Densely connected wavelet-based autoencoder for MR image reconstruction

Amir Aghabiglou; Ender M. Eksioglu

doi:10.1109/TSP55681.2022.9851354

Densely connected wavelet-based autoencoder for MR image reconstruction

Amir Aghabiglou^*, Ender M. Eksioglu

^*Bu çalışma için yazışmadan sorumlu yazar

Elektronik ve Hab.Müh.Bölümü

Istanbul Technical University

Araştırma sonucu: Kitap/Rapor/Konferans Bildirisinde Bölüm › Konferans katkısı › bilirkişi

5 Atıf (Scopus)

Özet

Recently, methods based on deep learning have been introduced to the literature as a solution for accelerating magnetic resonance imaging technique. However, Image reconstruction from subsampled data is an ill-posed problem. In the current study, the wavelet package has been applied to deep networks. The replacement of the conventional downsampling and upsampling layers with Discrete Wavelet Transform (DWT) and Inverse Wavelet Transform (IWT) improved the reconstruction results. Moreover, the consequence of this substitution has been investigated on potent densely connected deep networks. The proposed novelty resulted in promising performance improvement in MR Image reconstruction.

Orijinal dil	İngilizce
Ana bilgisayar yayını başlığı	2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022
Editörler	Norbert Herencsar
Yayınlayan	Institute of Electrical and Electronics Engineers Inc.
Sayfalar	212-215
Sayfa sayısı	4
ISBN (Elektronik)	9781665469487
DOI'lar	https://doi.org/10.1109/TSP55681.2022.9851354
Yayın durumu	Yayınlandı - 2022
Etkinlik	45th International Conference on Telecommunications and Signal Processing, TSP 2022 - Virtual, Online, Czech Republic Süre: 13 Tem 2022 → 15 Tem 2022

Yayın serisi

Adı	2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022

???event.eventtypes.event.conference???

???event.eventtypes.event.conference???	45th International Conference on Telecommunications and Signal Processing, TSP 2022
Ülke/Bölge	Czech Republic
Şehir	Virtual, Online
Periyot	13/07/22 → 15/07/22

Bibliyografik not

Publisher Copyright:
© 2022 IEEE.

Finansman

This work is supported by TUBITAK (The Scientific and Technological Research Council of Turkey) under project no. 119E248.

Finansörler	Finansör numarası
Türkiye Bilimsel ve Teknolojik Araştırma Kurumu	119E248

Belgeye Erişim

10.1109/TSP55681.2022.9851354

Diğer Dosyalar ve Linkler

Link to publication in Scopus

Alıntı Yap

Aghabiglou, A., & Eksioglu, E. M. (2022). Densely connected wavelet-based autoencoder for MR image reconstruction. In N. Herencsar (Ed.), 2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022 (pp. 212-215). (2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/TSP55681.2022.9851354

Aghabiglou, Amir ; Eksioglu, Ender M. / Densely connected wavelet-based autoencoder for MR image reconstruction. 2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022. editör / Norbert Herencsar. Institute of Electrical and Electronics Engineers Inc., 2022. pp. 212-215 (2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022).

@inproceedings{354351bef8ee424dba70fa6af2a7f62b,

title = "Densely connected wavelet-based autoencoder for MR image reconstruction",

abstract = "Recently, methods based on deep learning have been introduced to the literature as a solution for accelerating magnetic resonance imaging technique. However, Image reconstruction from subsampled data is an ill-posed problem. In the current study, the wavelet package has been applied to deep networks. The replacement of the conventional downsampling and upsampling layers with Discrete Wavelet Transform (DWT) and Inverse Wavelet Transform (IWT) improved the reconstruction results. Moreover, the consequence of this substitution has been investigated on potent densely connected deep networks. The proposed novelty resulted in promising performance improvement in MR Image reconstruction.",

keywords = "Deep learning, Densely Connected Residual Network, Magnetic resonance imaging, MR Image Reconstruction",

