A multichannel block-matching denoising algorithm for spectral photon-counting CT images

Adam P. Harrison, Ziyue Xu, Amir Pourmorteza, David A. Bluemke, Daniel J. Mollura

Research output: Contribution to journalArticle

Abstract

Purpose: We present a denoising algorithm designed for a whole-body prototype photon-counting computed tomography (PCCT) scanner with up to 4 energy thresholds and associated energy-binned images. Methods: Spectral PCCT images can exhibit low signal to noise ratios (SNRs) due to the limited photon counts in each simultaneously-acquired energy bin. To help address this, our denoising method exploits the correlation and exact alignment between energy bins, adapting the highly-effective blockmatching 3D (BM3D) denoising algorithm for PCCT. The original single-channel BM3D algorithm operates patch-by-patch. For each small patch in the image, a patch grouping action collects similar patches from the rest of the image, which are then collaboratively filtered together. The resulting performance hinges on accurate patch grouping. Our improved multi-channel version, called BM3D-PCCT, incorporates two improvements. First, BM3D-PCCTuses a more accurate shared patch grouping based on the image reconstructed from photons detected in all 4 energy bins. Second, BM3D-PCCT performs a cross-channel decorrelation, adding a further dimension to the collaborative filtering process. These two improvements produce a more effective algorithm for PCCT denoising. Results: Preliminary results compare BM3D-PCCT against BM3D-Naive, which denoises each energy bin independently. Experiments use a three-contrast PCCT image of a canine abdomen. Within five regions of interest, selected from paraspinal muscle, liver, and visceral fat, BM3D-PCCT reduces the noise standard deviation by 65.0%, compared to 40.4% for BM3D-Naive. Attenuation values of the contrast agents in calibration vials also cluster much tighter to their respective lines of best fit. Mean angular differences (in degrees) for the original, BM3D-Naive, and BM3D-PCCT images, respectively, were 15.61, 7.34, and 4.45 (iodine); 12.17, 7.17, and 4.39 (galodinium); and 12.86, 6.33, and 3.96 (bismuth). Conclusion: We outline a multi-channel denoising algorithm tailored for spectral PCCT images, demonstrating improved performance over an independent, yet state-of-the-art, single-channel approach.

Original languageEnglish (US)
Pages (from-to)2447-2452
Number of pages6
JournalMedical Physics
Volume44
Issue number6
DOIs
StatePublished - Jun 1 2017
Externally publishedYes

Fingerprint

Photons
Tomography
X-Ray Computed Tomography Scanners
Paraspinal Muscles
Bismuth
Intra-Abdominal Fat
Signal-To-Noise Ratio
Iodine
Abdomen
Contrast Media
Calibration
Noise
Canidae

Keywords

  • Block-matching 3D
  • Collaborative filtering
  • Denoising
  • Multi-channel
  • Photon-counting computed tomography

ASJC Scopus subject areas

  • Biophysics
  • Radiology Nuclear Medicine and imaging

Cite this

Harrison, A. P., Xu, Z., Pourmorteza, A., Bluemke, D. A., & Mollura, D. J. (2017). A multichannel block-matching denoising algorithm for spectral photon-counting CT images. Medical Physics, 44(6), 2447-2452. https://doi.org/10.1002/mp.12225

A multichannel block-matching denoising algorithm for spectral photon-counting CT images. / Harrison, Adam P.; Xu, Ziyue; Pourmorteza, Amir; Bluemke, David A.; Mollura, Daniel J.

In: Medical Physics, Vol. 44, No. 6, 01.06.2017, p. 2447-2452.

