A joint project of the Graduate School, Peabody College, and the Jean & Alexander Heard Library

Title page for ETD etd-11212014-132445


Type of Document Dissertation
Author Kumar, Ankur N.
URN etd-11212014-132445
Title Quantifying in vivo motion in video sequences using image registration
Degree PhD
Department Electrical Engineering
Advisory Committee
Advisor Name Title
Dr. Benoit Dawant Committee Chair
Dr. Alan Peters Committee Member
Dr. Bobby Bodenheimer Committee Member
Dr. Dave Piston Committee Member
Dr. Michael Miga Committee Member
Dr. Reid Thompson Committee Member
Keywords
  • stereovision
  • image registration
  • in vivo
  • brain tumor surgery
  • image guided surgery
  • magnification
Date of Defense 2014-11-19
Availability unrestricted
Abstract
Image registration is a pivotal part of many medical imaging analysis systems that provide clinically relevant medical information. One fundamental problem addressed by image registration is the accounting of a subject’s motion. This dissertation broadly addresses the problem of quantifying in vivo motion in video sequences for two different applications using image registration. The first problem involves the correction of motion in in vivo time-series microscopy imaging of islets of Langerhans in mice. The second problem focuses on delivering near real-time 3D intraoperative movements of the cortical surface to a computational biomechanical model framework for the compensation of brain shift during brain tumor surgery. For the first application, a fully automatic algorithm is developed for the correction of in vivo time-series microscopy images of islets of Langerhans. The second application focuses on delivering near real-time 3D intraoperative movements of the cortical surface to a computational biomechanical model framework for the compensation of brain shift during brain tumor surgery. This dissertation demonstrates a clinical microscope-based digitization platform capable of reliably providing temporally dense 3D textured point clouds in near real-time of the FOV for the entire duration and under realistic conditions of neurosurgery. A fully automatic technique has been developed for robustly digitizing 3D points intraoperatively using an operating microscope at 1Hz. Another algorithm has been developed for tracking points on the cortical surface intraoperatively, which can potentially deliver intraoperative 3D displacements of the cortical surface at different time points during brain tumor surgery.
Files
  Filename       Size       Approximate Download Time (Hours:Minutes:Seconds) 
 
 28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access 
  Kumar.pdf 22.15 Mb 01:42:32 00:52:44 00:46:08 00:23:04 00:01:58

Browse All Available ETDs by ( Author | Department )

If you have more questions or technical problems, please Contact LITS.