ICoNIK: Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit Representations in k-Space

Veronika Spieker; Wenqi Huang; Hannah Eichhorn; Jonathan Stelter; Kilian Weiss; Veronika A. Zimmer; Rickmer F. Braren; Dimitrios C. Karampinos; Kerstin Hammernik; Julia A. Schnabel

doi:10.1007/978-3-031-53767-7_18

ICoNIK: Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit Representations in k-Space

Veronika Spieker^*, Wenqi Huang, Hannah Eichhorn, Jonathan Stelter, Kilian Weiss, Veronika A. Zimmer, Rickmer F. Braren, Dimitrios C. Karampinos, Kerstin Hammernik, Julia A. Schnabel

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

2 Citations (Scopus)

Abstract

Motion-resolved reconstruction for abdominal magnetic resonance imaging (MRI) remains a challenge due to the trade-off between residual motion blurring caused by discretized motion states and undersampling artefacts. In this work, we generate blurring-free motion-resolved abdominal reconstructions by learning a neural implicit representation directly in k-space (NIK). Using measured sampling points and a data-derived respiratory navigator signal, we train a network to generate continuous signal values. To aid the regularization of sparsely sampled regions, we introduce an additional informed correction layer (ICo), which leverages information from neighboring regions to correct NIK’s prediction. The proposed generative reconstruction methods, NIK and ICoNIK, outperform standard motion-resolved reconstruction techniques and provide a promising solution to address motion artefacts in abdominal MRI.

Original language	English
Title of host publication	Deep Generative Models - Third MICCAI Workshop, DGM4MICCAI 2023, Held in Conjunction with MICCAI 2023, Proceedings
Editors	Anirban Mukhopadhyay, Ilkay Oksuz, Sandy Engelhardt, Dajiang Zhu, Yixuan Yuan
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	183-192
Number of pages	10
ISBN (Print)	9783031537660
DOIs	https://doi.org/10.1007/978-3-031-53767-7_18
Publication status	Published - 2024
Event	3rd Workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention, DGM4MICCAI 2023 Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023 - Vancouver, Canada Duration: 8 Oct 2023 → 12 Oct 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14533 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	3rd Workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention, DGM4MICCAI 2023 Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023
Country/Territory	Canada
City	Vancouver
Period	8/10/2023 → 12/10/2023

Keywords

Motion-Resolved Abdominal MRI
MRI Reconstruction
Neural Implicit Representations
Parallel Imaging

Access to Document

10.1007/978-3-031-53767-7_18

Cite this

Spieker, V., Huang, W., Eichhorn, H., Stelter, J., Weiss, K., Zimmer, V. A., Braren, R. F., Karampinos, D. C., Hammernik, K., & Schnabel, J. A. (2024). ICoNIK: Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit Representations in k-Space. In A. Mukhopadhyay, I. Oksuz, S. Engelhardt, D. Zhu, & Y. Yuan (Eds.), Deep Generative Models - Third MICCAI Workshop, DGM4MICCAI 2023, Held in Conjunction with MICCAI 2023, Proceedings (pp. 183-192). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14533 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-53767-7_18

Spieker, Veronika ; Huang, Wenqi ; Eichhorn, Hannah et al. / ICoNIK : Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit Representations in k-Space. Deep Generative Models - Third MICCAI Workshop, DGM4MICCAI 2023, Held in Conjunction with MICCAI 2023, Proceedings. editor / Anirban Mukhopadhyay ; Ilkay Oksuz ; Sandy Engelhardt ; Dajiang Zhu ; Yixuan Yuan. Springer Science and Business Media Deutschland GmbH, 2024. pp. 183-192 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{bab7b37a87b944a986f6fd2e2ee8bb1e,

title = "ICoNIK: Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit Representations in k-Space",

abstract = "Motion-resolved reconstruction for abdominal magnetic resonance imaging (MRI) remains a challenge due to the trade-off between residual motion blurring caused by discretized motion states and undersampling artefacts. In this work, we generate blurring-free motion-resolved abdominal reconstructions by learning a neural implicit representation directly in k-space (NIK). Using measured sampling points and a data-derived respiratory navigator signal, we train a network to generate continuous signal values. To aid the regularization of sparsely sampled regions, we introduce an additional informed correction layer (ICo), which leverages information from neighboring regions to correct NIK{\textquoteright}s prediction. The proposed generative reconstruction methods, NIK and ICoNIK, outperform standard motion-resolved reconstruction techniques and provide a promising solution to address motion artefacts in abdominal MRI.",

keywords = "Motion-Resolved Abdominal MRI, MRI Reconstruction, Neural Implicit Representations, Parallel Imaging",

author = "Veronika Spieker and Wenqi Huang and Hannah Eichhorn and Jonathan Stelter and Kilian Weiss and Zimmer, {Veronika A.} and Braren, {Rickmer F.} and Karampinos, {Dimitrios C.} and Kerstin Hammernik and Schnabel, {Julia A.}",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.; 3rd Workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention, DGM4MICCAI 2023 Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023 ; Conference date: 08-10-2023 Through 12-10-2023",

