Motion Compensated 4D Fluoroscopy And Dynamic Roadmaps For Real Time Device Guidance In Minimally Invasive Liver Procedures

University Of Wisconsin-Madison, Madison WI

Investigators

Linked publications & trials

Abstract

ABSTRACT Liver cirrhosis is an extremely costly, morbid condition responsible for over one million deaths per year worldwide. Liver embolization is an established treatment for patients with HCC, where a catheter is guided to specific branches of the visceral or hepatic arteries under fluoroscopic guidance to deliver particles or spheres, which may be loaded with chemotherapeutic drugs or radioactive isotopes. A static digital subtraction angiography (DSA) is usually performed prior to placing the catheter to visualize the hepatic arteries. However, due to respiratory motion, the actual shape and position of the vessels changes constantly during the procedure. This can reduce the accuracy of the device placement and increase procedure times. Aim 1 of this proposal will address this problem with a novel, fluoroscopic device guidance technique which provides motion compensated overlays of vasculature in real-time. A respiratory motion tracking technique will be developed based on curvilinear image features to determine the respiratory state. Additionally, a technique for the creation of vascular motion models based on fluoroscopic image sequences is proposed. This technique estimates the deformation of the vasculature for any fluoroscopic image frame. The second part of this proposal focuses on 3D needle guidance for transjugular intrahepatic portosystemic shunt (TIPS) creation. TIPS is an effective way to treat refractory ascites and decrease risk of variceal bleeding in patients with cirrhosis-related portal hypertension. To create the shunt, a needle is passed from a hepatic vein through the parenchyma and into the portal vein. Typically, the localization of the entry point into the portal venous system is based on either a prior CT/MR acquisition or a wedged portal venogram performed at the time of the procedure. However, in both cases, the target is determined for a single point in time. This makes it difficult to accurately guide the needle to the target point, and in many cases, multiple punctures must be made before the portal venous system is accessed. Aim 2 of this proposal will address this problem with a novel, motion-compensated, 3D imaging system to help guide the needle through the parenchyma to the portal vein. A 3D vascular motion model is proposed based on a static 3D DSA and biplane 2D fluoroscopic sequences. Additionally, a new method for 3D device tracking from biplane fluoroscopy acquisitions is proposed to determine the position and orientation of the needle in real-time. This technique could considerably increase the accuracy of the device guidance for TIPS procedures and therefore reduce the number of required needle passes and procedure time.