Three-dimensional A-mode ultrasound calibration and registration for robotic orthopaedic knee surgery

Alon Mozes, Ta Cheng Chang, Louis Arata, Weizhao Zhao

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Background: Registration is a key step for computer-navigated robot-assisted surgery. Registration links the live patient anatomical location to the prescanned CT or MRI images, so that predesigned procedures can be performed accurately. Fiducial markers or mechanical probes are usually used to identify anatomical features or collect data points for registration. This conventional invasive approach is common; however, using ultrasound probes may provide a non-invasive alternative. Methods: This report presents investigations of selecting an A-mode ultrasound transducer, calibrating it, analysing the ultrasound signal and using it to register phantom-sawbones of tibia and femur as well as cadaveric specimens. To ensure accurate registration, the A-mode ultrasound probe is calibrated by a designed calibration system. Detailed mathematical derivation and procedures for the calibration are provided in the Appendix. The calibration and registration experiments were performed in conjunction with MAKO Surgical Corporation's Tactile Guidance System™ (TGS™) at their headquarters and at the South Florida Spine Clinic for cadaveric experiments. Results: Calibration results show that an A-mode ultrasound probe can reach the same accuracy level as a mechanical probe. By using the A-mode ultrasound probe, averaged root mean square errors (RMSE) are <0.5 mm for calibration, <1.0 mm for phantom-sawbones and <2.0 mm for cadaveric specimens. Conclusion: The registration results from phantom and cadaveric experiments are suitable for clinical applications. A-mode ultrasound registration is a viable option for registration of the bones in orthopaedic knee surgery but with reduced incision size.

Original languageEnglish
Pages (from-to)91-101
Number of pages11
JournalInternational Journal of Medical Robotics and Computer Assisted Surgery
Volume6
Issue number1
DOIs
StatePublished - Mar 1 2010

Fingerprint

Orthopedics
Robotics
Surgery
Calibration
Knee
Ultrasonics
Cadaveric experiments
Fiducial Markers
Touch
Transducers
Tibia
Femur
Spine
Mean square error
Magnetic resonance imaging
Bone and Bones
Bone
Robots
Industry

Keywords

  • 3D registration
  • A-mode ultrasound
  • Knee
  • Robotic surgery

ASJC Scopus subject areas

  • Computer Science Applications
  • Biophysics
  • Surgery

Cite this

@article{205cb73a346140479e2b1c9bf9227bee,
title = "Three-dimensional A-mode ultrasound calibration and registration for robotic orthopaedic knee surgery",
abstract = "Background: Registration is a key step for computer-navigated robot-assisted surgery. Registration links the live patient anatomical location to the prescanned CT or MRI images, so that predesigned procedures can be performed accurately. Fiducial markers or mechanical probes are usually used to identify anatomical features or collect data points for registration. This conventional invasive approach is common; however, using ultrasound probes may provide a non-invasive alternative. Methods: This report presents investigations of selecting an A-mode ultrasound transducer, calibrating it, analysing the ultrasound signal and using it to register phantom-sawbones of tibia and femur as well as cadaveric specimens. To ensure accurate registration, the A-mode ultrasound probe is calibrated by a designed calibration system. Detailed mathematical derivation and procedures for the calibration are provided in the Appendix. The calibration and registration experiments were performed in conjunction with MAKO Surgical Corporation's Tactile Guidance System™ (TGS™) at their headquarters and at the South Florida Spine Clinic for cadaveric experiments. Results: Calibration results show that an A-mode ultrasound probe can reach the same accuracy level as a mechanical probe. By using the A-mode ultrasound probe, averaged root mean square errors (RMSE) are <0.5 mm for calibration, <1.0 mm for phantom-sawbones and <2.0 mm for cadaveric specimens. Conclusion: The registration results from phantom and cadaveric experiments are suitable for clinical applications. A-mode ultrasound registration is a viable option for registration of the bones in orthopaedic knee surgery but with reduced incision size.",
keywords = "3D registration, A-mode ultrasound, Knee, Robotic surgery",
author = "Alon Mozes and Chang, {Ta Cheng} and Louis Arata and Weizhao Zhao",
year = "2010",
month = "3",
day = "1",
doi = "10.1002/rcs.294",
language = "English",
volume = "6",
pages = "91--101",
journal = "International Journal of Medical Robotics and Computer Assisted Surgery",
issn = "1478-596X",
publisher = "John Wiley and Sons Ltd",
number = "1",