year = "2024",

doi = "10.1007/978-3-031-53767-7_18",

language = "English",

isbn = "9783031537660",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "183--192",

editor = "Anirban Mukhopadhyay and Ilkay Oksuz and Sandy Engelhardt and Dajiang Zhu and Yixuan Yuan",

booktitle = "Deep Generative Models - Third MICCAI Workshop, DGM4MICCAI 2023, Held in Conjunction with MICCAI 2023, Proceedings",

address = "Germany",

}

Spieker, V, Huang, W, Eichhorn, H, Stelter, J, Weiss, K, Zimmer, VA, Braren, RF, Karampinos, DC, Hammernik, K & Schnabel, JA 2024, ICoNIK: Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit Representations in k-Space. in A Mukhopadhyay, I Oksuz, S Engelhardt, D Zhu & Y Yuan (eds), Deep Generative Models - Third MICCAI Workshop, DGM4MICCAI 2023, Held in Conjunction with MICCAI 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14533 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 183-192, 3rd Workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention, DGM4MICCAI 2023 Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023, Vancouver, Canada, 8/10/2023. https://doi.org/10.1007/978-3-031-53767-7_18

ICoNIK: Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit Representations in k-Space. / Spieker, Veronika; Huang, Wenqi; Eichhorn, Hannah et al.
Deep Generative Models - Third MICCAI Workshop, DGM4MICCAI 2023, Held in Conjunction with MICCAI 2023, Proceedings. ed. / Anirban Mukhopadhyay; Ilkay Oksuz; Sandy Engelhardt; Dajiang Zhu; Yixuan Yuan. Springer Science and Business Media Deutschland GmbH, 2024. p. 183-192 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14533 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

TY - CHAP

T1 - ICoNIK

T2 - 3rd Workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention, DGM4MICCAI 2023 Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023

AU - Spieker, Veronika

AU - Huang, Wenqi

AU - Eichhorn, Hannah

AU - Stelter, Jonathan

AU - Weiss, Kilian

AU - Zimmer, Veronika A.

AU - Braren, Rickmer F.

AU - Karampinos, Dimitrios C.

AU - Hammernik, Kerstin

AU - Schnabel, Julia A.

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

PY - 2024

Y1 - 2024

N2 - Motion-resolved reconstruction for abdominal magnetic resonance imaging (MRI) remains a challenge due to the trade-off between residual motion blurring caused by discretized motion states and undersampling artefacts. In this work, we generate blurring-free motion-resolved abdominal reconstructions by learning a neural implicit representation directly in k-space (NIK). Using measured sampling points and a data-derived respiratory navigator signal, we train a network to generate continuous signal values. To aid the regularization of sparsely sampled regions, we introduce an additional informed correction layer (ICo), which leverages information from neighboring regions to correct NIK’s prediction. The proposed generative reconstruction methods, NIK and ICoNIK, outperform standard motion-resolved reconstruction techniques and provide a promising solution to address motion artefacts in abdominal MRI.

AB - Motion-resolved reconstruction for abdominal magnetic resonance imaging (MRI) remains a challenge due to the trade-off between residual motion blurring caused by discretized motion states and undersampling artefacts. In this work, we generate blurring-free motion-resolved abdominal reconstructions by learning a neural implicit representation directly in k-space (NIK). Using measured sampling points and a data-derived respiratory navigator signal, we train a network to generate continuous signal values. To aid the regularization of sparsely sampled regions, we introduce an additional informed correction layer (ICo), which leverages information from neighboring regions to correct NIK’s prediction. The proposed generative reconstruction methods, NIK and ICoNIK, outperform standard motion-resolved reconstruction techniques and provide a promising solution to address motion artefacts in abdominal MRI.

KW - Motion-Resolved Abdominal MRI

KW - MRI Reconstruction

KW - Neural Implicit Representations

KW - Parallel Imaging

UR - http://www.scopus.com/inward/record.url?scp=85187660449&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-53767-7_18

DO - 10.1007/978-3-031-53767-7_18

M3 - Conference paper

AN - SCOPUS:85187660449

SN - 9783031537660

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

SP - 183

EP - 192

BT - Deep Generative Models - Third MICCAI Workshop, DGM4MICCAI 2023, Held in Conjunction with MICCAI 2023, Proceedings

A2 - Mukhopadhyay, Anirban

A2 - Oksuz, Ilkay

A2 - Engelhardt, Sandy

A2 - Zhu, Dajiang

A2 - Yuan, Yixuan

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 8 October 2023 through 12 October 2023

ER -

Spieker V, Huang W, Eichhorn H, Stelter J, Weiss K, Zimmer VA et al. ICoNIK: Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit Representations in k-Space. In Mukhopadhyay A, Oksuz I, Engelhardt S, Zhu D, Yuan Y, editors, Deep Generative Models - Third MICCAI Workshop, DGM4MICCAI 2023, Held in Conjunction with MICCAI 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2024. p. 183-192. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-53767-7_18