Objective To propose an innovative self-supervised learning method for vascular segmentation in computed tomography angiography (CTA) images by integrating feature reconstruction with masked autoencoding. Methods A 3D masked autoencoder-based framework is developed, where in 3D histogram of oriented gradients (HOG) is utilized for multi-scale vascular feature extraction. During pre-training, random masking is applied to local patches of CTA images, and the model is trained to jointly reconstruct original voxels and HOG features of masked regions. The pre-trained model is further fine-tuned on two annotated datasets for clinical-level vessel segmentation. Results Evaluated on two independent datasets (30 labeled CTA images each), our method achieves superior segmentation accuracy to the supervised neural network U-Net (nnU-Net) baseline, with dice similarity coefficients (DSC) of 91.2% vs. 89.7% (aorta) and 84.8% vs. 83.2% (coronary arteries). Conclusion The proposed self-supervised model significantly reduces manual annotation costs without compromising segmentation precision, showing substantial potential for enhancing clinical workflows in vascular disease management.