A Contrast Enhancement Algorithm for Low-Dose CT Images Based on Local Histogram Equalization

As an important and necessary step in many medical image processing applications, contrast enhancement can amplify tiny anatomies, such as airways, vessels, lung nodules and pulmonary fissures in lung CT (computerized tomography) images. In this paper we describe a more useful contrast enhancement algorithm based on localized histogram equalization for low-dose CT images. This algorithm applies a two-stage approach: (1) computing the local statistics using a novel scheme which uses two parameters (the sub-block size and the motion step) as guides in reducing the calculating time and avoiding block artifacts at the block boundaries; and then (2) performing histogram equalization based on a modified local contrast-stretching manipulation. We tested the proposed algorithm on 2 CT sets which include 258 HRCT ( high resolution computerized tomography) images and 476 low-dose CT images respectively. Comparing with the visual inspection and the processing rate, the algorithm is proved as a flexible and effective way for low-dose lung CT image enhancement and can be use as a pre-process method for low dose CT images understanding and analysis.