Computed Tomography Image Quality Evaluation of a New Iterative Reconstruction Algorithm in the Abdomen (Adaptive Statistical Iterative Reconstruction-V) a Comparison with Model-Based Iterative Reconstruction, Adaptive Statistical Iterative Reconstruction, and Filtered Back Projection Reconstructions

Martin Halmes Goodenberger, Nicolaus A. Wagner-Bartak, Shiva Gupta, Xinming Liu, Ramon Q. Yap, Jia Sun, Eric P. Tamm, Corey T. Jensen

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Objective The purpose of this study was to compare abdominopelvic computed tomography images reconstructed with adaptive statistical iterative reconstruction-V (ASIR-V) with model-based iterative reconstruction (Veo 3.0), ASIR, and filtered back projection (FBP). Methods and Materials Abdominopelvic computed tomography scans for 36 patients (26 males and 10 females) were reconstructed using FBP, ASIR (80%), Veo 3.0, and ASIR-V (30%, 60%, 90%). Mean ± SD patient age was 32 ± 10 years with mean ± SD body mass index of 26.9 ± 4.4 kg/m 2 . Images were reviewed by 2 independent readers in a blinded, randomized fashion. Hounsfield unit, noise, and contrast-to-noise ratio (CNR) values were calculated for each reconstruction algorithm for further comparison. Phantom evaluation of low-contrast detectability (LCD) and high-contrast resolution was performed. Results Adaptive statistical iterative reconstruction-V 30%, ASIR-V 60%, and ASIR 80% were generally superior qualitatively compared with ASIR-V 90%, Veo 3.0, and FBP (P < 0.05). Adaptive statistical iterative reconstruction-V 90% showed superior LCD and had the highest CNR in the liver, aorta, and, pancreas, measuring 7.32 ± 3.22, 11.60 ± 4.25, and 4.60 ± 2.31, respectively, compared with the next best series of ASIR-V 60% with respective CNR values of 5.54 ± 2.39, 8.78 ± 3.15, and 3.49 ± 1.77 (P <0.0001). Veo 3.0 and ASIR 80% had the best and worst spatial resolution, respectively. Conclusions Adaptive statistical iterative reconstruction-V 30% and ASIR-V 60% provided the best combination of qualitative and quantitative performance. Adaptive statistical iterative reconstruction 80% was equivalent qualitatively, but demonstrated inferior spatial resolution and LCD.

Original languageEnglish (US)
Pages (from-to)184-190
Number of pages7
JournalJournal of Computer Assisted Tomography
Volume42
Issue number2
DOIs
StatePublished - Mar 1 2018

Fingerprint

Abdomen
Noise
Tomography
Aorta
Pancreas
Body Mass Index
Liver

Keywords

  • ASIR
  • CT image quality
  • MBIR
  • abdomen
  • iterative reconstruction

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Cite this

@article{a542a204341f4f6cbb77d62a488df6c9,
title = "Computed Tomography Image Quality Evaluation of a New Iterative Reconstruction Algorithm in the Abdomen (Adaptive Statistical Iterative Reconstruction-V) a Comparison with Model-Based Iterative Reconstruction, Adaptive Statistical Iterative Reconstruction, and Filtered Back Projection Reconstructions",
abstract = "Objective The purpose of this study was to compare abdominopelvic computed tomography images reconstructed with adaptive statistical iterative reconstruction-V (ASIR-V) with model-based iterative reconstruction (Veo 3.0), ASIR, and filtered back projection (FBP). Methods and Materials Abdominopelvic computed tomography scans for 36 patients (26 males and 10 females) were reconstructed using FBP, ASIR (80{\%}), Veo 3.0, and ASIR-V (30{\%}, 60{\%}, 90{\%}). Mean ± SD patient age was 32 ± 10 years with mean ± SD body mass index of 26.9 ± 4.4 kg/m 2 . Images were reviewed by 2 independent readers in a blinded, randomized fashion. Hounsfield unit, noise, and contrast-to-noise ratio (CNR) values were calculated for each reconstruction algorithm for further comparison. Phantom evaluation of low-contrast detectability (LCD) and high-contrast resolution was performed. Results Adaptive statistical iterative reconstruction-V 30{\%}, ASIR-V 60{\%}, and ASIR 80{\%} were generally superior qualitatively compared with ASIR-V 90{\%}, Veo 3.0, and FBP (P < 0.05). Adaptive statistical iterative reconstruction-V 90{\%} showed superior LCD and had the highest CNR in the liver, aorta, and, pancreas, measuring 7.32 ± 3.22, 11.60 ± 4.25, and 4.60 ± 2.31, respectively, compared with the next best series of ASIR-V 60{\%} with respective CNR values of 5.54 ± 2.39, 8.78 ± 3.15, and 3.49 ± 1.77 (P <0.0001). Veo 3.0 and ASIR 80{\%} had the best and worst spatial resolution, respectively. Conclusions Adaptive statistical iterative reconstruction-V 30{\%} and ASIR-V 60{\%} provided the best combination of qualitative and quantitative performance. Adaptive statistical iterative reconstruction 80{\%} was equivalent qualitatively, but demonstrated inferior spatial resolution and LCD.",
keywords = "ASIR, CT image quality, MBIR, abdomen, iterative reconstruction",
author = "Goodenberger, {Martin Halmes} and Wagner-Bartak, {Nicolaus A.} and Shiva Gupta and Xinming Liu and Yap, {Ramon Q.} and Jia Sun and Tamm, {Eric P.} and Jensen, {Corey T.}",
year = "2018",
month = "3",
day = "1",
doi = "10.1097/RCT.0000000000000666",
language = "English (US)",
volume = "42",
pages = "184--190",
journal = "Journal of Computer Assisted Tomography",
issn = "0363-8715",
publisher = "Lippincott Williams and Wilkins",
number = "2",

