D. Metaxas et al. (Eds.): MICCAI 2008, Part II, LNCS 5242, pp. 710–717, 2008. © Springer-Verlag Berlin Heidelberg 2008 Real-Time Simulation of 4D Lung Tumor Radiotherapy Using a Breathing Model Anand P. Santhanam 1,2 , Twyla Willoughby 2 , Amish Shah 2 , Sanford Meeks 2 , Jannick P. Rolland 1 , and Patrick Kupelian 2 1 College of Optics and Photonics, University of Central Florida jannick@odalab.ucf.edu 2 Department of Radiation Physics, M.D. Anderson Cancer Center Orlando {Anand.santhanam,twyla.willoughby,amish.shah, Sanford.meeks, Patrick.kupelian }@orhs.org Abstract. In this paper, we present a real-time simulation and visualization framework that models a deformable surface lung model with tumor, simulates the tumor motion and predicts the amount of radiation doses that would be de- posited in the moving lung tumor during the actual delivery of radiation. The model takes as input a subject-specific 4D Computed Tomography (4D CT) of lungs and computes a deformable lung surface model by estimating the defor- mation properties of the surface model using an inverse dynamics approach. Once computed, the deformable model is used to simulate and visualize lung tumor motion that would occur during radiation therapy accounting for varia- tions in the breathing pattern. A radiation treatment plan for the lung tumor is developed using one of the 4D CT phases. During the simulation of radiation delivery, the dose on the lung tumor is computed for each beam independently. 1 Introduction Physics and physiology based lung deformation models have been extensively dis- cussed in application areas ranging from computer aided diagnostics to medical image processing. Recent advances in 4D medical imaging modalities have led to effective medical applications by coupling medical simulations with patient-specific 4D ana- tomical models and their physically and physiologically realistic lung deformations. Of particular importance is the application of these deformable models for areas related to radiotherapy. Lung tumors move during breathing in unpredictable trajecto- ries that depend on patient’s body orientation and physiological condition. The uncer- tainty in tumor location compromises the accurate deposition of radiation doses during radiotherapy. This is particularly a concern when high radiation doses are de- livered with the intent of ablating tumors and subsequently gaining local regional con- trol of tumor growth in an environment that includes radiation sensitive structures such as normal lung tissue and the esophagus. In this paper, we present a method to model, simulate and visualize the radiation dose delivery on a moving 3D lung tumor. The contributions of this paper are (a) a