Complex wavelets for registration of tagged MRI sequences

Tagged magnetic resonance imaging (MRI) is currently the reference MR modality for myocardial motion and strain analysis. Mutual information (MI) based non rigid registration has proven to be an accurate method to retrieve cardiac deformation fields. However, this technique ignores high frequency information in tags. In a previous work this information was included by using feature vectors formed with wavelet coefficients and kNN graphs to estimate alphaMI. It was shown that cardiac motion estimation was feasible with these features. In this work, features were derived from complex wavelet transform (CWT), which is shift invariant and provides more high frequency subimages than conventional wavelets. Results show that lower errors are obtained with respect to the use of pixel intensity