Stochastic Extraction of Elongated Curvilinear Structures in Mammographic Images

The extraction of elongated curvilinear structure in mammographic images is an important objective for the automated detection of breast cancers. We develop an approach which relies on a fixed-grid, localized Radon transform for line segment extraction and a Markov random field model to incorporate local interactions and refine the line structure. The energy of the resulting distribution is minimized stochastically via a Markov chain Monte Carlo iterative procedure. Experimental results demonstrate that the method can accurately extract blurred and low-contrast elongated continuous curvilinear structures, including those radiating from cancerous masses.