Meeting Banner
Abstract #1744

Template Matching Can Accurately Track Tumor Motion in Cine MRI Images from Lung Cancer Patients

Xiutao Shi1, Tejan Diwanji1, Karen E. Mooney1, Warren D. D'Souza1, Nilesh Mistry1

1Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, MD, United States

Accurate tracking of tumors in lung cancer is crucial for safe and effective delivery of radiation treatment. In the current study, an automated template matching technique has been applied to track respiration induced tumor motion based on cine MRI images. The performance of the technique is compared to manual estimates of tumor position. For tumors smaller than 50 mm in the longest dimension, the prediction error is < 3.9 mm (less than two pixels) and is in the order of intra-operator or inter-operator variability during manual tracking. Further hardware developments can enable clinical translation.

Abstract PDF

Keywords

accurate accurately acknowledgments acquisition allowed among anterior applied automate automated automatic bars biggest board boxes breathe calculating cancer challenging cine complete contoured contribute correlated correlation correlations coverage create cross crucial cyan delivery determine determined determining development deviation diaphragm dimension dimensions distinctive drawing enable enables enclosed ensure error errors establish established evaluating event eventually examined fast frame frames furthermore ground heart help hybrid induce induced instructed intra laborious longest lung lungs management manual manually matching matrix medicine motion move necessary normally occurred oncology patient patients patterns performance pixels position positions positively prediction process prospective radiation ranging real recruited repeat representative resolution respiration respiratory restricted restrictions school search selected series served session sessions signifying slice slices smaller smallest solid starting structure structures subsequent subsequently substantially successfully suggest support surrogates surrounding systems table template temporal thoracic tissue track tracking treatment true truth tumor tumors univ utility utilize validated variability various vascular verification weeks window