The transjugular intrahepatic portosystemic shunt (TIPS) procedure is one of the most technically challenging procedures in interventional radiology. During the procedure, interventional radiologists (IRs) insert very thin and long instruments through a little incision in the patient’s neck. They then navigate the instruments through the blood vessels towards the liver to create a shunt in the liver. The most difficult step of the procedure is to puncture through the liver into to the portal vein, often referred to as the intrahepatic puncture. The difficulty of the puncture lies in the limited availability of real-time three-dimensional (3D) information about the anatomy and instruments during the puncture. IRs do get information to navigate, for example from X-ray imaging modalities, but insufficient. A promising technique to provide IRs the required information is by using three-dimensional ultrasound (3D US). However, the user interfaces (UI) of current 3D US machines are complicated and do not meet IRs’ expectations. The goal of this research is to design a real-time 3D US based UI for IRs to minimize the number of punctures during the TIPS procedure. To do so, the main objective of this thesis was to discover what information should be presented in an interactive 3D US based UI to minimize the number of intrahepatic punctures in the TIPS procedure. By using co-design research, the author was able to create an overview of the TIPS procedure, identify the related challenges and UI needs, and investigate the possibilities of using 3D US technology to improve the procedure. She found that IRs mainly miss information to perform the elementary cognitive actions and have limited possibility to create and maintain proper situation awareness, to predict the consequence of actions, and to learn from previous attempts. Based on the insights, two 3D US based UIs were designed to plan and perform the intrahepatic puncture, respectively. The UIs were tested with IRs and the results showed that both UIs provided the participating IRs with 3D information, feedback and navigation support in a more effective and efficient manner.