Optimal control in high intensity focused ultrasound surgery

Abstract

When an ultrasound wave is focused in biological tissue, a part of the energy of the wave is absorbed and turned into heat. This phenomena is used as a distributed heat source in ultrasound surgery, in which the aim is to destroy cancerous tissue by causing thermal damage. The main advantages of the ultrasound surgery are that it is noninvasive, there are no harmful side effects and spatial accuracy is good. The main disadvantage is that the treatment time is long for large cancer volumes when current treatment techniques are used. This is due to the undesired temperature rise in healthy tissue during the treatment. The interest for optimization of ultrasound surgery has been increased recently. With proper mathematical models and optimization algorithms the treatment time can be shortened and temperature rise in tissues can be better localized. In this study, two alternative control procedures for thermal dose optimization during ultrasound surgery are presented. In the first method, the scanning path between individual foci is optimized in order to decrease the treatment time. This method uses the prefocused ultrasound fields and predetermined focus locations. In the second method, combined feedforward and feedback controls are used to produce desired thermal dose in tissue. In the feedforward part, the phase and amplitude of the ultrasound transducers are changed as a function of time to produce the desired thermal dose distribution in tissue. The foci locations do not need to be predetermined. In addition, inequality constraint approximations for maximum input amplitude and maximum temperature can be used with the proposed method. The feedforward control is further expanded with a feedback controller which can be used during the treatment to compensate the modeling errors. All of the proposed control methods are tested with numerical simulations in 2D or 3D.