1 / 8

A Computationally Efficient Approach for 2D-3D Image Registration

A Computationally Efficient Approach for 2D-3D Image Registration. Juri Minxha Medical Image Analysis Professor Benjamin Kimia Spring 2011. Brown University. Brown University. Problem Statement. 2 Signal Sources 3D CT volume, 2D images (fluoroscopy

marilu
Download Presentation

A Computationally Efficient Approach for 2D-3D Image Registration

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. A Computationally Efficient Approach for 2D-3D Image Registration Juri Minxha Medical Image Analysis Professor Benjamin Kimia Spring 2011 Brown University

  2. Brown University Problem Statement • 2 Signal Sources 3D CT volume, 2D images (fluoroscopy • Project 3D data onto a 2D plane and compare it to existing 2D image. The projected image is also known as the digitally reconstructed radiograph

  3. Brown University Background General Approach Approach outlined in this paper Plan of Action

  4. Brown University Background • Image guided surgery (on spinal cord for example) • Pre-operative data (CT/MRI acquisitions) • Good resolution • 3D data • Slow Acquisition • Intra-operative data (fluoroscopy images) • Can be quickly acquired • Poor resolution, more noise

  5. Brown University Typical Approach to Registration 1. A geometric transform is applied on the 3D CT volume to find the warped position of the bone. 2. The warped volume is projected onto a 2D plane to produce the 2D DRR. 3. Both the DRR and 2D fluoroscopy frame are then filtered with a Laplacian-of-Gaussian (LoG) filter. 4. The similarity between the DRR and the fluoroscopy frame is then measured. 5. An update vector for the geometric transform is then found. 6. The algorithm is repeated with the updated transform parameters. The associated challenges are: Different dimensionalities, Minimize computation time to fit operation

  6. Brown University Approach Outlined in this paper Pre-compute: 1. Construct the 2D DRR image of the original CT volume. 2. Estimate the value of zˆi. Iterate the following steps: 1. For the start of each level and for the last 5 iterations apply a 3D transform on the CT volume and construct a DRR image from the updated volume. Otherwise apply a 2D transform on the DRR image. 2. Apply (LoG) filter to both the DRR and Fluoroscopy frames to highlight the edges of the object. 3. Update the transformation parameters using the SCV similarity measure [1]. 4. Repeat the algorithm with the updated parameters.

  7. Brown University Plan of Action Weeks 1: Problem definition Build data set Week 2-4: Implement the general approach and fast approach in Matlab Week 5: testing on data, analyze results , compare general approach with approach outlined in the paper see if my performance results match what they claim in the paper

  8. Brown University References • A computationally efficient approach for 2D-3D image registration Haque, M.N.; Pickering, M.R.; Biswas, M.; Frater, M.R.; Scarvell, J.M.; Smith, P.N.; 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Issue Date: Aug. 31 2010-Sept. 4 2010, On page(s): 6268 – 6271

More Related