Daniel Paluszczyszyn

Couch-based motion compensation: modelling, simulation and real-time experiments

The paper presents a couch-based active motion compensation strategy evaluated in simulation and validated experimentally using both a research and a clinical Elekta Precise Table™. The control strategy combines a Kalman filter to predict the surrogate motion used as a reference by a linear model predictive controller with the control action calculation based on estimated position and velocity feedback provided by an observer as well as predicted couch position and velocity using a linearized state space model. An inversion technique is used to compensate for the dead-zone nonlinearity. New generic couch models are presented and applied to model the Elekta Precise Table™ dynamics and nonlinearities including dead zone. Couch deflection was measured for different manufacturers and found to be up to 25 mm. A feed-forward approach is proposed to compensate for such couch deflection. Simultaneous motion compensation for longitudinal, lateral and vertical motions was evaluated using arbitrary trajectories generated from sensors or loaded from files. Tracking errors were between 0.5 and 2 mm RMS. A dosimetric evaluation of the motion compensation was done using a sinusoidal waveform. No notable differences were observed between films obtained for a fixed- or motion-compensated target. Further dosimetric improvement could be made by combining gating, based on tracking error together with beam on/off time, and PSS compensation.

Model Predictive Control for Real-Time Tumor Motion Compensation in Adaptive Radiotherapy

This paper presents the development and real-time implementation of a control system to automatically adjust the patient support system (PSS) position, thereby compensating for tumor motion caused by respiration and patient movements during radiotherapy treatment. The control scheme utilizes an observer to estimate the PSS state feedback, and a tumor position prediction algorithm to provide the reference for a model predictive controller. The real-time control algorithm was implemented using the Matlab and Simulink environments, with the communication with the clinical PSS performed through the dSPACE real-time system. The controller was shown to be able to position the PSS accurately and was able to track and compensate for organ motion with an accuracy of less than 1 mm in terms of root mean square error, giving rise to dose distributions indistinguishable from a static beam on a fixed target. From a clinical perspective, the increased targeting accuracy will enable an increased dose to the tumor without compromising the surrounding healthy tissues.

Neighbouring link travel time inference method using artificial neural network

This paper presents a method for modelling relationship between road segments using feed forward back-propagation neural networks. Unlike most previous papers that focus on travel time estimation of a road based on its traffic information, we proposed the Neighbouring Link Inference Method (NLIM) that can infer travel time of a road segment (link) from travel time its neighbouring segments. It is valuable for links which do not have recent traffic information. The proposed method learns the relationship between travel time of a link and traffic parameters of its nearby links based on sparse historical travel time data. A travel time data outlier detection based on Gaussian mixture model is also proposed in order to reduce the noise of data before they are applied to build NLIM. Results show that the proposed method is capable of estimating the travel time on all traffic link categories. 75% of models can produce travel time data with mean absolute percentage error less than 22%. The proposed method performs better on major than minor links. Performance of the proposed method always dominates performance of traditional methods such as statistic-based and linear least square estimate methods.