Massachusetts General Hospital, Center for Morphometric Analysis
Internet Brain Segmentation Repository

Introduction

The IBSR is supported by the NIH under Grant number 1 R01 NS34189-01 from the National Institute of Neurological Disorders and Stroke (NINDS). This grant funds research in MR brain segmentation by researchers at Boston University, Draper Laboratory, Northeastern University, Massachusetts Institute of Technology, and Massachusetts General Hospital / Harvard Medical School.

Definition

The IBSR is a World Wide Web resource providing access to magnetic resonance brain image data and segmentation results contributed and utilized by researchers from all over the world.

Purpose

Its purpose is to encourage the development and evaluation of segmentation methods by providing raw test and image data, human expert segmentation results, and methods for comparing segmentation results.

Justification

The field of medical image processing is growing rapidly. Many novel approaches to the segmentation, classification and shape analysis are being developed and reported. This poses a problem in the identification and selection of an optimal methodology once a specific analysis task has been identified. It has become virtually impossible for even a well equipped image processing laboratory to implement all of the methods that are published. An alternative to local implementation of each method is the establishment of a local repository for image data and the results of any analysis. In this way, new analysis algorithms can be tested on common images, and the results directly compared between various methods and the human expert interpretation of the images.

General data description

This repository is meant to contain standard test image data sets which will permit a standardized mechanism for evaluation of the sensitivity of a given analysis method to signal to noise ratio, contrast to noise ratio, shape complexity, degree of partial volume effect, etc. This capability is felt to be essential to further development in the field since many published algorithms tend to only operate successfully under a narrow range of conditions which may not extend to those experienced under the typical clinical imaging setting. This repository is also meant to describe and discuss methods for the comparison of results.

Credits

Thanks to all those who have used and contributed to the IBSR!

The IBSR was initially created in April 1996 and is maintained by Andrew Worth at the CMA .

Michael Farmer converted scan 788_6 into 8-bit TGA format and modified the program overlap.c to read 8-bit TGA image files in a Windows 32-bit environment.

YOU are encouraged to contribute and participate in this resource. If you would like, please see the list of things to do. Full credit for anything you do will appear here.


Go to [IBSR Main Page]. Questions about this page should be directed to Andy.