A Fast Model for Prediction of Respiratory Lung Motion for Image-Guided Radiotherapy: A Feasibility Study



Zehtabian M¹ ; Faghihi R¹ ; Moslehshirazi MA² ; Shakibafard AR³ ; Mohammadi M⁴ ; Baradaranghahfarokhi M⁵ Show Affiliations
Source: Iranian Journal of Radiation Research Published:2012

Abstract

Background: The aim of this work was to study the feasibility of constructing a fast thorax model suitable for simulating lung motion due to respiration using only one CT dataset. Materials and Methods: For each of six patients with different thorax sizes, two sets of CT images were obtained in single-breath-hold inhale and exhale stages in the supine position. The CT images were then analyzed by measurements of the displacements due to respiration in the thorax region. Lung and thorax were 3D reconstructed and then transferred to the ABAQUS software for biomechanical fast finite element (FFE) modeling. The FFE model parameters were tuned based on three of the patients, and then was tested in a predictive mode for the remaining patients to predict lung and thorax motion and deformation following respiration. Results: Starting from end-exhale stage, the model, tuned for a patient created lung wall motion at end-inhale stage that matched the measurements for that patient within 1 mm (its limit of accuracy). In the predictive mode, the mean discrepancy between the imaged landmarks and those predicted by the model (formed from averaged data of two patients) was 4.2 mm. The average computation time in the fast predictive mode was 89 sec. Conclusion: Fast prediction of approximate, lung and thorax shapes in the respiratory cycle has been feasible due to the linear elastic material approximation, used in the FFE model.