[visionlist] CfP: MICCAI 2012 Multimodal Brain Tumor Segmentation Challenge (MICCAI-BRATS)

[Bjoern Menze]

CALL FOR PARTICIPATION:

MICCAI 2012 Multimodal Brain Tumor Segmentation Challenge (MICCAI-BRATS)

http://www2.imm.dtu.dk/projects/BRATS2012

Workshop during MICCAI 2012 in Nice, France

October 1st 2012

---

DATA & CHALLENGE DESCRIPTION:

We will provide preprocessed (co-registered, skull-stripped, resampled) and
annotated training data from glioma patients with low- and high-grade
disease. Voxel labels are 'edema' and 'tumor core'.

The data sets comprise synthetically generated tumor images as well as real
data acquired from different patients prior to and post therapy. MR image
contrasts are T1, contrast-enhanced T1-gad, T2-FLAIR, T2.

Training data can be downloaded NOW (30 patient data sets, 50 synthetic data
sets), test data will be evaluated as on-site challenge during MICCAI 2012.
The tumor segmentation methods of all participants will be ranked using Dice
score, Hausdorff distance, etc.

All automated, semi-automated or inter-active tumor segmentation approaches
are permitted. Performances will be evaluated separately.

---

PARTICIPATION IN A NUT-SHELL:

* NOW: download the training data and adjust your tumor segmentation method

* before Jul 3th: submit a 2-4 page paper that describes your method; there
is a quick review and decisions are out by Jul 6th

* before Aug 15th: register to the workshop via the MICCAI conference

* [Jul-Sep: continue adjusting your method if you want to]

* Oct 1st: attend the on-site challenge in Nice, get the test data, present
your method, report results

* [Oct 1st + a few hours: report late results if your method has not
finished by the end of the day]

* Oct-Nov: contribute to the joint journal paper

---

For more detailed information see the homepage, or contact the organizers
directly.


Bjoern Menze, ETH Zurich

Andras Jakab, ETH Zurich & Debrecen Medical School

Stefan Bauer, U Bern

Mauricio Reyes, U Bern

Marcel Prastawa, U Utah

Koen Van Leemput, TU Denmark & Harvard Medical School