Learning to Avoid Poor Images: Towards Task-aware C-arm Cone-beam CT Trajectories

Jan Nico Zaech, Cong Gao, Bastian Bier, Russell Taylor, Andreas Maier, Nassir Navab, Mathias Unberath

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Metal artifacts in computed tomography (CT) arise from a mismatch between physics of image formation and idealized assumptions during tomographic reconstruction. These artifacts are particularly strong around metal implants, inhibiting widespread adoption of 3D cone-beam CT (CBCT) despite clear opportunity for intra-operative verification of implant positioning, e. g. in spinal fusion surgery. On synthetic and real data, we demonstrate that much of the artifact can be avoided by acquiring better data for reconstruction in a task-aware and patient-specific manner, and describe the first step towards the envisioned task-aware CBCT protocol. The traditional short-scan CBCT trajectory is planar, with little room for scene-specific adjustment. We extend this trajectory by autonomously adjusting out-of-plane angulation. This enables C-arm source trajectories that are scene-specific in that they avoid acquiring “poor images”, characterized by beam hardening, photon starvation, and noise. The recommendation of ideal out-of-plane angulation is performed on-the-fly using a deep convolutional neural network that regresses a detectability-rank derived from imaging physics.

Original languageEnglish (US)
Title of host publicationMedical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings
EditorsDinggang Shen, Pew-Thian Yap, Tianming Liu, Terry M. Peters, Ali Khan, Lawrence H. Staib, Caroline Essert, Sean Zhou
PublisherSpringer
Pages11-19
Number of pages9
ISBN (Print)9783030322533
DOIs
StatePublished - Jan 1 2019
Event22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019 - Shenzhen, China
Duration: Oct 13 2019Oct 17 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11768 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019
CountryChina
CityShenzhen
Period10/13/1910/17/19

Fingerprint

Computed Tomography
Tomography
Cones
Cone
Trajectories
Trajectory
Implant
Physics
Metals
Detectability
Surgery
Hardening
Image processing
Fusion reactions
Photons
Positioning
Recommendations
Neural networks
Adjustment
Fusion

Keywords

  • Deep reinforcement learning
  • Robotic imaging

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Zaech, J. N., Gao, C., Bier, B., Taylor, R., Maier, A., Navab, N., & Unberath, M. (2019). Learning to Avoid Poor Images: Towards Task-aware C-arm Cone-beam CT Trajectories. In D. Shen, P-T. Yap, T. Liu, T. M. Peters, A. Khan, L. H. Staib, C. Essert, ... S. Zhou (Eds.), Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings (pp. 11-19). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11768 LNCS). Springer. https://doi.org/10.1007/978-3-030-32254-0_2

Learning to Avoid Poor Images : Towards Task-aware C-arm Cone-beam CT Trajectories. / Zaech, Jan Nico; Gao, Cong; Bier, Bastian; Taylor, Russell; Maier, Andreas; Navab, Nassir; Unberath, Mathias.

Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings. ed. / Dinggang Shen; Pew-Thian Yap; Tianming Liu; Terry M. Peters; Ali Khan; Lawrence H. Staib; Caroline Essert; Sean Zhou. Springer, 2019. p. 11-19 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11768 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zaech, JN, Gao, C, Bier, B, Taylor, R, Maier, A, Navab, N & Unberath, M 2019, Learning to Avoid Poor Images: Towards Task-aware C-arm Cone-beam CT Trajectories. in D Shen, P-T Yap, T Liu, TM Peters, A Khan, LH Staib, C Essert & S Zhou (eds), Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11768 LNCS, Springer, pp. 11-19, 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019, Shenzhen, China, 10/13/19. https://doi.org/10.1007/978-3-030-32254-0_2
Zaech JN, Gao C, Bier B, Taylor R, Maier A, Navab N et al. Learning to Avoid Poor Images: Towards Task-aware C-arm Cone-beam CT Trajectories. In Shen D, Yap P-T, Liu T, Peters TM, Khan A, Staib LH, Essert C, Zhou S, editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings. Springer. 2019. p. 11-19. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-32254-0_2
Zaech, Jan Nico ; Gao, Cong ; Bier, Bastian ; Taylor, Russell ; Maier, Andreas ; Navab, Nassir ; Unberath, Mathias. / Learning to Avoid Poor Images : Towards Task-aware C-arm Cone-beam CT Trajectories. Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings. editor / Dinggang Shen ; Pew-Thian Yap ; Tianming Liu ; Terry M. Peters ; Ali Khan ; Lawrence H. Staib ; Caroline Essert ; Sean Zhou. Springer, 2019. pp. 11-19 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{02052b9752dc43b3971db73428a76bf3,
title = "Learning to Avoid Poor Images: Towards Task-aware C-arm Cone-beam CT Trajectories",
abstract = "Metal artifacts in computed tomography (CT) arise from a mismatch between physics of image formation and idealized assumptions during tomographic reconstruction. These artifacts are particularly strong around metal implants, inhibiting widespread adoption of 3D cone-beam CT (CBCT) despite clear opportunity for intra-operative verification of implant positioning, e. g. in spinal fusion surgery. On synthetic and real data, we demonstrate that much of the artifact can be avoided by acquiring better data for reconstruction in a task-aware and patient-specific manner, and describe the first step towards the envisioned task-aware CBCT protocol. The traditional short-scan CBCT trajectory is planar, with little room for scene-specific adjustment. We extend this trajectory by autonomously adjusting out-of-plane angulation. This enables C-arm source trajectories that are scene-specific in that they avoid acquiring “poor images”, characterized by beam hardening, photon starvation, and noise. The recommendation of ideal out-of-plane angulation is performed on-the-fly using a deep convolutional neural network that regresses a detectability-rank derived from imaging physics.",
keywords = "Deep reinforcement learning, Robotic imaging",
author = "Zaech, {Jan Nico} and Cong Gao and Bastian Bier and Russell Taylor and Andreas Maier and Nassir Navab and Mathias Unberath",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-32254-0_2",
language = "English (US)",
isbn = "9783030322533",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer",
pages = "11--19",
editor = "Dinggang Shen and Pew-Thian Yap and Tianming Liu and Peters, {Terry M.} and Ali Khan and Staib, {Lawrence H.} and Caroline Essert and Sean Zhou",
booktitle = "Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings",

