Neural computational approaches to clinical male reproductive data analysis problems

Abstract

Clinical male reproductive medical problems are particularly complex due to the variety of systems that interact to achieve the ultimate reproductive outcome, fertilization. Neural computation offers a robust nonlinear computational modeling tool for andrological data sets. In this mini-symposium, neural computation is reviewed, and aspects of neural computation are discussed, including cross-validation, overlearning and feature extraction. Real world neural computational solutions for clinical andrological problems are given, including modeling outcomes of gamete micromanipulation, testis biopsy, and outcomes after varicocele surgery.