Comparing Autoencoder to Geometrical Features for Vascular Bifurcations Identification

The cerebrovascular tree is a complex anatomical structure that plays a crucial role in the brain irrigation. A precise identification of the bifurcations in the vascular network is essential for understanding various cerebral pathologies. Traditional methods often require manual intervention and are sensitive to variations in data quality. In recent years, deep learning techniques, and particularly autoencoders, have shown promising performances for feature extraction and pattern recognition in a variety of domains. In this paper, we propose two novel approaches for vascular bifurcation identification based respectiveley on Autoencoder and geometrical features. The performance and effectiveness of each method in terms of classification of vascular bifurcations using medical imaging data is presented. The evaluation was performed on a sample database composed of 91 TOF-MRA, using various evaluation measures, including accuracy, F1 score and confusion matrix.