Software evaluations of EPID based Winston-Lutz measurements

Thursday, September 23 2021


“SOFTWARE EVALUATIONS OF EPID BASED WINSTON-LUTZ MEASUREMENTS.” Fourie, H, African Journal of Medical Physics, Volume 2 (Suppl), 2019,


H Fourie - Netcare Hospitals Pty. Ltd


Linacs used for SRS and SBRT treatments require routine, and sometimes patient-specific, Winston-Lutz (WL) tests to confirm the precision and accuracy of the machine isocentre. EPID based WL measurement is a time and resource saving method, however analysis of the images by software and the believability of the results are still being investigated. 

Materials and Methods:
We aim to investigate the influence of WL marker size (8 mm, 5 mm, and 3 mm diameter ball bearings) and radiation field size (10x10 cm2, 6x6 cm2, and 2x2 cm2) on the WL analyses. Additionally the results of different analysis software (PTW IsoCheck and Sun Nuclear Machine Check) will be compared. We will attempt to determine the accuracy of the software results by comparing them to film measurements. 

The differences between the methods and algorithms used to calculate the WL isocenter make comparisons between the PTW and Sun Nuclear softwares difficult. The PTW software allows the use of the Minimum Distribution algorithm, which reports the mean WL isocenter of the EPID images and the radius of the isocenter sphere (defined as the difference between the WL isocenter and the measurement point farthest away from it). The Sun Nuclear software reports the maximum deviation from the image centre as the radius of the isocenter sphere. In both sofwares the 6x6 cm2 and 2x2 cm2 resulted in better WL results independent of marker size, for both gantry and couch rotations. More successful ball bearing detections occurred with the 8 mm and 5 mm WL pointers. 

The use of “smaller” field sizes along with “larger” WL pointers resulted in better and more successful WL analyses, independent of software, for both gantry and couch motions.