Abstract:
This study focuses on developing a geometric algorithm to identify the position and size of bones in X-ray images, aimed at supporting diagnostic and tracking processes in medical treatment. Accurate bone positioning assists in analyzing patients' anatomical structures and identifying injuries or other medical information effectively. The algorithm leverages statistical data from existing datasets to enhance accuracy and applies the Active Shape Model (ASM) method using an elliptical model, which adjusts in size and orientation to align with the X-ray images. Results show a Jaccard index above 0.85, indicating high accuracy in estimating the positions of the pelvis bones on both the left and right sides. This achievement forms a crucial foundation for detecting other structures, such as the lower spine or coccyx, and holds potential for applications in the medical field to improve diagnostic efficiency.