Deep Network Construction using Autoencoder for Abnormality Detection in Radiotherapy Service

Abstract

In this paper, we propose an automatic system based on machine learning algorithms to detect the abnormalities in radiotherapy service. However, challenges are posed regarding quality of service in radiotherapy. Thus, the presented system increases the quality of the control in service. Mainly, the control platform is composed of a set of computers connected with the network for the server. The stored data include checklist of the machines, the temperature, the humidity and the pressure, and operators and management of patients. The main contribution in this field is the use of the classification techniques for avoiding fatal mistakes during the treatment. An encouraging result is obtained by deep network constructed using autoencoders.