Introduction
Radiotherapy is a technique used for the cancer treatment; it uses ionizing radiation to destroy tumor tissues. During radiotherapy, healthy tissues must be preserved, while the dose of irradiation has to be optimally concentrated on the tumor. These tasks are difficult to achieve, especially in the case of pulmonary tumors due to the patient's breathing which induces tumor alterations. Given this situation, the knowledge of the lung tumor position at each step of the respiratory cycle (RC) could dramatically improve the radiological treatment of lung cancer. Using a numerical model is essential to assist the medical staff in this task.
Existing respiratory numerical model simulations are based on respiratory mechanics and physical laws. Such models are available in the laboratory, but they do not consider the electrophysiological variables involved in the process. Several works have been developed in the field of processing of electrophysiological respiratory signals, but these works only describe the variability and complexity of the process in humans, and have not been used as input in a numerical respiratory model. Therefore, the possibility of including information from electrophysiological signals and mechanical process would allow obtaining a better breathing model.
Investigation proposal
The LIRIS-SAARA team has a great experience on computer graphics simulation and has been working for a long time on developing thorax and pulmonary models. The purpose of this thesis is to incorporate electrophysiological variables involved in the respiratory process, using signal and image processing and analysis tools to correlate this information with the mechanical aspects of the process. The aim of this work is to incorporate information provided for a “physiological model”, including:
1.Diaphragm deformation, based on the electrical activity of this muscle measured by electromyography (EMG) and electrical or magnetic stimulation of phrenic nerve.
2.Variation of volume and pressure during RC, measured using spirometry, plethysmography, or extracted from medical image and manometry (mouth and esophageal pressures).
3.Airway features (resistance) and lung elasticity (compliance), deduced using volume-pressure curves.
Finally, we expect to guide a 3D respiratory model combining biomechanical and electrophysiological behaviors. In addition, signal processing and analysis tools development will be adapted to be applied to polysomnographic recordings to classify sleep states and stages and sleep patterns detection. This latter work will be realized within the framework of the collaboration with U. de Chile and some Chilean public hospitals.
Objectives
The main objective of this study is the following: to incorporate electrophysiological variables of the respiratory process as an input for the biomechanical model of breathing (respiratory movements). This could improve the numerical model simulation and allow predicting the respiratory movements of lung tumor during the irradiation process, improving the dosimetry and reducing the adverse effects of radiotherapy treatment. This work will be realized within the framework on the ETOILE project, in the “tumor tracking” workpackage for which the LIRIS is the supervising partner.