A deep learning framework for transforming image reconstruction into pixel classification

Kamlesh Pawar, Zhaolin Chen, N. Jon Shah, Gary F. Egan

Research output: Contribution to journalArticleResearchpeer-review

Abstract

A deep learning framework is presented that transforms the task of MR image reconstruction from randomly undersampled k-space data into pixel classification. A DL network was trained to remove incoherent undersampling artifacts from MR images. The underlying, fully sampled, target image was represented as a discrete quantized image. The quantization step enables the design of a convolutional neural network (CNN) that can classify each pixel in the input image to a discrete quantized level. The reconstructed image quality of the proposed DL classification model was compared with conventional compressed sensing (CS) and a DL regression model. The reconstructed images using the DL classification model outperformed the state-of-the-art compressed sensing and DL regression models with a similar number of parameters assessed using quantitative measures. The experiments reveal that the proposed deep learning method is robust to noise and is able to reconstruct high-quality images in low SNR scenarios where conventional CS reconstructions and DL regression networks perform poorly. A generic design framework for transforming MR image reconstruction into pixel classification is developed. The proposed method can be easily incorporated into other DL-based image reconstruction methods.

Original languageEnglish
Article number8931762
Pages (from-to)177690-177702
Number of pages13
JournalIEEE Access
Volume7
DOIs
Publication statusPublished - 2019

Keywords

  • compressive sensing
  • deep learning
  • Magnetic resonance imaging

Cite this

@article{4260ff07d8134de09500c9c5306a95ca,
title = "A deep learning framework for transforming image reconstruction into pixel classification",
abstract = "A deep learning framework is presented that transforms the task of MR image reconstruction from randomly undersampled k-space data into pixel classification. A DL network was trained to remove incoherent undersampling artifacts from MR images. The underlying, fully sampled, target image was represented as a discrete quantized image. The quantization step enables the design of a convolutional neural network (CNN) that can classify each pixel in the input image to a discrete quantized level. The reconstructed image quality of the proposed DL classification model was compared with conventional compressed sensing (CS) and a DL regression model. The reconstructed images using the DL classification model outperformed the state-of-the-art compressed sensing and DL regression models with a similar number of parameters assessed using quantitative measures. The experiments reveal that the proposed deep learning method is robust to noise and is able to reconstruct high-quality images in low SNR scenarios where conventional CS reconstructions and DL regression networks perform poorly. A generic design framework for transforming MR image reconstruction into pixel classification is developed. The proposed method can be easily incorporated into other DL-based image reconstruction methods.",
keywords = "compressive sensing, deep learning, Magnetic resonance imaging",
author = "Kamlesh Pawar and Zhaolin Chen and Shah, {N. Jon} and Egan, {Gary F.}",
year = "2019",
doi = "10.1109/ACCESS.2019.2959037",
language = "English",
volume = "7",
pages = "177690--177702",
journal = "IEEE Access",
issn = "2169-3536",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",

}

A deep learning framework for transforming image reconstruction into pixel classification. / Pawar, Kamlesh; Chen, Zhaolin; Shah, N. Jon; Egan, Gary F.

In: IEEE Access, Vol. 7, 8931762, 2019, p. 177690-177702.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - A deep learning framework for transforming image reconstruction into pixel classification

AU - Pawar, Kamlesh

AU - Chen, Zhaolin

AU - Shah, N. Jon

AU - Egan, Gary F.

PY - 2019

Y1 - 2019

N2 - A deep learning framework is presented that transforms the task of MR image reconstruction from randomly undersampled k-space data into pixel classification. A DL network was trained to remove incoherent undersampling artifacts from MR images. The underlying, fully sampled, target image was represented as a discrete quantized image. The quantization step enables the design of a convolutional neural network (CNN) that can classify each pixel in the input image to a discrete quantized level. The reconstructed image quality of the proposed DL classification model was compared with conventional compressed sensing (CS) and a DL regression model. The reconstructed images using the DL classification model outperformed the state-of-the-art compressed sensing and DL regression models with a similar number of parameters assessed using quantitative measures. The experiments reveal that the proposed deep learning method is robust to noise and is able to reconstruct high-quality images in low SNR scenarios where conventional CS reconstructions and DL regression networks perform poorly. A generic design framework for transforming MR image reconstruction into pixel classification is developed. The proposed method can be easily incorporated into other DL-based image reconstruction methods.

AB - A deep learning framework is presented that transforms the task of MR image reconstruction from randomly undersampled k-space data into pixel classification. A DL network was trained to remove incoherent undersampling artifacts from MR images. The underlying, fully sampled, target image was represented as a discrete quantized image. The quantization step enables the design of a convolutional neural network (CNN) that can classify each pixel in the input image to a discrete quantized level. The reconstructed image quality of the proposed DL classification model was compared with conventional compressed sensing (CS) and a DL regression model. The reconstructed images using the DL classification model outperformed the state-of-the-art compressed sensing and DL regression models with a similar number of parameters assessed using quantitative measures. The experiments reveal that the proposed deep learning method is robust to noise and is able to reconstruct high-quality images in low SNR scenarios where conventional CS reconstructions and DL regression networks perform poorly. A generic design framework for transforming MR image reconstruction into pixel classification is developed. The proposed method can be easily incorporated into other DL-based image reconstruction methods.

KW - compressive sensing

KW - deep learning

KW - Magnetic resonance imaging

UR - http://www.scopus.com/inward/record.url?scp=85077184052&partnerID=8YFLogxK

U2 - 10.1109/ACCESS.2019.2959037

DO - 10.1109/ACCESS.2019.2959037

M3 - Article

AN - SCOPUS:85077184052

VL - 7

SP - 177690

EP - 177702

JO - IEEE Access

JF - IEEE Access

SN - 2169-3536

M1 - 8931762

ER -