A Lattice Boltzmann Method Based Medical Image Denoising and Enhancement

This paper presents a new LBM-based model for image denoising and enhancement. The mathematical deduction showed that the global equation of our model was very similar to the diffusion reaction equation. By testing our LBM-based model on the artificial images, the behavior of image denoising and enhancement was given. We also demonstrated the effectiveness of the algorithm in clinical images such as blood cell image and MR image. Experiment results show that our model is effective.

I. Introduction

Lattice Boltzmann Method (LBM) is a numerical scheme based on kinetic theory for modeling various mathematical physical problems [1]. It is a new numerical method originated from the Lattice Gas Automata (LGA) which is a kind of Cellular Automata. Since 1990s, many interests have been aroused in the application of LBM to the field of computational physics. Nowadays, LBM has become an alternative tool to the traditional numerical technology for resolving the partial differential equation (PDE) which also has been used in the image smoothing [2] and image segmentation [3] since 1980's. LBM is easy to implement and has the ability to incorporate massive parallel computation [4] [5]. The most important difference between the LBM and the traditional numerical methods is that LBM is based on the microscopic description of the physical systems while the traditional numerical method is based on the discretization of the PDE which is a macroscopic description of the physical systems. By simulating the microscopic behavior of the physical systems, the global equation of LBM can match a certain PDE.