Real-time image-features detection and tracking in non-injected fluoroscopic images
X-ray fluoroscopic images are widely used as guidance in minimally invasive endovascular procedures.
They allow the clinician to visualize in real-time both the anatomy and the surgical instrument navigating within the blood vessels.
Nevertheless, when a contrast medium is not injected, these grey-scale fluoroscopic images are difficult to be interpreted, given the lack of depth perception and the overlaying anatomical structures with similar contrast.
The aim of this internship is to implement a calibration technique for X-ray imaging system combined with a method for real-time image-features detection and tracking in non-injected fluoroscopic images.
An additional validation study of the designed method will be carried out, in order to quantify the errors associated with the procedure.