}

TY - GEN

T1 - Learning to Avoid Poor Images

T2 - Towards Task-aware C-arm Cone-beam CT Trajectories

AU - Zaech, Jan Nico

AU - Gao, Cong

AU - Bier, Bastian

AU - Taylor, Russell

AU - Maier, Andreas

AU - Navab, Nassir

AU - Unberath, Mathias

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Metal artifacts in computed tomography (CT) arise from a mismatch between physics of image formation and idealized assumptions during tomographic reconstruction. These artifacts are particularly strong around metal implants, inhibiting widespread adoption of 3D cone-beam CT (CBCT) despite clear opportunity for intra-operative verification of implant positioning, e. g. in spinal fusion surgery. On synthetic and real data, we demonstrate that much of the artifact can be avoided by acquiring better data for reconstruction in a task-aware and patient-specific manner, and describe the first step towards the envisioned task-aware CBCT protocol. The traditional short-scan CBCT trajectory is planar, with little room for scene-specific adjustment. We extend this trajectory by autonomously adjusting out-of-plane angulation. This enables C-arm source trajectories that are scene-specific in that they avoid acquiring “poor images”, characterized by beam hardening, photon starvation, and noise. The recommendation of ideal out-of-plane angulation is performed on-the-fly using a deep convolutional neural network that regresses a detectability-rank derived from imaging physics.

AB - Metal artifacts in computed tomography (CT) arise from a mismatch between physics of image formation and idealized assumptions during tomographic reconstruction. These artifacts are particularly strong around metal implants, inhibiting widespread adoption of 3D cone-beam CT (CBCT) despite clear opportunity for intra-operative verification of implant positioning, e. g. in spinal fusion surgery. On synthetic and real data, we demonstrate that much of the artifact can be avoided by acquiring better data for reconstruction in a task-aware and patient-specific manner, and describe the first step towards the envisioned task-aware CBCT protocol. The traditional short-scan CBCT trajectory is planar, with little room for scene-specific adjustment. We extend this trajectory by autonomously adjusting out-of-plane angulation. This enables C-arm source trajectories that are scene-specific in that they avoid acquiring “poor images”, characterized by beam hardening, photon starvation, and noise. The recommendation of ideal out-of-plane angulation is performed on-the-fly using a deep convolutional neural network that regresses a detectability-rank derived from imaging physics.

KW - Deep reinforcement learning

KW - Robotic imaging

UR - http://www.scopus.com/inward/record.url?scp=85075667701&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075667701&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-32254-0_2

DO - 10.1007/978-3-030-32254-0_2

M3 - Conference contribution

AN - SCOPUS:85075667701

SN - 9783030322533

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 11

EP - 19

BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings

A2 - Shen, Dinggang

A2 - Yap, Pew-Thian

A2 - Liu, Tianming

A2 - Peters, Terry M.

A2 - Khan, Ali

A2 - Staib, Lawrence H.

A2 - Essert, Caroline

A2 - Zhou, Sean

PB - Springer

ER -