Analysis of Generalized Entropies in Mutual Information Medical Image Registration

TL;DR

This paper compares different mutual information functions, including Shannon and Tsallis entropies, for 3D medical image registration, demonstrating improved robustness and speed, especially with GPU acceleration.

Contribution

It introduces novel GPU-accelerated MI algorithms based on generalized entropies and analyzes their effectiveness in improving registration accuracy.

Findings

GPU-based MI algorithms outperform traditional methods in speed.

Generalized entropies enhance registration robustness.

The proposed methods reduce registration failures in challenging scenarios.

Abstract

Mutual information (MI) is the standard method used in image registration and the most studied one but can diverge and produce wrong results when used in an automated manner. In this study we compared the results of the ITK Mattes MI function, used in 3D Slicer and ITK derived software solutions, and our own MICUDA Shannon and Tsallis MI functions under the translation, rotation and scale transforms in a 3D mathematical space. This comparison allows to understand why registration fails in some circumstances and how to produce a more robust automated algorithm to register medical images. Since our algorithms were designed to use GPU computations we also have a huge gain in speed while improving the quality of registration.