An OpenSource Solution for Interactive Acquisition, Processing and Transfer of Interventional Ultras

1
An Open-Source Solution for Interactive
Acquisition, Processing and Transfer of
Interventional Ultrasound Images

• Jonathan Boisvert, David Gobbi, Siddharth Vikal,
  Robert Rohling, Gabor Fichtinger and Purang
  Abolmaesumi
• Medical Image Analysis (MedIA) Laboratory
• http//www.media.queensu.ca
• Queens University
• Kingston, Ontario, Canada

2
Introduction

• Ultrasound is a very common interventional image
  modality
• Open-interface machines makes it even more
  attractive to IGT developers
• Often used with motion tracking devices.
• Problem
• Acquisition, synchronization and transfer to
  existing IGT systems is more difficult than it
  should be.

3
Prior work

• Integrated navigation software such as
• CustusX1, IGSonic2, SonoNav3 or Stradx4.
• Software frameworks such as
• IGSTK5
• Slicer specific software development
• A module for Slicer 2 initially developed by
  Boctor et al.6

4
Open-source software platform Slicer

• Sophisticated registration, data-fusion and
  visualization capabilities
• Modular architecture
• Multi-platform
• Support for the new OpenIGTLink protocol

5
Objectives

• Ultrasound Image Acquisition
• 3D Tracking
• Synchronization
• Ultrasound Volumes Reconstruction
• Open-Source
• Standardized communication with independent IGT
  platforms.

6
Architecture SynchroGrab

• Provide real-time ultrasound processing
  capabilities.
• Delegate visualization, registration and
  navigation to an OpenIGTLink compatible software
  (Slicer).

7
Architecture SynchroGrab

• Provide real-time ultrasound processing
  capabilities.
• Delegate visualization, registration and
  navigation to an OpenIGTLink compatible software
  (Slicer).

8
Architecture SynchroGrab

• Provide real-time ultrasound processing
  capabilities.
• Delegate visualization, registration and
  navigation to an OpenIGTLink compatible software
  (Slicer).

9
Architecture SynchroGrab

• Provide real-time ultrasound processing
  capabilities.
• Delegate visualization, registration and
  navigation to an OpenIGTLink compatible software
  (Slicer).

10
Architecture SynchroGrab

• Provide real-time ultrasound processing
  capabilities.
• Delegate visualization, registration and
  navigation to an OpenIGTLink compatible software
  (Slicer).

11
Ultrasound image acquisition

• SonixRP
• Open interface
• Access to pre-scan converted images, post-scan
  converted images, imaging parameters, RF
  raw-data, etc.
• New vtk class vtkSonixVideoSource
• Easy integration
• Highly reusable

12
Using vtkSonixVideoSource
vtkSonixVideoSource sonixGrabber
vtkSonixVideoSourceNew() sonixGrabber-gtSetSonix
IP("127.0.0.1") sonixGrabber-gtSetImagingMode(BMod
e) sonixGrabber-gt SetAcquisitionDataType(udtBP
ost) sonixGrabber-gtRecord() sonixGrabber-gtStop()
sonixGrabber-gtRewind() // Process the first
image here sonixGrabber-gtSeek(1) // Process the
second image here sonixGrabber-gtDelete()

• Inherit from vtkVideoSource
• Support many data types
• BPreScan, BPost 8 bit, BPost 32 bit, RF 16 bit,
  MPost 8 bit, Color B 32 bits

13
3D Tracking

• Object oriented
• Extensible
• Multi-threaded
• Circular buffers for transformations
  synchronization

14
Synchronization

• Tracking and ultrasound imaging
• not performed at the same frequency.
• Need to interpolate the poses (using spherical
  linear interpolation)

15
Synchronization

• Data preprocessing
• Hardware latency, pose computation, ultrasound
  image processing, and copy to user-space memory.
• OS related
• Context switching, memory swapping, etc.
• Compensate for the mean lag

