Intraoperative segmentation and nonrigid registration for image guided therapy

Abstract

Our goal was to improve image guidance during minimally invasive image guided therapy by developing an intraoperative segmentation and nonrigid registration algorithm. The algorithm was designed to allow for improved navigation and quantitative monitoring of treatment progress in order to reduce the time required in the operating room and to improve outcomes. The algorithm has been applied to intraoperative images from cryotherapy of the liver and from surgery of the brain. Empirically the algorithm has been found to be robust with respect to imaging characteristics such as noise and intensity inhomogeneity and robust with respect to parameter selection. Serial and parallel implementations of the algorithm are sufficiently fast to be practical in the operating room. The contributions of this work are an algorithm for intraoperative segmentation and intraoperative registration, a method for quantitative monitoring of cryotherapy from real-time imaging, quantitative monitoring of brain tumor resection by comparison to a preoperative treatment plan and an extensive validation study assessing the reproducibility of the intraoperative segmentation. We have evaluated our algorithm with six neurosurgical cases and two liver cryotherapy cases with promising results. Further clinical validation with larger numbers of cases will be necessary to determine if our algorithm succeeds in improving intraoperative navigation and intraoperative therapy delivery and hence improves therapy outcomes.