Automatic detection of liver tumor motion by fluoroscopy images

Background: A method to track liver tumor motion signals from fluoroscopic images without any implanted gold fiducial markers was proposed in this study to overcome the adverse effects on precise tumor irradiation caused by respiratory movement. Materials and Methods: The method was based on the following idea: (i) Before treatment, a series of fluoroscopic images corresponding to different breathing phases and tumor positions were acquired after patient set-up; (iii) The wavelet transform method and Canny edge detection algorithm were used to detect motion trajectory of the diaphragm; (iv) The motion curves of center of lipiodol in the images were obtained by mathematical morphology and median filtering algorithm. The method was evaluated using by five sequences of fluoroscopic images from TACE patients who received transcatheter arterial chemoembolization therapy.  Results: The position of liver tumor was significantly affected by respiratory motion; the motion trajectories of the diaphragm and lipiodolagreed well with the manually marked locations in amplitude and period; the motion trajectories of the diaphragm and lipiodol almost had similar period and amplitude in one  treatment fraction. The respiratory period and amplitude of the same patient in different fractions had no significant differences; however, the difference was obvious for different patients. The proposed lipiodol detection methods can effectively reflect the relevant rules of tumor location caused by respiratory movement. Conclusion: Direct tracking of liver tumor motion in fluoroscopic images is feasible. The automatic detection method can reflect the characteristics of respiratory and tumor motions, which can save much time and significantly improve measurement precision compared with manual measurement.