Comparative Study between Bias Correction Algorithms in Magnetic Resonance Imaging
-
1
University Saad Dahlab of Blida, Laboratory of Research in Medical Imagery, Algeria
Magnetic resonance images are often corrupted by an artifact called bias field or intensity inhomogeneity or shading artifact: a slow and smooth variation in the intensity of the image for the same tissue. The bias field is mainly due to non-homogeneity of the RF field, and increases significantly with increasing field strengths. It is generally modelled by a multiplicative field variable, which changes smoothly and slowly.
Bias field is unique to MRI. Given the brain’s ability to correct image imperfections, it is virtually invisible to the human eye. However, it can cause errors in segmentation, and consequent misinterpretation. To avoid this difficulty, MRI images are often preprocessed to reduce the effects of bias before segmentation.
In our work, we make a comparative study of seven bias correction algorithms: the Eq algorithm from Cohen & al, the algorithm proposed by Mangin and al. incorporated in the Brainvisa software, Ashburner and Friston’s algorithm, developed with SPM5, Styner and al.’s algorithm provided with the ITK library, EMS from Van Leemput and al. and finally Shattuck et al’s BFC algorithm, available in Brainsuite. All these algorithms can be classified into two broad categories: those which aim only to correct the bias and and those which alternate correction and segmentation.
Our study was conducted in 3 steps: - Estimation of corrected image bias and comparison with real bias present in images. - Generation of histograms for corrected images and comparison with the ideal histograms for the images. - Segmentation of corrected images and comparison with the ideal segmentation. We have tested these approaches on brain MRI corrupted with a known percentage of bias (40%).
Acknowledgment: A big thank you to the neurology department of Grenoble hospital for providing images, and to Catherine Garbay from UJF of Grenoble for her help.
Conference:
2nd NEUROMED Workshop, Fez, Morocco, 10 Jun - 12 Jun, 2010.
Presentation Type:
Poster Presentation
Topic:
Poster Session 4: Technological developments
Citation:
Cherfa
A and
Cherfa
Y
(2010). Comparative Study between Bias Correction Algorithms in Magnetic Resonance Imaging.
Front. Neurosci.
Conference Abstract:
2nd NEUROMED Workshop.
doi: 10.3389/conf.fnins.2010.12.00035
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
04 Jun 2010;
Published Online:
04 Jun 2010.
*
Correspondence:
Assia Cherfa, University Saad Dahlab of Blida, Laboratory of Research in Medical Imagery, Blida, Algeria, nemoABS01@frontiersin.org