We have been involved in developing new strategies to calculate laterality indices from functional MRI activations, particularly those caused by memory encoding tasks. We found that incorporating the totality of voxels present in a certain region, rather than only voxels activated above an arbitrary threshold, allows for the determination of more meaningful lateralizations. We were able, for example, to better differentiate lateralizations in patients from those obtained from healthy subjects by using techniques that rely on whole voxel distributions. This can be useful for surgical planning of epilepsy and tumor patients. We are now further developing this technique by studying the influence of successful memorization and of negatively correlated voxels in the calculation of laterality indices, which can help us increase our accuracy in detecting abnormal normalizations in patients.