Research output: Contribution to journalArticle

Harrison, AP, Xu, Z, Pourmorteza, A, Bluemke, DA & Mollura, DJ 2017, 'A multichannel block-matching denoising algorithm for spectral photon-counting CT images', Medical Physics, vol. 44, no. 6, pp. 2447-2452. https://doi.org/10.1002/mp.12225
Harrison AP, Xu Z, Pourmorteza A, Bluemke DA, Mollura DJ. A multichannel block-matching denoising algorithm for spectral photon-counting CT images. Medical Physics. 2017 Jun 1;44(6):2447-2452. https://doi.org/10.1002/mp.12225
Harrison, Adam P. ; Xu, Ziyue ; Pourmorteza, Amir ; Bluemke, David A. ; Mollura, Daniel J. / A multichannel block-matching denoising algorithm for spectral photon-counting CT images. In: Medical Physics. 2017 ; Vol. 44, No. 6. pp. 2447-2452.
@article{626e310861cb4f7d84d13269263171f6,
title = "A multichannel block-matching denoising algorithm for spectral photon-counting CT images",
abstract = "Purpose: We present a denoising algorithm designed for a whole-body prototype photon-counting computed tomography (PCCT) scanner with up to 4 energy thresholds and associated energy-binned images. Methods: Spectral PCCT images can exhibit low signal to noise ratios (SNRs) due to the limited photon counts in each simultaneously-acquired energy bin. To help address this, our denoising method exploits the correlation and exact alignment between energy bins, adapting the highly-effective blockmatching 3D (BM3D) denoising algorithm for PCCT. The original single-channel BM3D algorithm operates patch-by-patch. For each small patch in the image, a patch grouping action collects similar patches from the rest of the image, which are then collaboratively filtered together. The resulting performance hinges on accurate patch grouping. Our improved multi-channel version, called BM3D-PCCT, incorporates two improvements. First, BM3D-PCCTuses a more accurate shared patch grouping based on the image reconstructed from photons detected in all 4 energy bins. Second, BM3D-PCCT performs a cross-channel decorrelation, adding a further dimension to the collaborative filtering process. These two improvements produce a more effective algorithm for PCCT denoising. Results: Preliminary results compare BM3D-PCCT against BM3D-Naive, which denoises each energy bin independently. Experiments use a three-contrast PCCT image of a canine abdomen. Within five regions of interest, selected from paraspinal muscle, liver, and visceral fat, BM3D-PCCT reduces the noise standard deviation by 65.0{\%}, compared to 40.4{\%} for BM3D-Naive. Attenuation values of the contrast agents in calibration vials also cluster much tighter to their respective lines of best fit. Mean angular differences (in degrees) for the original, BM3D-Naive, and BM3D-PCCT images, respectively, were 15.61, 7.34, and 4.45 (iodine); 12.17, 7.17, and 4.39 (galodinium); and 12.86, 6.33, and 3.96 (bismuth). Conclusion: We outline a multi-channel denoising algorithm tailored for spectral PCCT images, demonstrating improved performance over an independent, yet state-of-the-art, single-channel approach.",
keywords = "Block-matching 3D, Collaborative filtering, Denoising, Multi-channel, Photon-counting computed tomography",
author = "Harrison, {Adam P.} and Ziyue Xu and Amir Pourmorteza and Bluemke, {David A.} and Mollura, {Daniel J.}",
year = "2017",
month = "6",
day = "1",
doi = "10.1002/mp.12225",
language = "English (US)",
volume = "44",
pages = "2447--2452",
journal = "Medical Physics",
issn = "0094-2405",
publisher = "AAPM - American Association of Physicists in Medicine",
number = "6",

}

TY - JOUR

T1 - A multichannel block-matching denoising algorithm for spectral photon-counting CT images

AU - Harrison, Adam P.

AU - Xu, Ziyue

AU - Pourmorteza, Amir

AU - Bluemke, David A.

AU - Mollura, Daniel J.

