Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Purpose Liver cancer is the 5 most common cancer and shows poor prognosis. Surgical removal of liver tumours, the only existing curative treatment, can merely be used in 10-20% of the case. Increasing surgical precision is a key-challenge to give more patients access to a potentially curative treatment. Recent progress in computer science enables the use of instrument guidance systems for open liver surgery by providing improved orientation and guidance support during planning and intraoperative realization. However, challenge remains when precise alignment between preoperative image data and the intraoperative situation is required, since the liver is subject to deformation and movements during the surgical treatment. The CASOne liver navigation system (CAScination AG, Switzerland) applies a landmark based registration technic to perform the alignment. Major drawbacks of this technic reside in the difficulties of identifying accurately correspondences between the preoperative image data and the intraoperative situation. In a recent study, including more than 50 surgeries performed with the CALS system, the authors measure a median alignment precision of 6.3 mm. We present a framework to improve such alignment using intraoperative ultrasound imaging (US) and preoperative computed tomography (MeVis-CT) data.
Kristina Schoonjans, Petar Petrov