An Adaptive Control Method for Ultrasound Prostate Hyperthermia

L. Sun, J. Schiano, and N.B. Smith (USA)

Keywords

self-tuning regulator, ultrasound hyperthermia, prostate disease

Abstract

However, in the clinical treatments, the temperature distribution is controlled by a clinician’s intervention and judgment through adjusting the power to the array elements based on temperature information from thermocouples placed in the prostate. Previously, various control strategies have been designed for controlling temperature in either actual or simulated hyperthermia treatment, such as proportional-integral (PI) bang-bang control [6], reduced-order multi-input multi-output (MIMO) control [7], adaptive MIMO control [8], multipoint adaptive control [9], linear quadratic regulator (LQR) [10], magnetic resonance imaging (MRI) compatible single input single output (SISO) PI control [11]. The purpose of this research is to introduce an adaptive feedback control method for ultrasound hyperthermia treatment to achieve the desired prostate temperature with minimal overshoots, rapid rise time, fast settle time, and small oscillations. For thermal treatment of prostate diseases, an ultrasound phased array was designed and operated with a computer controlled amplifier system which adjusted the power and phase of each transducer element. The clinical application of such a system is to increase the temperature in the prostate to 43-45o C and maintain the temperature for 30 60 minutes for a hyperthermia therapy treatment. In this paper, an adaptive self-tuning regulator (STR) controller has been designed and implemented with this hyperthermia system using a negative feedback transfer operator and a feedforward transfer operator. The transfer operators’ parameters were obtained directly from plant input and output with recursive least square estimation (RLSE). The advantage of this controller was that it did not need a priori knowledge of the tissue properties and could adaptively change its control variables according to the perfusion rate or other dynamic properties. Simulations indicated that the prostate reached the target temperature without overshoots and oscillations within 100 seconds. The system also successfully adapted to the dynamic tissue properties. Phantom experiments showed that the measured temperature tracked the reference temperature closely and reached the target temperature with a little oscillations and without overshoots within 150 seconds, which was consistent with the computer simulations. 2. Materials and Methods 2.1 The Ultrasound Hyperthermia System The ultrasound hyperthermia system consisted of an ultrasound phased array driven by a computer controlled amplifier system (Fig. 1) [4;10-15]. In vitro prostate phantom experiments used a 16-element ultrasound unfocused array immersed in a Plexiglas® tank (25 x 38 x 53 cm3 ) filled with distilled degassed water. The array was driven by a programmable 64-channel radio frequency (RF) amplifier (Advanced Surgical System Incorporated, Tucson, AZ) connected to the computer through a RS-232 port [16;17]. Temperatures change within the in vitro phantom was determined by a multi-element fiber optic thermometer (Luxtron Corporation, Mountain View, CA) inserted in the prostate phantom. Each probe was shielded with copper tube to prevent heat absorption of the ultrasound by the probe [18]. The prostate phantom was made of agar powder simulating the response of human tissue to capture the temperature of the heated tissue [19]. The adaptive feedback control algorithm was used to maintain a desired temperature evolution by controlling the electrical power of the individual phased array element.

Important Links:



Go Back