Analytical Identification of Isolated Clustered Microcalcifications Based on the Distribution of Effective Atomic Numbers

Abstract

The goal of this work was to improve the efficiency of early detection of microcalcifications, the earliest indicators of breast cancer. Studies have shown that the presence of isolated clustered microcalcifications in the mammary gland significantly increases the range of variation of effective atomic number and considerably changes its distribution. Conventional, subtraction, or dividing mammography or their linear convex combination may fail to visualize isolated clustered microcalcifications. However, the presence of microcalcifications in the mammary gland can be determined analytically from the distribution of the extreme terms of an ordered series of effective atomic numbers in the mammary gland. Then, the coordinates of the isolated clustered microcalcifications can be calculated. The distribution of the effective atomic numbers in healthy tissue without microcalcifications is symmetric and close to normal, while breast tissue with even a single isolated cluster of microcalcifications is characterized by an asymmetric distribution and high values of the maximum effective atomic numbers. An algorithm for calculating the coordinates of isolated microcalcifications is suggested.