Automatic 3D Vascular Tree Construction of Perforator Flaps for Plastic Surgery Planning

Perforator flaps have been increasingly used in the past few years for trauma and reconstructive surgical cases. With the thinned flap design, greater survivability and a decrease in donor site morbidity have been reported. Knowledge of the 3D vascular tree will provide insight information about the dissection region, vascular territory, and fascia levels. In this paper, we will propose a computational framework for the automatic 3D vascular tree construction. The computational framework begins with an image segmentation algorithm, spedge-and-medge, which is an integration of Canny edge detector, edge-linking, and split-and-merge to initially segment out the vessels from the background. To deal with the possible broken vessels, a vascular cross-sectional tree repairing and interpolation algorithm is then developed based on the 3D connectivity and root-converging properties of the tree branches. Furthermore, to extract the essential characteristics of the vascular structure, 3D thinning algorithms are used to build up the skeletons of the tree. At each stage of the framework, 3D rendering results are provided for the visualization of the computed results. The proposed method achieves good performance and has been used for the 3D vascular tree construction and surgical danger zone measurements on 39 harvested cadaver perforator flaps with the types of ALTP, GAP, and TAP.