Application of Weber's Law to Medical Image Registration to Accommodate Intensity Inhomogeneities

Medical image registration is often confronted by the presence of image intensity inhomogeneity, even when the images are acquired of the same subject using mono-modality techniques. An intensity- invariant dense feature extraction method as inspired by the Weber's Law is proposed as a tool to align and warp medical images containing intensity inhomogeneity issues through a feature-based registration method. Registration performance was evaluated using magnetic resonance and histology images and the results were compared to two established intensity correction methods. Our findings show that the proposed method performs comparably in the case of affine registration, and outperforms the other two methods when non-linear warping is used as part of the registration process.