A new model to determine asymmetry coefficients on MR images using PSNR and SSIM

The human brain consists of two hemispheres, right and left. These two hemispheres are almost symmetrical, not perfectly. However, in neurological diseases, the volumetric losses in the brain begin to deteriorate asymmetrically between the two hemispheres. This deterioration can be local or global in the brain. Symmetry deterioration can be a biomarker in the early stage diagnosis and the following of neurological diseases. However, it has been stated that the analysis of asymmetry in the brain by numerical methods is problematic. In this study, a new approach is proposed to analyze the brain symmetry deterioration numerically. In order to perform asymmetry analysis in MR images, two hemispheres must be separated from each other by finding the midsagittal plane which are known symmetry axis. The PSNR and SSIM coefficients are often used for quality measurements between two images. In the study, these coefficients were tested for asymmetry measurement. Statistical analysis was performed by determining PSNR-SSIM coefficients between 70 Control and 70 Alzheimer Disease MR images from the OASIS database. It was determined that the use of PSNR and SSIM coefficients in the asymmetry analysis of MR images gave meaningful results.

[1]  William D. Hopkins,et al.  Gray matter asymmetries in chimpanzees as revealed by voxel-based morphometry , 2008, NeuroImage.

[2]  Yanxi Liu,et al.  Automatic bilateral symmetry (midsagittal) plane extraction from pathological 3D neuroradiological images , 1998, Medical Imaging.

[3]  R. Koeppe,et al.  A diagnostic approach in Alzheimer's disease using three-dimensional stereotactic surface projections of fluorine-18-FDG PET. , 1995, Journal of nuclear medicine : official publication, Society of Nuclear Medicine.

[4]  Koenraad Van Leemput,et al.  Quantification of Cerebral Grey and White Matter Asymmetry from MRI , 1999, MICCAI.

[5]  Sheena Xin Liu,et al.  Symmetry and asymmetry analysis and its implications to computer-aided diagnosis: A review of the literature , 2009, J. Biomed. Informatics.

[6]  Lutz Jäncke,et al.  A voxel-based approach to gray matter asymmetries , 2004, NeuroImage.

[7]  Mark Mühlau,et al.  Grey-matter atrophy in Alzheimer's disease is asymmetric but not lateralized. , 2011, Journal of Alzheimer's disease : JAD.

[8]  Sébastien Ourselin,et al.  Computation of the mid-sagittal plane in 3-D brain images , 2002, IEEE Transactions on Medical Imaging.

[9]  et al.,et al.  Spatial patterns of brain atrophy in MCI patients, identified via high-dimensional pattern classification, predict subsequent cognitive decline , 2008, NeuroImage.

[10]  Daniel Rueckert,et al.  Automatic morphometry in Alzheimer's disease and mild cognitive impairment☆☆☆ , 2011, NeuroImage.

[11]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[12]  Arno Klein,et al.  Evaluation of 14 nonlinear deformation algorithms applied to human brain MRI registration , 2009, NeuroImage.

[13]  Max A. Viergever,et al.  Assessment of methods to extract the mid-sagittal plane from brain MR images , 2013, Medical Imaging.

[14]  Djemel Ziou,et al.  Image Quality Metrics: PSNR vs. SSIM , 2010, 2010 20th International Conference on Pattern Recognition.

[15]  John G. Csernansky,et al.  Open Access Series of Imaging Studies (OASIS): Cross-sectional MRI Data in Young, Middle Aged, Nondemented, and Demented Older Adults , 2007, Journal of Cognitive Neuroscience.