mypacs.net data requires completing a separate license agreement and sending it to Henning Müller.

Some files were updated and/or had their names changed on or before 25 May 2007:

The file of metadata annotations, ImageCLEFmed2007.xml was corrected and the original file is now named Imageclefmed2007old.xml.
The readme file from 2006 has been renamed to README2006.
The baseline textual and visual runs have been designated as baseline06.
Seven empty (0 byte file size) images were discovered in the MyPACS collection and removed. There are other images for the two cases in which they occur, so no changes to the cases have been made.
A baseline run has been made available using the FIRE retrieval system of University of Aachen, baseline07.FIRE.gz. The run uses visual retrieval techniques, using color histograms, tamura texture features, and image thumbnails weighted in a 3:2:1 manner.

The directory for the data contains the following files in the /iclefmed directory:

CASImage.tar		25-Jan-2006 19:08 	1.3G
DataCheck.java		06-Feb-2006 12:34 	1.1K
Imageclefmed.xml	01-Apr-2006 14:26 	19M
Imageclefmed2007.xml	15-May-2007 12:08 	23M
Imageclefmed2007old.xml	26-Mar-2007 18:13 	23M
MIR.tar			25-Jan-2006 19:08 	63M
PathoPic.tar		25-Jan-2006 19:09 	879M
Peir.tar		01-Apr-2006 12:56 	2.5G
Peir.tar_aa		01-Apr-2006 13:03 	954M
Peir.tar_ab		01-Apr-2006 13:04 	954M
Peir.tar_ac		01-Apr-2006 13:05 	651M
README2006		01-Apr-2006 14:09 	2.4K
README2007		15-May-2007 12:35 	3.3K
baseline06.text.txt	07-Jun-2006 07:13 	23M
baseline06.visual.txt	07-Jun-2006 07:13 	1.3M
baseline07.FIRE.gz	25-May-2007 05:41 	276K
endoscopic.tar		15-May-2007 13:24 	33M
iclefmed2005images.zip	27-Dec-2006 09:21 	2.8M
iclefmed2005topics.doc	27-Dec-2006 09:21 	176K
iclefmed2005topics.txt	27-Dec-2006 09:21 	4.9K
iclefmed2006images.zip	27-Dec-2006 09:21 	6.7M
iclefmed2006topics.doc	27-Dec-2006 09:21 	7.0M
iclefmed2006topics.txt	27-Dec-2006 09:21 	4.8K
iclefmed2007images.zip	15-May-2007 16:26 	2.8M
iclefmed2007topics.txt	16-May-2007 17:19 	3.1K
iclefmed2007topics.xml	16-May-2007 18:13 	6.2K
icm05.qrels.lenient.txt	27-Dec-2006 09:24 	620K
icm05.qrels.txt		27-Dec-2006 09:22 	620K
icm06.qrels.txt		28-Jul-2006 05:57 	783K
topic#conversion.txt 	27-Dec-2006 09:59 	234
updatePeir.java		01-Apr-2006 13:07 	858

And this file in the iclefmed2 directory:

mypacs.tar		31-Mar-2007 07:00	390M

The README files describes how to extract the collection from the .tar files. Here is a description of the files:

Images

CASImage
MIR
PathoPic
Peir - also split into three files for systems that cannot handle >2 GB files
myPACS
Endoscopic

Image Annotations (metadata)

Imageclefmed2007.xml (ImageCLEFmed.xml is metadata file from 2005-2006 tracks)

Topics

From 2005: iclefmed2005*
From 2006: iclefmed2006*

Note: the 2006 files have some corrections in the German topics from the originally released topic files

Relevance judgments

icm05.qrels.txt - official qrels; only definitely relevant designated as relevant
icm05.qrels.lenient.txt - definitely and possibly relevant designated as relevant
icm06.qrels.txt - official qrels; definitely and possibly relevant designated as relevant

Utilities

baseline.* files - baseline runs for textual and visual retrieval systems using 2006 data
DataCheck.java
updatePeir.java
topic#conversion.txt

Here is more detail of the corrections to the 2007 images: MyPACS case 258423 had 5 0kb images: mp_258441, mp_258443, mp_258445, mp_258447, mp_258449. MyPACS case 317190 had 2 0kb images: mp_317190 and mp_317192. As these files were also malformed in our source, they were removed from myPACS/images/. References to these images were removed from the cases MP_258423.xml and MP_111310.xml, as well as Imageclef2007.xml. Initially, collections/endospic/IMAGES/155000029461.jpg was also 0kb. However, it was replaced with the correct image. The updates have been noted in README2007.

Organisers

Henning Müller, University and Hospitals of Geneva, Geneva, Switzerland	Email: henning.mueller@sim.hcuge.ch Web: http://www.sim.hcuge.ch/medgift/02_Henning_EN.htm
Paul Clough, Sheffield University, Sheffield, England	Email: p.d.clough@sheffield.ac.uk Web: http://ir.shef.ac.uk/cloughie/
William Hersh, Oregon Health &Science University (OHSU), Portland, OR, USA	Email: hersh@ohsu.edu Web: http://www.billhersh.info/
Jayashree Kalpathy-Cramer, Oregon Health &Science University (OHSU), Portland, OR, USA	Email: kalpathy@ohsu.edu