A Variational Level Set Approach to Segmentation and Bias Correction of Images with Intensity Inhomogeneity

Abstract

This paper presents a variational level set approach to joint segmentation and bias correction of images with intensity inhomogeneity. Our method is based on an observation that intensities in a relatively small local region are separable, despite of the inseparability of the intensities in the whole image caused by the intensity inhomogeneity. We first define a weighted K-means clustering objective function for image intensities in a neighborhood around each point, with the cluster centers having a multiplicative factor that estimates the bias within the neighborhood. The objective function is then integrated over the entire domain and incorporated into a variational level set formulation. The energy minimization is performed via a level set evolution process. Our method is able to estimate bias of quite general profiles. Moreover, it is robust to initialization, and therefore allows automatic applications. The proposed method has been used for images of various modalities with promising results.

Compensation of the non-uniformity of back-lighting sources when digitising X-ray films for video-densitometric measures

“…The set-up algorithm has been studied in order to eliminate one specific source of noise; further elaborations are needed in order to take into account the radiometric response and other vignetting effects; in the field of medical imaging; very promising algorithms are those introduced by Leemput et al [18] in relation to brain RM, based on 'a priori' knowledge, and the algorithm by Li et al [17] where a 'variational level set approach' to segmentation has been introduced.…”

“…However it is opinion of the authors that a significant portion of digitisation noise is not randomly distributed, being related to the non-uniformity of the back-lighting source; a proof is given by the peculiar pattern of subtraction images obtainable when performing repeatability tests [11] ; a further confirm was found by means of a rigorous statistical analysis [12] . More refined algorithms are summarized by Kim [15] ; they take into account spatial frequency distribution [16] or are based on segmentation [17,18] ; maximum likelihood estimators are employed [15] or a priori knowledge in relation to a given kind of images [18] . Such probabilistic or heuristic approaches are not required here because the source of noise is well known (the back-lighting system) and it can be quantified in a deterministic way (having assumed that the relative position between the camera and the back-lighting system remains fixed, in the case of camera acquisition).…”

Volume Estimation of Various Brain Components Using MR Images - A Technical Report

“…If the threshold for gray matter is set higher, then the white matter at surface layer may be misclassified as gray matter; if the threshold for gray matter is set lower, then some central gray matter may be misclassified as white matter. Many unsupervised segmentation methods have been developed to address this issue, such as non-local methods [3,4] and bias-correction methods [5][6][7]. However, researchers…”

“…However, due to the closeness of intensities between central gray matter and white matter, much central gray matter was misclassified to white matter. There have been some papers addressing the problem of inhomogeneity of intensities, such as biascorrection based multiphase image segmentations [5,7,[11][12][13] and nonlocal information or global information based image segmentations [3,4]. However, all of those methods, including the unsupervised segmentation method employed inBrainImg Processing, do not work very well for real MR brain images due to the existence of central gray matter.…”

A Level Set Method for Image Segmentation in the Presence of Intensity Inhomogeneities With Application to MRI

“…As an important application, our method can be used for segmentation and bias correction of magnetic resonance (MR) images. Note that this paper is an extended version of our preliminary work presented in our conference paper [9]. This paper is organized as follows.…”