Lung Tumor Tracking and Couch Control - A Feasibility Study

We look at the problem of real-time tumor tracking and motion compensation in radiotherapy techniques from a pure feedback-control point of view and try to understand it using simple mechanical models of the treatment couch and breathing patterns. Since we want to control a robotic couch based on the data collected in real-time, we propose the use of an acceleration sensor to develop a feedback control system to correct for the random movements of the tumor. An acceleration sensor offers a much richer set of data and in situations of rapid change in the breathing pattern, an acceleration sensor can provide a quick measurement and prediction of the trend. Our simplified models and the prototype show stable response under normal conditions.