PY - 2017/6/1

Y1 - 2017/6/1

N2 - Purpose: We present a denoising algorithm designed for a whole-body prototype photon-counting computed tomography (PCCT) scanner with up to 4 energy thresholds and associated energy-binned images. Methods: Spectral PCCT images can exhibit low signal to noise ratios (SNRs) due to the limited photon counts in each simultaneously-acquired energy bin. To help address this, our denoising method exploits the correlation and exact alignment between energy bins, adapting the highly-effective blockmatching 3D (BM3D) denoising algorithm for PCCT. The original single-channel BM3D algorithm operates patch-by-patch. For each small patch in the image, a patch grouping action collects similar patches from the rest of the image, which are then collaboratively filtered together. The resulting performance hinges on accurate patch grouping. Our improved multi-channel version, called BM3D-PCCT, incorporates two improvements. First, BM3D-PCCTuses a more accurate shared patch grouping based on the image reconstructed from photons detected in all 4 energy bins. Second, BM3D-PCCT performs a cross-channel decorrelation, adding a further dimension to the collaborative filtering process. These two improvements produce a more effective algorithm for PCCT denoising. Results: Preliminary results compare BM3D-PCCT against BM3D-Naive, which denoises each energy bin independently. Experiments use a three-contrast PCCT image of a canine abdomen. Within five regions of interest, selected from paraspinal muscle, liver, and visceral fat, BM3D-PCCT reduces the noise standard deviation by 65.0%, compared to 40.4% for BM3D-Naive. Attenuation values of the contrast agents in calibration vials also cluster much tighter to their respective lines of best fit. Mean angular differences (in degrees) for the original, BM3D-Naive, and BM3D-PCCT images, respectively, were 15.61, 7.34, and 4.45 (iodine); 12.17, 7.17, and 4.39 (galodinium); and 12.86, 6.33, and 3.96 (bismuth). Conclusion: We outline a multi-channel denoising algorithm tailored for spectral PCCT images, demonstrating improved performance over an independent, yet state-of-the-art, single-channel approach.

AB - Purpose: We present a denoising algorithm designed for a whole-body prototype photon-counting computed tomography (PCCT) scanner with up to 4 energy thresholds and associated energy-binned images. Methods: Spectral PCCT images can exhibit low signal to noise ratios (SNRs) due to the limited photon counts in each simultaneously-acquired energy bin. To help address this, our denoising method exploits the correlation and exact alignment between energy bins, adapting the highly-effective blockmatching 3D (BM3D) denoising algorithm for PCCT. The original single-channel BM3D algorithm operates patch-by-patch. For each small patch in the image, a patch grouping action collects similar patches from the rest of the image, which are then collaboratively filtered together. The resulting performance hinges on accurate patch grouping. Our improved multi-channel version, called BM3D-PCCT, incorporates two improvements. First, BM3D-PCCTuses a more accurate shared patch grouping based on the image reconstructed from photons detected in all 4 energy bins. Second, BM3D-PCCT performs a cross-channel decorrelation, adding a further dimension to the collaborative filtering process. These two improvements produce a more effective algorithm for PCCT denoising. Results: Preliminary results compare BM3D-PCCT against BM3D-Naive, which denoises each energy bin independently. Experiments use a three-contrast PCCT image of a canine abdomen. Within five regions of interest, selected from paraspinal muscle, liver, and visceral fat, BM3D-PCCT reduces the noise standard deviation by 65.0%, compared to 40.4% for BM3D-Naive. Attenuation values of the contrast agents in calibration vials also cluster much tighter to their respective lines of best fit. Mean angular differences (in degrees) for the original, BM3D-Naive, and BM3D-PCCT images, respectively, were 15.61, 7.34, and 4.45 (iodine); 12.17, 7.17, and 4.39 (galodinium); and 12.86, 6.33, and 3.96 (bismuth). Conclusion: We outline a multi-channel denoising algorithm tailored for spectral PCCT images, demonstrating improved performance over an independent, yet state-of-the-art, single-channel approach.

KW - Block-matching 3D

KW - Collaborative filtering

KW - Denoising

KW - Multi-channel

KW - Photon-counting computed tomography

UR - http://www.scopus.com/inward/record.url?scp=85024370635&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85024370635&partnerID=8YFLogxK

U2 - 10.1002/mp.12225

DO - 10.1002/mp.12225

M3 - Article

C2 - 28332211

AN - SCOPUS:85024370635

VL - 44

SP - 2447

EP - 2452

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

IS - 6

ER -