Polycystic Ovarian Syndrome Identification through Self-Attention Guided Convolutional Neural Network

Abstract

Polycystic Ovarian Syndrome (PCOS) is a hormonal disorder that impacts women during their reproductive years, marked by indicators like multiple ovarian follicles or cysts that can be visualized through ultrasound imaging. Convolution Neural Networks (ConvNets) have been enhanced with self-attention mechanisms to improve their efficacy across a variety of computer vision applications, according to researchers. This study uses self-attention to improve the effectiveness of a ConvNet classifier in classifying PCOS, yielding a superior 99% accuracy, exceeding the 96% accuracy of a regular ConvNet classifier.