Project Proposal: Surfacetosurface tracking of breast models in radiation oncology

1
Project ProposalSurface-to-surface tracking of
breast models in radiation oncology

• Shidong Li, Russell H. TaylorRobert Jacques,
  John Rittenhouse

2
Overview

• The Problem
• Surface Tracking Methods
• FEM Models of the Breast
• Project Expectations
• Dependencies
• Reading List

3
The Problem

• The Procedure
• Equipment
• Procedure Questions
• Project Questions

4
The Procedure

• Radiation beam therapy of breast tumours
• Treatment is gated to respiration
• Single marker
• Trade off between treatment time and accuracy
• Large safety boundary 15mm

5
Equipment

• Stereo Camera pair with structured light
• Currently 0.1Hz maximum
• 1Hz capture
• Co-registered with CT
• Rigid ICP registration

6
Procedure Questions

• Optimum gate length
• Save OR time
• Reduction of safety boundary
• Less healthy tissue damaged
• Correlation between skin and tumour locations
• Active collimation

7
Project Questions

• Investigation of surface breast motion
• Noise/Random motion
• Respiration
• Evaluate correlation with current metric
• Coupling with FEM analysis

8
Surface Tracking Methods

• 2D Methods
• Template tracking
• NURB tracking package
• 3D Methods
• Active surfaces
• ICP vs. GVF
• Free vs. Continuous
• Elastance
• Surface area constraint

9
FEM Models of the Breast

• Current uses
• Compression studies for biopsy
• Image processing validation
• Accuracy of 2.5-4.5mm
• Sources of error
• Soft tissue segmentation in CT
• Boundary conditions
• Skin
• Chest Wall

10
Project Expectations

• Minimum
• Medium
• High
• Maximum

11
Minimum Project Expectations

• Deliverables
• Basic offline deformable surface tracker
• Analysis of tracker using SSD
• Limitations
• Direct analysis of mesh modes not possible
• Difficult to couple to a FEM model
• Not real time
• User supervision

12
Medium Project Expectations

• Deliverables
• Offline deformable elastic surface tracker
• Analysis of tracker using volume between surfaces
• Generation of patient specific atlas of
  respiration
• Plot of modes versus time
• Limitations
• Not real time
• User supervision

13
High Project Expectations

• Deliverables
• Medium Project Expectations
• Simple user interface for clinical use
• Per frame processing equal to capture device
• Creation and analysis of modes of variation
• Analysis of the modes of a new frame
• Limitations
• Not real time

14
Maximum Project Expectations

• Deliverables
• High Project Expectations in real time
• Near-fully automatic
• Determination of optimum treatment times
• Coupling to a FEM model

15
Dates for Medium Expectations

• February 18, 2005
• Robert Tracking of fiducial points
• John Basic GUI for visualization
• February 25, 2005
• Robert ICP based elastic mesh deformation
• John Volumetric accuracy metric
• March 4, 2005
• Robert Generation of a statistical atlas
• John Mode visualization

16
Project Checkpoint (Mid-March)

• Determine current status
• Identify bottlenecks
• Performance
• Developmental issues
• Update project plan
• Project extensions
• New requirements
• Final goals

17
Plan for Project Extensions

• Time and complexity analysis
• Mesh simplification
• Direct stereo methods
• Euler vs. PDE solvers
• Mode Analysis
• Determination of clinical user interface
• Automatic fiducial search
• FEM coupling and analysis

18
Dependencies

• Developmental
• Standard data format
• Image sequence at frame rate of consecutive
  respiration cycles
• Raw Stereo data
• Color data sets
• Implementation
• Clinical trial coupling
• 3rd party image acquisition at 0.1Hz

19
Reading list

• Julia A. Schnabel, Christine Tanner, Andy D.
  Castellano-Smith, Andreas Degenhard, Martin O.
  Leach, D. Rodney Hose, Derek L.G. Hill, and David
  J. Hawkes. Validation of Nonrigid Image
  Registration Using Finite-Element Methods
  Application to Breat MR Images. IEEE Transactions
  on Medical Imaging, Vol. 22, No. 2, February 2003
• David J. Hawks, and Derek L. G. Hill. Medical
  Imaging at Guys Hospital, Kings College London.
  IEE Transactions on Medical Imaging, Vol. 22, No.
  9, September 2003

20
Reading List Cont.

• Kass, Witkin, and Terzopoulos. Snakes Active
  Contour Models. 1987
• C. Xu, and J.L. Prince. Gradient Vector Flow A
  New External Force for Snakes. Proc. IEEE Conf.
  on Comp. Vis. Patt. Recog. (CVPR), Los Alamitos
  Comp. Soc. Press, pp. 66-71, June 1997
• Jason J. Corso, Nicholas Ramey, and Gregory D.
  Hager. Stereo-Based Direct Surface Tracking with
  Deformable Parametric Models. Technical report,
  The Johns Hopkins University, 2003. CIRL Lab
  Technical Report 2003-02.