The Electronic Portal Image Device (EPID) is commonly used to evaluate the difference between the mechanical and radiation centers of a linear accelerator; however, the gantry and EPID are limited in some non-coplanar angles due to collisions with the treatment couch. This study aimed to compare the results of mathematical calculations and actual measurements to evaluate the feasibility of assessing the differences in the mechanical and radiation field isocenters. A 20 mm Stereotactic Circular Collimator was mounted on the head of the linear accelerator (LINAC), and an Elekta 8 mm ball-bearing (BB) was placed on the couch, following the Winston-Lutz (WL) measurement method. Measurements were taken with varying gantry and couch positions, and Image J software was used to analyze the images. Two regions with a difference of more than 0.1 mm were identified. The first region was between T270 and T330, with gantry angles ranging from G120 to G160. The second region was between T280 and T340, with gantry angles ranging from G290 to G330, and the two regions showed a symmetric diagonal relationship in terms of the gantry angles. The paired sample t-test showed no significant difference between the actual measurements and the calculated results (p-value = 0.11). The study concluded that mathematical calculations can complement the limitations imposed by equipment collisions, providing valuable insights for treatment planning.