Prediction of Respiratory Induced Lung Tumor Motion Based on Phenomenological Models of Hysteresis

Dynamic Tumor Tracking Radiotherapy (DTTRT) is a medical technology that irradiates a tumor exhibiting respiratory induced motion inside the body of a patient. It is known that tumor motion prediction is necessary in the implementation of DTT-RT to compensate for the positioning lag of a multi-leaf collimator of a medical linear accelerator. We accordingly develop a tumor motion predictor in this article. The proposed prediction model utilizes the Bouc-Wen hysteresis models that were combined with multi-model estimation algorithm based on a particle filter. Prediction accuracy of the proposed predictor has been validated with five lung tumor motion trajectories.