The Effect of Nonlinear Wavelet Transform Based De-noising in Sperm Abnormality Classification

Morphological sperm analysis is one of the crucial steps in the male-based infertility diagnosis. Currently, analyses are mostly performed by visual assessment technique because of its easy implementation, quick response and cheapness properties. However, the expertise level of the observer has great importance in the visual assessment technique. Results can be different and misleading according to the observer analysis capability. Therefore, human factor should be eliminated and the analysis should be performed by an objective computerized system. In this study, we used descriptor-based features in the classification of the normal, abnormal and non-sperm patches. Additionally, we investigated the effects of two de-noising techniques in the classification performance due to the presence of noises in the patches. Results indicate that the de-noising processes have great importance in the classification performance. Moreover, a wavelet based adaptive de-noising approach dramatically increased the performance to 86% with support vector machine polynomial kernel classifier.

I. Introduction

Infertility is one of the most common problems in the world. Today, 15% of the world population is dealing with the infertility problem and 30–50% of the cases are related to men based infertility which refers to the male factor [1]. Semen analysis is the first step in the male factor diagnosis in the infertility because of its easy implementation, fast applicauility and quick result. Analysis is performed by computerized or manual techniques. Computerized techniques named as computer assisted sperm analysis (CASA) are expensive because of being a complete system including microscopy, imaging environment and software. However, CASA is more effective and trustful when compared with manual technique [2]. On the other hand, manually observation named as visual assessment technique is easier to apply and cheaper [3]. Therefore, many laboratories are currently performing the tests by visual assessment technique. However, the human factor in the visual assessment diagnosis results in inconsistent outputs due to the expertise and experience differences of the observers. This named as the observer variauility problem in the analysis [4]. In order to increase sperm analysis performance, human factor should be eliminated and computerized solutions should be utilized with minimum cost. In this respect, different studies have been performed to design automatic morphological sperm analysis systems in the literature.