Ultrasound probe and needle-guide calibration for robotic ultrasound scanning and needle targeting

Chunwoo Kim, Doyoung Chang, Doru Petrisor, Gregory Chirikjian, Misop Han, Dan Stoianovici

Research output: Contribution to journalArticle

Abstract

Image-to-robot registration is a typical step for robotic image-guided interventions. If the imaging device uses a portable imaging probe that is held by a robot, this registration is constant and has been commonly named probe calibration. The same applies to probes tracked by a position measurement device. We report a calibration method for 2-D ultrasound probes using robotic manipulation and a planar calibration rig. Moreover, a needle guide that is attached to the probe is also calibrated for ultrasound-guided needle targeting. The method is applied to a transrectal ultrasound (TRUS) probe for robot-assisted prostate biopsy. Validation experiments include TRUS-guided needle targeting accuracy tests. This paper outlines the entire process from the calibration to image-guided targeting. Freehand TRUS-guided prostate biopsy is the primary method of diagnosing prostate cancer, with over 1.2 million procedures performed annually in the U.S. alone. However, freehand biopsy is a highly challenging procedure with subjective quality control. As such, biopsy devices are emerging to assist the physician. Here, we present a method that uses robotic TRUS manipulation. A 2-D TRUS probe is supported by a 4-degree-of-freedom robot. The robot performs ultrasound scanning, enabling 3-D reconstructions. Based on the images, the robot orients a needle guide on target for biopsy. The biopsy is acquired manually through the guide. In vitro tests showed that the 3-D images were geometrically accurate, and an image-based needle targeting accuracy was 1.55 mm. These validate the probe calibration presented and the overall robotic system for needle targeting. Targeting accuracy is sufficient for targeting small, clinically significant prostatic cancer lesions, but actual in vivo targeting will include additional error components that will have to be determined.

Original languageEnglish (US)
Article number6416032
Pages (from-to)1728-1734
Number of pages7
JournalIEEE Transactions on Biomedical Engineering
Volume60
Issue number6
DOIs
StatePublished - 2013

Fingerprint

Needles
Biopsy
Robotics
Ultrasonics
Calibration
Scanning
Robots
Imaging techniques
Position measurement
Quality control

Keywords

  • Image-guided robot
  • needle-guide calibration
  • prostate biopsy
  • registration
  • ultrasound calibration

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

Ultrasound probe and needle-guide calibration for robotic ultrasound scanning and needle targeting. / Kim, Chunwoo; Chang, Doyoung; Petrisor, Doru; Chirikjian, Gregory; Han, Misop; Stoianovici, Dan.

In: IEEE Transactions on Biomedical Engineering, Vol. 60, No. 6, 6416032, 2013, p. 1728-1734.

Research output: Contribution to journalArticle

@article{8c60535750c94291ab16f2eb6615f119,
title = "Ultrasound probe and needle-guide calibration for robotic ultrasound scanning and needle targeting",
abstract = "Image-to-robot registration is a typical step for robotic image-guided interventions. If the imaging device uses a portable imaging probe that is held by a robot, this registration is constant and has been commonly named probe calibration. The same applies to probes tracked by a position measurement device. We report a calibration method for 2-D ultrasound probes using robotic manipulation and a planar calibration rig. Moreover, a needle guide that is attached to the probe is also calibrated for ultrasound-guided needle targeting. The method is applied to a transrectal ultrasound (TRUS) probe for robot-assisted prostate biopsy. Validation experiments include TRUS-guided needle targeting accuracy tests. This paper outlines the entire process from the calibration to image-guided targeting. Freehand TRUS-guided prostate biopsy is the primary method of diagnosing prostate cancer, with over 1.2 million procedures performed annually in the U.S. alone. However, freehand biopsy is a highly challenging procedure with subjective quality control. As such, biopsy devices are emerging to assist the physician. Here, we present a method that uses robotic TRUS manipulation. A 2-D TRUS probe is supported by a 4-degree-of-freedom robot. The robot performs ultrasound scanning, enabling 3-D reconstructions. Based on the images, the robot orients a needle guide on target for biopsy. The biopsy is acquired manually through the guide. In vitro tests showed that the 3-D images were geometrically accurate, and an image-based needle targeting accuracy was 1.55 mm. These validate the probe calibration presented and the overall robotic system for needle targeting. Targeting accuracy is sufficient for targeting small, clinically significant prostatic cancer lesions, but actual in vivo targeting will include additional error components that will have to be determined.",
keywords = "Image-guided robot, needle-guide calibration, prostate biopsy, registration, ultrasound calibration",
author = "Chunwoo Kim and Doyoung Chang and Doru Petrisor and Gregory Chirikjian and Misop Han and Dan Stoianovici",
year = "2013",
doi = "10.1109/TBME.2013.2241430",
language = "English (US)",
volume = "60",
pages = "1728--1734",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "6",