}

TY - JOUR

T1 - Three-dimensional A-mode ultrasound calibration and registration for robotic orthopaedic knee surgery

AU - Mozes, Alon

AU - Chang, Ta Cheng

AU - Arata, Louis

AU - Zhao, Weizhao

PY - 2010/3/1

Y1 - 2010/3/1

N2 - Background: Registration is a key step for computer-navigated robot-assisted surgery. Registration links the live patient anatomical location to the prescanned CT or MRI images, so that predesigned procedures can be performed accurately. Fiducial markers or mechanical probes are usually used to identify anatomical features or collect data points for registration. This conventional invasive approach is common; however, using ultrasound probes may provide a non-invasive alternative. Methods: This report presents investigations of selecting an A-mode ultrasound transducer, calibrating it, analysing the ultrasound signal and using it to register phantom-sawbones of tibia and femur as well as cadaveric specimens. To ensure accurate registration, the A-mode ultrasound probe is calibrated by a designed calibration system. Detailed mathematical derivation and procedures for the calibration are provided in the Appendix. The calibration and registration experiments were performed in conjunction with MAKO Surgical Corporation's Tactile Guidance System™ (TGS™) at their headquarters and at the South Florida Spine Clinic for cadaveric experiments. Results: Calibration results show that an A-mode ultrasound probe can reach the same accuracy level as a mechanical probe. By using the A-mode ultrasound probe, averaged root mean square errors (RMSE) are <0.5 mm for calibration, <1.0 mm for phantom-sawbones and <2.0 mm for cadaveric specimens. Conclusion: The registration results from phantom and cadaveric experiments are suitable for clinical applications. A-mode ultrasound registration is a viable option for registration of the bones in orthopaedic knee surgery but with reduced incision size.

AB - Background: Registration is a key step for computer-navigated robot-assisted surgery. Registration links the live patient anatomical location to the prescanned CT or MRI images, so that predesigned procedures can be performed accurately. Fiducial markers or mechanical probes are usually used to identify anatomical features or collect data points for registration. This conventional invasive approach is common; however, using ultrasound probes may provide a non-invasive alternative. Methods: This report presents investigations of selecting an A-mode ultrasound transducer, calibrating it, analysing the ultrasound signal and using it to register phantom-sawbones of tibia and femur as well as cadaveric specimens. To ensure accurate registration, the A-mode ultrasound probe is calibrated by a designed calibration system. Detailed mathematical derivation and procedures for the calibration are provided in the Appendix. The calibration and registration experiments were performed in conjunction with MAKO Surgical Corporation's Tactile Guidance System™ (TGS™) at their headquarters and at the South Florida Spine Clinic for cadaveric experiments. Results: Calibration results show that an A-mode ultrasound probe can reach the same accuracy level as a mechanical probe. By using the A-mode ultrasound probe, averaged root mean square errors (RMSE) are <0.5 mm for calibration, <1.0 mm for phantom-sawbones and <2.0 mm for cadaveric specimens. Conclusion: The registration results from phantom and cadaveric experiments are suitable for clinical applications. A-mode ultrasound registration is a viable option for registration of the bones in orthopaedic knee surgery but with reduced incision size.

KW - 3D registration

KW - A-mode ultrasound

KW - Knee

KW - Robotic surgery

UR - http://www.scopus.com/inward/record.url?scp=77149156625&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77149156625&partnerID=8YFLogxK

U2 - 10.1002/rcs.294

DO - 10.1002/rcs.294

M3 - Article

C2 - 20014154

AN - SCOPUS:77149156625

VL - 6

SP - 91

EP - 101

JO - International Journal of Medical Robotics and Computer Assisted Surgery

JF - International Journal of Medical Robotics and Computer Assisted Surgery

SN - 1478-596X

IS - 1

ER -