ESTRO 2024 - Abstract Book

S3839

Physics - Image acquisition and processing

ESTRO 2024

Figure 1. Methodology of FACT for determining the magnification factor using two photos taken at pre-established heights within the electron applicator, ensuring consistent and repeatable measurements.

The captured images were post-processed using MATLAB software. Initially, images were cropped to isolate the square region that held the contour. This procedure entailed converting the image to grayscale, applying filters, binarization, and skeletonization, ultimately producing a binary representation of the contour against a contrasting background. The resulting binary images were then rigorously processed to generate sorted arrays of the contour's x- and y-coordinates. Using these processed images, a comprehensive geometric framework was employed to derive the 3D coordinates of the contour. By methodically comparing two images, we could ascertain magnification factors and subsequently derive the final contour's coordinates. In this study, contour digitization was performed at every 5 degree interval. Note digitization interval can be easily adjusted as needed (e.g., can be coarser for simpler geometries and finer for more complicated ones). Figure 2 illustrates the workflow of the FACT method.