Accurate extraction of vessel centerlines is critical for quantitative assessment of vascular morphology and hemodynamic biomarkers. This study proposes a semi-automatic algorithm for aortic centerline extraction applicable to 4D Flow MRI and contrast-enhanced CT images in both healthy individuals and patients with aortic pathologies, including abdominal aortic aneurysms (AAA) and type A and B dissections. The method combines region growing, morphological filtering, skeletonization, and shortest-path computation, and was tested on 40 MRI and 80 CT volumes. Centerlines were successfully extracted in 100% of MRI cases, 95% of CT scans with and 80% without pathology. Centerline length and tortuosity were quantified in both modalities, while pulse wave velocity (PWV) and aortic volume were derived as modality-specific biomarkers. Measurements by two independent observers were compared using Bland-Altman plots and paired t-tests. In CT scans, for centerlines measured within equivalent anatomical regions, dissected aortas were significantly longer (491.8 ± 39.1 mm vs. 453.4 ± 41.7 mm, p < 0.05), more tortuous (0.8 ± 0.2 vs. 0.7 ± 0.1, p < 0.05), and had greater volume (351.6 ± 92.9 ml vs. 197.4 ± 48.3 ml, p < 0.001) than non-pathologic aortas. In 4D Flow MRI, when only AAA cases with valid PWV estimates were analyzed (i.e. R2 > 0.6 in wave transit time vs distance correlations), PWV did not significantly differ from healthy subjects (9.0 ± 3.0 m/s vs. 8.1 ± 2.6 m/s, p = 0.158). Interobserver variability was low across all metrics and modalities. Largest disparities were observed in CT for AAA length (7.2 ± 4.9 mm) and volume (6.5 ± 3.2 ml, p < 0.001), and in type B dissection volume (7.8 ± 7.6 ml, p < 0.01). These results support the algorithm’s robustness and highlight its potential for clinical application.