}

TY - JOUR

T1 - Computed Tomography Image Quality Evaluation of a New Iterative Reconstruction Algorithm in the Abdomen (Adaptive Statistical Iterative Reconstruction-V) a Comparison with Model-Based Iterative Reconstruction, Adaptive Statistical Iterative Reconstruction, and Filtered Back Projection Reconstructions

AU - Goodenberger, Martin Halmes

AU - Wagner-Bartak, Nicolaus A.

AU - Gupta, Shiva

AU - Liu, Xinming

AU - Yap, Ramon Q.

AU - Sun, Jia

AU - Tamm, Eric P.

AU - Jensen, Corey T.

PY - 2018/3/1

Y1 - 2018/3/1

N2 - Objective The purpose of this study was to compare abdominopelvic computed tomography images reconstructed with adaptive statistical iterative reconstruction-V (ASIR-V) with model-based iterative reconstruction (Veo 3.0), ASIR, and filtered back projection (FBP). Methods and Materials Abdominopelvic computed tomography scans for 36 patients (26 males and 10 females) were reconstructed using FBP, ASIR (80%), Veo 3.0, and ASIR-V (30%, 60%, 90%). Mean ± SD patient age was 32 ± 10 years with mean ± SD body mass index of 26.9 ± 4.4 kg/m 2 . Images were reviewed by 2 independent readers in a blinded, randomized fashion. Hounsfield unit, noise, and contrast-to-noise ratio (CNR) values were calculated for each reconstruction algorithm for further comparison. Phantom evaluation of low-contrast detectability (LCD) and high-contrast resolution was performed. Results Adaptive statistical iterative reconstruction-V 30%, ASIR-V 60%, and ASIR 80% were generally superior qualitatively compared with ASIR-V 90%, Veo 3.0, and FBP (P < 0.05). Adaptive statistical iterative reconstruction-V 90% showed superior LCD and had the highest CNR in the liver, aorta, and, pancreas, measuring 7.32 ± 3.22, 11.60 ± 4.25, and 4.60 ± 2.31, respectively, compared with the next best series of ASIR-V 60% with respective CNR values of 5.54 ± 2.39, 8.78 ± 3.15, and 3.49 ± 1.77 (P <0.0001). Veo 3.0 and ASIR 80% had the best and worst spatial resolution, respectively. Conclusions Adaptive statistical iterative reconstruction-V 30% and ASIR-V 60% provided the best combination of qualitative and quantitative performance. Adaptive statistical iterative reconstruction 80% was equivalent qualitatively, but demonstrated inferior spatial resolution and LCD.

AB - Objective The purpose of this study was to compare abdominopelvic computed tomography images reconstructed with adaptive statistical iterative reconstruction-V (ASIR-V) with model-based iterative reconstruction (Veo 3.0), ASIR, and filtered back projection (FBP). Methods and Materials Abdominopelvic computed tomography scans for 36 patients (26 males and 10 females) were reconstructed using FBP, ASIR (80%), Veo 3.0, and ASIR-V (30%, 60%, 90%). Mean ± SD patient age was 32 ± 10 years with mean ± SD body mass index of 26.9 ± 4.4 kg/m 2 . Images were reviewed by 2 independent readers in a blinded, randomized fashion. Hounsfield unit, noise, and contrast-to-noise ratio (CNR) values were calculated for each reconstruction algorithm for further comparison. Phantom evaluation of low-contrast detectability (LCD) and high-contrast resolution was performed. Results Adaptive statistical iterative reconstruction-V 30%, ASIR-V 60%, and ASIR 80% were generally superior qualitatively compared with ASIR-V 90%, Veo 3.0, and FBP (P < 0.05). Adaptive statistical iterative reconstruction-V 90% showed superior LCD and had the highest CNR in the liver, aorta, and, pancreas, measuring 7.32 ± 3.22, 11.60 ± 4.25, and 4.60 ± 2.31, respectively, compared with the next best series of ASIR-V 60% with respective CNR values of 5.54 ± 2.39, 8.78 ± 3.15, and 3.49 ± 1.77 (P <0.0001). Veo 3.0 and ASIR 80% had the best and worst spatial resolution, respectively. Conclusions Adaptive statistical iterative reconstruction-V 30% and ASIR-V 60% provided the best combination of qualitative and quantitative performance. Adaptive statistical iterative reconstruction 80% was equivalent qualitatively, but demonstrated inferior spatial resolution and LCD.

KW - ASIR

KW - CT image quality

KW - MBIR

KW - abdomen

KW - iterative reconstruction

UR - http://www.scopus.com/inward/record.url?scp=85036589508&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85036589508&partnerID=8YFLogxK

U2 - 10.1097/RCT.0000000000000666

DO - 10.1097/RCT.0000000000000666

M3 - Article

VL - 42

SP - 184

EP - 190

JO - Journal of Computer Assisted Tomography

JF - Journal of Computer Assisted Tomography

SN - 0363-8715

IS - 2

ER -