author = "Amir Aghabiglou and Eksioglu, \{Ender M.\}",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 45th International Conference on Telecommunications and Signal Processing, TSP 2022 ; Conference date: 13-07-2022 Through 15-07-2022",

year = "2022",

doi = "10.1109/TSP55681.2022.9851354",

language = "English",

series = "2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "212--215",

editor = "Norbert Herencsar",

booktitle = "2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022",

address = "United States",

}

Aghabiglou, A & Eksioglu, EM 2022, Densely connected wavelet-based autoencoder for MR image reconstruction. in N Herencsar (ed.), 2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022. 2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022, Institute of Electrical and Electronics Engineers Inc., pp. 212-215, 45th International Conference on Telecommunications and Signal Processing, TSP 2022, Virtual, Online, Czech Republic, 13/07/22. https://doi.org/10.1109/TSP55681.2022.9851354

Densely connected wavelet-based autoencoder for MR image reconstruction. / Aghabiglou, Amir; Eksioglu, Ender M.
2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022. ed. / Norbert Herencsar. Institute of Electrical and Electronics Engineers Inc., 2022. p. 212-215 (2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022).

Araştırma sonucu: Kitap/Rapor/Konferans Bildirisinde Bölüm › Konferans katkısı › bilirkişi

TY - GEN

T1 - Densely connected wavelet-based autoencoder for MR image reconstruction

AU - Aghabiglou, Amir

AU - Eksioglu, Ender M.

PY - 2022

Y1 - 2022

N2 - Recently, methods based on deep learning have been introduced to the literature as a solution for accelerating magnetic resonance imaging technique. However, Image reconstruction from subsampled data is an ill-posed problem. In the current study, the wavelet package has been applied to deep networks. The replacement of the conventional downsampling and upsampling layers with Discrete Wavelet Transform (DWT) and Inverse Wavelet Transform (IWT) improved the reconstruction results. Moreover, the consequence of this substitution has been investigated on potent densely connected deep networks. The proposed novelty resulted in promising performance improvement in MR Image reconstruction.

AB - Recently, methods based on deep learning have been introduced to the literature as a solution for accelerating magnetic resonance imaging technique. However, Image reconstruction from subsampled data is an ill-posed problem. In the current study, the wavelet package has been applied to deep networks. The replacement of the conventional downsampling and upsampling layers with Discrete Wavelet Transform (DWT) and Inverse Wavelet Transform (IWT) improved the reconstruction results. Moreover, the consequence of this substitution has been investigated on potent densely connected deep networks. The proposed novelty resulted in promising performance improvement in MR Image reconstruction.

KW - Deep learning

KW - Densely Connected Residual Network

KW - Magnetic resonance imaging

KW - MR Image Reconstruction

UR - https://www.scopus.com/pages/publications/85138051852

U2 - 10.1109/TSP55681.2022.9851354

DO - 10.1109/TSP55681.2022.9851354

M3 - Conference contribution

AN - SCOPUS:85138051852

T3 - 2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022

SP - 212

EP - 215

BT - 2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022

A2 - Herencsar, Norbert

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 45th International Conference on Telecommunications and Signal Processing, TSP 2022

Y2 - 13 July 2022 through 15 July 2022

ER -

Aghabiglou A, Eksioglu EM. Densely connected wavelet-based autoencoder for MR image reconstruction. In Herencsar N, editör, 2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022. Institute of Electrical and Electronics Engineers Inc. 2022. p. 212-215. (2022 45th International Conference on Telecommunications and Signal Processing, TSP 2022). doi: 10.1109/TSP55681.2022.9851354

Densely connected wavelet-based autoencoder for MR image reconstruction

Özet

Yayın serisi

???event.eventtypes.event.conference???

Bibliyografik not

Finansman

Belgeye Erişim

Diğer Dosyalar ve Linkler

Parmak izi

Alıntı Yap