TRACKING OF TUMOR MOTION IN LUNG CANCER USING PATIENT SPECIFIC FINITE ELEMENT MODELING AND 4D-MRI IMAGE DATA

摘要

This paper presents a study that demonstrates the potentialof using finite element (FE) lung model constructed using4D-MRI (3D + time) for tracking tumor motion during arespiratory cycle. A series of volumetric images of onelung cancer patient was acquired over time under freebreathing and sorted into respiratory phases. A FE modelof the lung with the tumor was constructed using thevolume which is at full exhale phase. Displacement fieldfrom this initial volume to the subsequent 3D volumes inthe respiratory phases were derived using a deformableimage registration technique. This displacement fieldwhich provides displacement information of the lungsurface is then used to predict the tumor motion in thelung interior using the FE model. Our results showed thatthe tumor motion (as represented by the trajectory of thetumor centroid) follows a highly non-linear path duringthe respiratory cycle from the full exhale phase to the fullinhale phase. We also showed that the predicted tumormotion from our FE model is in reasonable agreementwith that computed from 4D-MRI.