It is essential to verify the reconstruction techniques as suggested in ESTRO booklet 8 that institutions should perform the verification of any reconstruction technique employed clinically. However, the seed-position error that undergoes residual analysis is usually not noticed, but is important, since the error may not be in normal distribution. An isocentric beam checker device is used in this work. The device has a two-dimensional array of steel balls embedded on its surface. The checker was placed on a simulator couch and the simulator gantry was rotated to make the checker behave like a three-dimensional array of balls. All points were located within 20 x 20 x 20 cm3, with the isocenter in the middle, and were reconstructed using an orthogonal-film algorithm in the Abacus treatment system. To correctly calculate the standard deviation of the seed-position error, Jarque-Bera and Doornik-Hansen normality tests were used. All normality tests were passed. With 0.1° and -0.1° read errors of the gantry angle and collimator angle, respectively, the mean error was zero for each axis in Cartesian coordinates, with an uncertainty of around 0.74 mm with a 95% confidence interval, which implicated the importance of the normality test. The mean distance error was around 0.28 mm in each axis and the uncertainty of the seed-isocenter distance error did not depend on the seed-isocenter distance. The proposed method can be used for the verification of seed-position error in any facility, and can be applied to evaluate the delivered dose errors of brachytherapy.