Symmetry plane detection in neuroimages based on intensity profile analysis

Symmetry plane identification is a critical initial step in brain image analysis, either done manually or automatically. Automatic extraction of the symmetry plane can provide an initial estimate for brain image registration, pathology assessment and disease diagnosis. This paper presents a novel technique for extracting the mid plane from volumetric magnetic resonance images (MRI). It is a theoretically simple approach that utilizes anatomical and radiological properties. The method is straight-forward to implement without the need for any prior segmentation. The efficacy of the proposed method was tested on brain MRI while investigating the robustness against rotation, intensity non-uniformity, noise and pathology. The method was compared with a cross-correlation based technique and the results show the viability of the approach.