Identification of Regions of Interest in Digital Mammograms

Abstract

The main purpose of this paper is to compare clustering (region growing) and gradient based techniques for detecting regions of interest in digital mammograms. Such regions of interest form the basis of applying shape and texture techniques for detecting cancerous masses. In addition, the paper proposes a two-stage method, in which gradient based techniques are applied first, followed by a region growing method that will yield lesser numbers of regions for analysis. For this purpose, we first use histogram equalization and fuzzy enhancement techniques to improve the quality of the images and to compare their utility on our mammogram data. Image-enhanced mammograms are then subjected to clustering or to gradient operations (masking) for segmentation purposes. The segmented image is then analyzed for estimating the regions of interest, and the results are compared against the previously known diagnosis of the radiologist. A total of 30 mammograms from the University of South Florida database were used, for which the radiologist's hand-sketched boundaries of the masses were known. The results show that when compared with histogram equalization, fuzzy enhancement techniques are better suited for mammogram analysis, and when compared with gradient based segmentation, region growing segmentation will give a lesser number of regions for analysis without compromising on quality.

Modified Global and Modified Linear Contrast Stretching Algorithms: New Colour Contrast Enhancement Techniques for Microscopic Analysis of Malaria Slide Images

“…By using DSM, a measure of separation between these two PDFs would be an indicator of the performance of the proposed contrast enhancement techniques on malaria images. The procedures to apply the quantitative measure of contrast enhancement techniques on malaria images are as follows. Apply the four contrast enhancement techniques which are GCS, LCS, MGCS, and MLCS on malaria images. Apply manual segmentation on both original and enhanced malaria images in order to obtain the target, T (parasite), and background, B , regions as shown in Figure 3. Calculate the best decision boundary for the original image between the target and background regions based on the following equation [21]: $\begin{array}{c}D\mathrm{1}=\frac{\left({\mu }_B^O{\sigma }_T^O\right)+\left({\mu }_T^O{\sigma }_B^O\right)}{\left({\sigma }_T^O+{\sigma }_B^O\right)},\end{array}$ where μ T O , σ T O , μ B O , and σ B O are the mean and standard deviation for each of the RGB components comprising the parasite and background regions for the original image, respectively. Calculate the best decision boundary for the enhanced image between the target and background regions based on the following equation [21]: $\begin{array}{c}D\mathrm{2}=\frac{\left({\mu }_B^E{\sigma }_T^E\right)+\left({\mu }_T^E{\sigma }_B^E\right)}{\left({\sigma }_T^E+{\sigma }_B^E\right)},\end{array}$ where μ T E , σ T E , μ B E , and σ B E are the mean and standard deviation for each of the RGB components comprising the parasite and background regions for the enhanced image, respectively.Calculate the value of DSM based on the following equation [21]: …”

Mammogram image visual enhancement, mass segmentation and classification

“…Feature based enhancement methods used in [12,19,46,48,52] are based on wavelet domain enhancement. Finally, the fuzzy enhancement techniques as in the work of [51,38,28,16,56] apply fuzzy operators and properties to enhance mammogram features.…”

“…Clustering segmentation methods have been employed in [50,42,59,45,11,61,60,14,64]. Regions of interest segmentation using a single view and the multiple views are used to segment both masses and calcifications, those methods have been used in the work of [49,29,23,51,37,58,66,10,40,53,39,9,15,18]. Graph segmentation methods can also be used to segment masses, and are applied in [4,33,34,20,22,21,67].…”

Predicting the Severity of Breast Masses with Ensemble of Bayesian Classifiers

“…It is the second largest cause of cancer deaths among women. Mammography is one of the most used methods to detect this kind of cancer (Choua et al, 2004;Singh and Al-Mansoori, 2000). The value of mammography is that it can identify breast abnormalities with 85-90% accuracy.…”