16
Synchronization

• vtkTaggedImageFilter
• Compute the transformation associated with an
  image.
• Two inputs a vtkTrackerTool and a
  vtkImageAlgorithm object

vtkTaggedImageFilter tagger
vtkTaggedImageFilterNew() tagger-gtSetInput(son
ixGrabber-gtGetOutput()) tagger-gtSetTrackerTool(tr
ackerTool) tagger-gtSetTemporalLag(xxx) tagger-gt
Update() currentTransforms tagger-gtGetTransfor
m() currentImage tagger-gtGetOutput()
17
3D Reconstruction

• Synchronized images and transformation
• Combined to produce 3D volumes
• Pixel-based reconstruction
• Pixel nearest neighbors (PNN)
• Pixel trilinear interpolation (PTL)

18
Collecting images

• Just start the application using the command line

SynchroGrab --calibration-file 071508_162653.Calib
ration.results --save-images --use-tracker-transf
orms --nb-frames 150 --fps 5
Saving usImage0.png Current rigid transform
0.0933926 0.982412 0.152506 -394.844 0.976081
-0.0744842 -0.208865 296.167 -0.196327 0.171225
-0.96587 -1923.87 0 0 0 1 Saving
usImage1.png Current rigid transform 0.0934317
0.982387 0.152642 -394.833 0.976065 -0.0744826
-0.208939 296.161 -0.196386 0.171371 -0.965833
-1923.89 0 0 0 1 Saving usImage2.png Current
rigid transform 0.093352 0.982413 0.152529
-394.84 0.976048 -0.0744086 -0.209039 296.141
-0.196509 0.171252 -0.965829 -1923.86 0 0 0 1 .
19
Reconstructing volumes

• Just start the application using the command line

SynchroGrab --calibration-file 071508_162653.Calib
ration.results --reconstruct-volume --nb-frames
150 --fps 5 --use-tracker-transforms
20
Communication with Slicer

• Real-time acquisition using SynchroGrab.
• Visualization with Slicer 3.

SynchroGrab --calibration-file 071508_162653.Calib
ration.results --oigtl-transfert-images
--nb-frames 150 --fps 5
21
Communication with Slicer
22
Conclusion Future work

• SynchroGrab
• Acquisition and synchronization of tracking and
  ultrasound data
• Can be used as a stand-alone command-line
  application.
• Can be used in conjunction to an OpenIGTLink
  compliant software.
• Development of temporal calibration protocols
• Future support for 3D probes

23
References

• T. Lango, G. A. Tangen, R. Marvik, B. Ystgaard,
  Y. Yavuz, J. H. Kaspersen, O. V. Solberg, and T.
  A. N. Hernes. Navigation in laparoscopyprototype
  research platform for improved image-guided
  surgery. Minim Invasive Ther Allied Technol,
  17(1)1733, 2008.
• BrainLAB (AG, Heimstetten, Germany). IGSonic.
  http//www.brainlab.com
• Medtronic(Louisville, CO, U.S.). SonoNav.
  http//www.medtronicnavigation.com
• Graham Treece, Richard Prager and Andrew Gee.
  Stradx/Stradwin. http//mi.eng.cam.ac.uk/rwp/st
  radwin/
• Andinet Enquobahrie, Patrick Cheng, Kevin Gary,
  Luis Ibanez, David Gobbi, Frank Lindseth, Ziv
  Yaniv, Stephen Aylward, Julien Jomier, and Kevin
  Cleary. The image-guided surgery toolkit IGSTK
  an open source C software toolkit. J Digit
  Imaging, 20 Suppl 12133, 2007.
• E. M. Boctor, A. Viswanathan, S. Pieper, M. A.
  Choti, R. H. Taylor, R. Kikinis, and G.
  Fichtinger. CISUS an integrated 3D ultrasound
  system for IGT using a modular tracking API.
  Proceedings of the SPIE, volume 5367, pages
  247256, May 2004

24
Acknowledgments

• Natural Sciences and Engineering Research Council
  (NSERC),
• Canadian Institutes of Health Research (CIHR),
• National Alliance in Medical Image Computing
  (NIH5U54EB005149-03),
• Fonds québécois de recherche sur la nature et les
  technologies.

25
Questions