Respiratory tumor motion estimation during lung brachytherapy

Abstract

Brachytherapy is a newly proposed treatment for lung cancer which involves direct application of radioactive seeds into the tumor. One of the major challenges to achieve an accurate seed implantation pattern which is consistent with the pre-defined dosimetry plan, is tumor motion caused by lung tissue deformation. Tissue deformation in this case is caused mainly by respiratory motion which can result in significant areas of underand/ or over-dosage. To minimize respiratory motion, the lung is usually fully deflated prior to the procedure. However, it is still necessary to predict tumor motion caused by diaphragm contact forces during respiration as the other lung continues breathing during the procedure. The aim of this study, therefore, is to estimate the deflated lung’s tumor motion resulting from the diaphragm contact forces, using a biomechanical approach. To accomplish this, a Finite Element (FE) mesh was generated from a segmented lung CT image using a software package - 3D slicer (a free, open source: http://www.slicer.org ). This mesh was then input into a FE software package for tissue deformation analysis. The FE analysis required for simulating the lung under diaphragm motion is a contact problem in which a contact body mimicking the diaphragm effect is applied to the FE mesh to calculate the deformation. To analyze the effect of the diaphragm’s contact force, Abaqus FE software (Hibbit, Karsson & Sorensen, Inc.) was used. To validate the simulation results, an ex vivo experiment was performed. Qualitative assessment of the deformation obtained from the simulation demonstrates a good agreement with the deformation observed in the experiment. After applying the loading and obtaining the tissue deformation field, an image corresponding to the deformed lung can be reconstructed, accordingly. This can help improve the accuracy of a brachytherapy procedure by predicting the exact location of the tumor during the surgery.