Evaluation of similarity measures in contrast enhanced echocardiography motion detection and registration

In the present work, a similarity metric based on image difference entropy has been investigated for myocardial contrast echocardiography non-rigid registration. The image histogram size has shown to influence registration accuracy, improving results for difference entropy, which has presented better results when compared to other two well known similarity metric in literature, namely, normalized cross correlation and mutual information. The displacement field has been smoothed with a Gaussian convolution, allowing controlling the smoothness in geometric transformation and attenuating displacement field miscalculations. Therefore, registration parameters can be set to privilege more local or global image transformation. Difference image entropy has shown robust accurate registration for myocardial contrast echocardiography at parameter values found in this study.