Closed Loop Image Guided Intervention

James Williams from Siemens Research

There are three primary drivers of innovation in image-guided intervention: Can we do it better? Can it be done faster? Can we do something that could not be done before?

The technologies that make this happen are primarily software. Image segmentation, registration and visualization are well established in the diagnostic domain, but making the jump into the interventional world is to change from offline to online formulations of these problems. Of particular concern are the speed and robustness of registration algorithms. In order to maintain an accurate mapping between pre-operative and inter-operative frames of reference, registration and imaging must be engaged in a frequently updated closed-loop. In such an environment, the boundaries between procedure planning and interactive control of imaging devices and tools can become blurred in a beneficial way.

The talk will first discuss the experience of creating a complete image-based guidance system for the cardiac electrophysiology (EP) lab which incorporates registration, segmentation and visualization algorithms designed specifically for this application domain. Subsequently, similar closed-loop systems under development for the MR interventional environment and radiation therapy suite will be described.