This paper proposes a method for linear-structure (LS) verification in mammography computer-aided detection (CAD)
systems that aims at reducing post-classification microcalcification (MCC) false-positives (FPs). It is an MCC cluster-driven
method that verifies linear structures with a small rotatable band that is centered on a given MCC cluster
candidate. The classification status of an MCC cluster candidate is changed if its association with a linear structure is
confirmed through LS verification. There are mainly four identifiable features that are extracted from the rotatable band
in the gradient-magnitude and Hough parameter spaces. The LS verification process applies cascade rules to the
extracted features to determine if an MCC cluster candidate resides in a linear structure area. The efficiency and efficacy
of the proposed method are demonstrated with results obtained by applying the LS verification method to over one
hundred cancer cases and over one thousand normal cases.
Image segmentation is an essential process for quantitative analysis. Segmentation of brain tissues in magnetic resonance (MR) images is very important for understanding the structural-functional relationship for various pathological conditions, such as dementia vs. normal brain aging. Different brain regions are responsible for certain functions and may have specific implication for diagnosis. Segmentation may facilitate the analysis of different brain regions to aid in early diagnosis. Region competition has been recently proposed as an effective method for image segmentation by minimizing a generalized Bayes/MDL criterion. However, it is sensitive to initial conditions -- the "seeds", therefore an optimal choice of “seeds” is necessary for accurate segmentation. In this paper, we present a new skeleton-based region competition algorithm for automated gray and white matter segmentation. Skeletons can be considered as good "seed regions" since they provide the morphological a priori information, thus guarantee a correct initial condition. Intensity gradient information is also added to the global energy function to achieve a precise boundary localization. This algorithm was applied to perform gray and white matter segmentation using simulated MRI images from a realistic digital brain phantom. Nine different brain regions were manually outlined for evaluation of the performance in these separate regions. The results were compared to the gold-standard measure to calculate the true positive and true negative percentages. In general, this method worked well with a 96% accuracy, although the performance varied in different regions. We conclude that the skeleton-based region competition is an effective method for gray and white matter segmentation.