A Concentric Morphology Model for the Detection of Masses in Mammography

Abstract

We propose a technique for the automated detection of malignant masses in screening mammography. The technique is based on the presence of concentric layers surrounding a focal area with suspicious morphological characteristics and low relative incidence in the breast region. Mammographic locations with high concentration of concentric layers with progressively lower average intensity are considered suspicious deviations from normal parenchyma. The multiple concentric layers (MCLs) technique was trained and tested using the craniocaudal views of 270 mammographic cases with biopsy proven malignant masses from the digital database of screening mammography. One-half of the available cases were used for optimizing the parameters of the detection algorithm. The remaining cases were used for testing. During testing, malignant masses were detected with 92%, 88%, and 81% sensitivity at 5.4, 2.4, and 0.6 false positive marks per image. Testing on 82 normal screening mammograms showed a false positive rate of 5.0, 1.7, and 0.2 marks per image at the previously reported operating points. Furthermore, additional evaluation on 135 benign cases produced a significantly lower detection rate for benign masses (61.6%, 58.3%, and 43.7% at 5.1, 2.8, and 1.2 false positives per image, respectively). Overall, MCL is a promising computer-assisted detection strategy for screening mammograms to identify malignant masses while maintaining the detection rate of benign masses considerably lower.

Automatic Mass Segmentation Based on Adaptive Pyramid and Sublevel Set Analysis

“…However, mass regions often yield a pattern of concentric rings with gradually declining intensities when moving from the inner to the outer ring. In [2], within a total number of 281 malignant masses analyzed, 96% of masses were found having multiple concentric rings. Only 7 of these 281 masses had concentric rings fewer than 3 concentric rings.…”

“…The set of pixels that lie within a level comprise a sublevel set. More formally, let Á be the intensity range of a mammogram, a partition of the intensity range is Á´Èµ In the literature, other names have been used to describe the sublevel set, such as isolabel contour in [9], iso-level contour [3] and concentric layers [2]. The different names indicate different approaches.…”

“…[3] used the term iso-level contour as that study also focused on the contour formed by the pixels in the same intensity level. Concentric layers used in [2] had the same definition as the sublevel set used in this paper. That method used the concentric properties between the layers.…”

“…The difficulty of segmentation in mammography is further compounded by the noise inevitably introduced during the image acquisition process. Medical image segmentation based on level sets has been studied in the literature [9,3,2] and has been shown capable of segmenting ROIs (In literature, different names were used. A discussion on this is presented in Section 3.1).…”

“…However, accuracy of the extraction was not evaluated. [2] first detected seeds from mammograms, and then extracted concentric layers for each seeds based on the distance between the centroids of consecutive layers. The method was evaluated on multiple sets of mammograms and sensitivities and detection rates were analyzed.…”