Title: Artificial neural networks for acquisition and processing of sensors data in a radiotherapy application

Authors: Kheireddine Lamamra; Abdelkrim Allam; M'hammed Afiane

Addresses: Department of Electrical Engineering, University of Larbi Ben Mhidi, Oum El Bouaghi, Algeria ' Division of Architecture and Multimedia Systems, Development Centre of Advanced Technologies, Baba Hassan, Algiers, Algeria ' Department of Radiotherapy, Hospital of Pierre and Marie Curie Centre, Place of May 1st, Algiers, Algeria

Abstract: This paper presents a practical part of work that we have begun to realise for several months and it is planned for several steps. In this paper, a part devoted to acquisition and processing of coded data from temperature sensor of type MS6503 used in radiotherapy rooms of the Hospital Pierre and Marie Curie Centre (PMCC) is presented. The aim is to acquire and check remotely the temperatures of rooms to trigger alarms and their control thereafter in order to avoid mistakes of manipulation which are deadly for patients if they happen or arise. For this, a system modelling is made before proceeding to the implementation in practice. During the implementation, several problems have occurred such as the legibility of the received data that has been encrypted. To overcome this problem, an artificial neural network type of multi-layer perceptron (MLP) is used to acquire and decrypt the temperature data received from the sensors placed in the treatment rooms. The obtained results show that the neural network used has decrypted well the received data. It is the reason for which this technique has been implemented in the realised solution.

Keywords: data acquisition and processing; temperature sensor; radiotherapy rooms control; artificial neural network; modelling.

DOI: 10.1504/IJSPM.2018.090257

International Journal of Simulation and Process Modelling, 2018 Vol.13 No.1, pp.15 - 23

Received: 07 Jun 2016
Accepted: 18 Dec 2016
Published online: 07 Mar 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article