Click here to


Are you sure ?

Yes, do it No, cancel

Study of Automated Edge Detection for QA of Proton Beam Distal Uniformity

V Ugarte*, Loma Linda University Medical Center, Loma Linda, California


(Sunday, 7/29/2018) 3:00 PM - 6:00 PM

Room: Exhibit Hall

Purpose: To study the viability of edge detection algorithm as a reliable QA technique for proton distal uniformity QA. The effect of the variation in distal depth, film optical density, and proton energy on the uniformity of dose coverage is investigated.

Methods: EDR2 films were exposed using for all cardinal gantry angles using mod wheel #60 and 13.8cm diameter aperture in a solid water phantom with the center of modulation (COM) depth at the gantry isocenter. Individual films were exposed to proton beam with energies of 149, 186, and 250 MeV; the films were placed behind a picket fence at the distal edge (DE). Dose profiles from each film were obtained by using department calibration curves and ASCII data was exported for analysis. The effect of film optical density on DE uniformity was studied using a lateral field penumbra defined within the range 50% to 90% to calibrate the edge detection profile; DE uniformity was obtained for two distinct cases: 1) small penumbra and 2) large penumbra.

Results: The 2D edge detection profile based on two optical density calibration schemes exhibit a DE uniformity in the range of 0.01mm to 0.14mm. DE uniformity displayed proton energy dependence consistent with distal penumbra. The efficiency of edge detection algorithms to correctly identify changes within proton range was achieved for smoothing windows between 11 pixels to 21 pixels.

Conclusion: The result indicates an opportunity to improve the workflow in the proton beam QAs. The automated process provides a significant decrease in time used for DE uniformity QA. With a decrease in error introduced by individual QA physicists while improving consistent monitoring of DE uniformity.


Protons, Quality Assurance, Image Processing


IM- Particle (e.g., proton) CT: Quality control

Contact Email