Skip to Main Content
Mutual information (MI) is a popular entropy-based similarity measure used in the medical imaging field for multimodal registration. The basic concept behind any approach using MI is to find a transformation, which when applied to an image, will maximize the MI between two images. A common implementation of MI involves the use of Parzen windows. This process generally requires two samples of image intensities: one to estimate the underlying intensity distributions and the second to estimate the entropy. This paper presents a novel gradient-based registration algorithm (MIGH) which uses Gauss-Hermite quadrature to estimate the image entropies. The use of this technique provides an effective and efficient way of estimating entropy while bypassing the need to draw a second sample of image intensities. With this technique, it is possible to achieve similar results and registration accuracy when compared to current Parzen-based MI techniques. These results are achieved using half the previously required sample sizes and also with an improvement in algorithm complexity.