International Journal of Radiation Research

en Evaluation of statistical prediction model fitting by combining fiducial markers and lung volume for stereotactic body radiotherapy Radiation Biology Radiation Biology تحقيق بديع Original Research <div style="text-align: justify;">Background: This study evaluates the tracking accuracy of the lung tumor volume and fiducial markers using four-dimensional computed tomography (4DCT) and fitted neural network models. Materials and Methods: This study utilized data from 31 patients (109 fiducial markers) who underwent stereotactic body radiotherapy (SBRT). The respiratory movements of fiducial markers, lung tumors, and lung volume were calculated using 4DCT. Cross-correlation coefficients were then calculated to analyze the phase movements of fiducial markers, lung tumors, lung volume, direction displacement, and lung area (upper, middle, and lower lobes). Statistical prediction models were used to evaluate the predictive accuracy of the left–right (LR), anterior–posterior (AP), cranial–caudal (CC), and three-dimensional (3D) cross-correlation coefficients between fiducial markers and lung tumors. The coefficient of determination (R2) was used to determine the accuracy of the statistical prediction models for the explanatory variables. Results: The correlation between fiducial marker and lung tumor, as well as lung tumor movement by time phase, yielded the following R2 values—LR: 0.920, AP: 0.319, CC: 0.675, and 3D: 0.449 for the upper lobe; LR: 0.567, AP: 0.627, CC: 0.955, and 3D: 0.939 for the middle/lower lobes. Statistically significant differences were observed in the CC and 3D directions within each lower lobe. Conclusions: The respiratory movements of fiducial markers and lung tumors in lung SBRT show stronger correlation with the movement of lung volume in the middle/lower lobes compared to that in the upper lobes. Combining a fiducial marker with lung volume improves the prediction accuracy of the respiratory movement of lung tumors.</div> Pelvis, Stereotactic body radiotherapy, Four-dimensional computed tomography, Respiratory movement, Statistical prediction model, lung cancer. 17 22 http://ijrr.com/browse.php?a_code=A-10-1-1132&slc_lang=en&sid=1 Y. Tanabe tanabey@okayama-u.ac.jp 7900319475328460025825 7900319475328460025825 Yes Faculty of Medicine, Graduate School of Health Sciences, Okayama University, 2-5-1, Shikata, Kita, Okayama 700-8525, Japan T. Deguchi 7900319475328460025826 7900319475328460025826 No Facility of Health Sciences, Okayama University Medical School, 2-5-1, Shikata, Kita, Okayama 700-8525, Japan M. Kiritani 7900319475328460025827 7900319475328460025827 No Facility of Health Sciences, Okayama University Medical School, 2-5-1, Shikata, Kita, Okayama 700-8525, Japan N. Hira 7900319475328460025828 7900319475328460025828 No Facility of Health Sciences, Okayama University Medical School, 2-5-1, Shikata, Kita, Okayama 700-8525, Japan S. Tomimoto 7900319475328460025829 7900319475328460025829 No Facility of Health Sciences, Okayama University Medical School, 2-5-1, Shikata, Kita, Okayama 700-8525, Japan H. Nishikawa 7900319475328460025830 7900319475328460025830 No Facility of Health Sciences, Okayama University Medical School, 2-5-1, Shikata, Kita, Okayama 700-8525, Japan S. Tsumoto 7900319475328460025831 7900319475328460025831 No Facility of Health Sciences, Okayama University Medical School, 2-5-1, Shikata, Kita, Okayama 700-8525, Japan H. Tanaka 7900319475328460025832 7900319475328460025832 No Department of Radiation Oncology, Yamaguchi University Graduate School of Medicine, 1-1-1 Minamikogushi, Ube, Yamaguchi 755-8505, Japan