Detection of Singularity in the Cell Nucleus of Synovial Sarcoma Using Wavelet Leaders

Abstract

Pathological examination is important for an accurate diagnosis of Synovial Sarcoma (SS). It is the most common cancer of the soft tissues of the limb in adolescents and adults. In this work, SS was used to determine the discriminant singularity characteristics using Wavelet Leaders (WL). The most popular technique for measuring the discriminant singularity characteristics of an image signal is the Lipschitz Exponent (LE). The singularity measurement was based on LE function by taking a slope of logarithmic scale versus logarithmic Wavelet Transform Modulus Maxima (WTMM). Here, the presence of the singularity was measured using WTMM and WL by summing each color component of an image signal. The performance characteristics of the statistical discrimination are evaluated and compared with the non-parametric hypothesis using the Wilcoxon rank-sum test. The most important difference between WTMM and WL was analyzed using the Receiver Operating Characteristics (ROC) curve. From the experimental analysis that the WL method provides excellent discriminant singularities or discontinuities performance characteristics such as area, standard error, z-statistics, and p-values. Finally, the results of experiments have proven that a WL can express practical, precise, robust, and satisfactory performance in practice.