Preliminary results for the supervised detection of lumen and stent from OCT pullbacks

In cardiology, optical coherence tomography (OCT) has been established as a reference for in vivo intravascular image. OCT provides high-resolution and enables us to visualize accurately vessel intimal layers and stents. The drawback is the time necessary to manually select stent struts and lumen boundary. We developed and evaluated an automatic stent strut detection method to apply to intravascular OCT images. The proposed approach innovates the strut shadow detection, which can be considered as an independent processing step that helps discriminating struts by using a profile map. The paper will be presented as a complete process, explaining the necessary steps in order of appearance.