}

TY - JOUR

T1 - Ultrasound probe and needle-guide calibration for robotic ultrasound scanning and needle targeting

AU - Kim, Chunwoo

AU - Chang, Doyoung

AU - Petrisor, Doru

AU - Chirikjian, Gregory

AU - Han, Misop

AU - Stoianovici, Dan

PY - 2013

Y1 - 2013

N2 - Image-to-robot registration is a typical step for robotic image-guided interventions. If the imaging device uses a portable imaging probe that is held by a robot, this registration is constant and has been commonly named probe calibration. The same applies to probes tracked by a position measurement device. We report a calibration method for 2-D ultrasound probes using robotic manipulation and a planar calibration rig. Moreover, a needle guide that is attached to the probe is also calibrated for ultrasound-guided needle targeting. The method is applied to a transrectal ultrasound (TRUS) probe for robot-assisted prostate biopsy. Validation experiments include TRUS-guided needle targeting accuracy tests. This paper outlines the entire process from the calibration to image-guided targeting. Freehand TRUS-guided prostate biopsy is the primary method of diagnosing prostate cancer, with over 1.2 million procedures performed annually in the U.S. alone. However, freehand biopsy is a highly challenging procedure with subjective quality control. As such, biopsy devices are emerging to assist the physician. Here, we present a method that uses robotic TRUS manipulation. A 2-D TRUS probe is supported by a 4-degree-of-freedom robot. The robot performs ultrasound scanning, enabling 3-D reconstructions. Based on the images, the robot orients a needle guide on target for biopsy. The biopsy is acquired manually through the guide. In vitro tests showed that the 3-D images were geometrically accurate, and an image-based needle targeting accuracy was 1.55 mm. These validate the probe calibration presented and the overall robotic system for needle targeting. Targeting accuracy is sufficient for targeting small, clinically significant prostatic cancer lesions, but actual in vivo targeting will include additional error components that will have to be determined.

AB - Image-to-robot registration is a typical step for robotic image-guided interventions. If the imaging device uses a portable imaging probe that is held by a robot, this registration is constant and has been commonly named probe calibration. The same applies to probes tracked by a position measurement device. We report a calibration method for 2-D ultrasound probes using robotic manipulation and a planar calibration rig. Moreover, a needle guide that is attached to the probe is also calibrated for ultrasound-guided needle targeting. The method is applied to a transrectal ultrasound (TRUS) probe for robot-assisted prostate biopsy. Validation experiments include TRUS-guided needle targeting accuracy tests. This paper outlines the entire process from the calibration to image-guided targeting. Freehand TRUS-guided prostate biopsy is the primary method of diagnosing prostate cancer, with over 1.2 million procedures performed annually in the U.S. alone. However, freehand biopsy is a highly challenging procedure with subjective quality control. As such, biopsy devices are emerging to assist the physician. Here, we present a method that uses robotic TRUS manipulation. A 2-D TRUS probe is supported by a 4-degree-of-freedom robot. The robot performs ultrasound scanning, enabling 3-D reconstructions. Based on the images, the robot orients a needle guide on target for biopsy. The biopsy is acquired manually through the guide. In vitro tests showed that the 3-D images were geometrically accurate, and an image-based needle targeting accuracy was 1.55 mm. These validate the probe calibration presented and the overall robotic system for needle targeting. Targeting accuracy is sufficient for targeting small, clinically significant prostatic cancer lesions, but actual in vivo targeting will include additional error components that will have to be determined.

KW - Image-guided robot

KW - needle-guide calibration

KW - prostate biopsy

KW - registration

KW - ultrasound calibration

UR - http://www.scopus.com/inward/record.url?scp=84877903485&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84877903485&partnerID=8YFLogxK

U2 - 10.1109/TBME.2013.2241430

DO - 10.1109/TBME.2013.2241430

M3 - Article

VL - 60

SP - 1728

EP - 1734

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 6

M1 - 6416032

ER -