Detection of the midsagittal plane in MR images using a sheetness measure from eigenanalysis of local 3D phase congruency responses

The midsagittal plane (MSP) separates the cerebrum into left and right hemispheres and its detection has a number of useful applications in brain image processing. We propose an automatic technique for the detection of the MSP in magnetic resonance (MR) images that uses a sheetness measure obtained from eigenanalysis of local matrix of second-order moments of 3D phase congruency responses to determine those voxels most likely to belong to the MSP. A weighted least-squares fitting algorithm is used in a coarse-to-fine iterative manner to find the best fitting plane (the MSP) through the selected voxels. Unlike most of the proposed approaches, which are mainly based on symmetric measures of the brain, our technique uses a direct measure to find the MSP. Our technique was applied to 202 MR images (40 clinical and 162 synthetic) and it has shown to be very effective for both symmetrical and asymmetrical brain images. Quantitative assessment, using the angle between unit normals of the detected and reference MSPs, yielded to a mean absolute angle value bellow 0.5°.