Vessel segmentation in computed tomography angiography images is a crucial step in the diagnosis and treatment planning of cerebrovascular diseases. This study analyzes the preliminary segmentation process of four major cerebral arteries: the anterior communicating artery, the basilar artery, the left middle cerebral artery, and the right middle cerebral artery. The segmentation was performed using the Frangi filter, which enhances tubular structures based on eigenvalue analysis of the Hessian matrix. The study was conducted on a dataset of 100 three-dimensional CTA brain images with corresponding binary vessel masks. The primary objective was to determine the optimal Frangi filter parameters— \(\alpha \) , \(\beta \) , \(\gamma \) , and set of \(\sigma \) values —to improve segmentation accuracy. Additionally, the influence of these parameters on segmentation quality was evaluated to understand their role in detecting and enhancing vascular structures. The results indicate that proper selection of these parameters significantly affects segmentation performance, and the identified optimal values can be utilized in future research on automated angiographic image analysis. This segmentation approach is intended as a tool to enhance the performance of deep learning models in vessel detection and analysis.