In-vivo study of quantitative ultrasound parameters in fatty rabbit livers

T. Nguyen, A. Podkowa, R. J. Miller, Michael L Oelze, Minh N Do

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease and can often lead to fibrosis, cirrhosis, cancer and complete liver failure. Liver biopsy is the current standard of care to quantify hepatic steatosis, but it comes with increased patient risk and only samples a small portion of the liver. Imaging approaches to assess NAFLD include proton density fat fraction (PDFF) estimated via MRI and shear wave elastography. However, MRI is often prohibitively expensive and shear wave elastography is not sensitive to fat content of the liver [1]. On the other hand, ultrasonic attenuation and the backscatter coefficient (BSC) have been observed to be sensitive to levels of fat in the liver. In this study, we explored the use of attenuation and a principal component analysis (PCA) of the BSC to detect and quantify hepatic steatosis in vivo in a rabbit model of fatty liver. Rabbits were maintained on a high fat diet for 0, 1, 2, 3 or 6 weeks with three rabbits per diet group (total N = 15). For analysis, rabbits were separated into three classes based on the total lipid content of the livers estimated using the Folch method: low fat (< 5%), medium fat (from 5% to 10%) and high fat (> 10%). An array transducer L9-4 with center frequency of 4 MHz connected to a SonixOne scanner was used to gather RF backscattered data in vivo from rabbits. The RF signals were used to estimate an average attenuation and BSC for each rabbit. The first five principal components of the PCA from the BSCs were used as input features to a support vector machine (SVM) for classification and comparison to total liver lipid levels. The slope of the attenuation coefficient provided statistically significant differences (p < 0.00058, using two-sample t-test) between low and high lipid fat groups. The proposed PCA-SVM based classification system yielded a classification accuracy of 63.7%, 32.6% and 71.2% for the low, medium and high fat groups, respectively. The results suggest that attenuation and BSC analysis can differentiate low versus high fat livers in a rabbit model. This work was supported by a grant from NIH (R21EB020766).

Original languageEnglish (US)
Title of host publication2017 IEEE International Ultrasonics Symposium, IUS 2017
PublisherIEEE Computer Society
ISBN (Electronic)9781538633830
DOIs
StatePublished - Oct 31 2017
Event2017 IEEE International Ultrasonics Symposium, IUS 2017 - Washington, United States
Duration: Sep 6 2017Sep 9 2017

Publication series

NameIEEE International Ultrasonics Symposium, IUS
ISSN (Print)1948-5719
ISSN (Electronic)1948-5727

Other

Other2017 IEEE International Ultrasonics Symposium, IUS 2017
CountryUnited States
CityWashington
Period9/6/179/9/17

Fingerprint

rabbits
liver
fats
principal components analysis
attenuation
lipids
diets
coefficients
S waves
fibrosis
attenuation coefficients
scanners
transducers
ultrasonics
cancer
slopes

Keywords

  • Fatty liver
  • PCA
  • QUS
  • SVM

ASJC Scopus subject areas

  • Acoustics and Ultrasonics

Cite this

Nguyen, T., Podkowa, A., Miller, R. J., Oelze, M. L., & Do, M. N. (2017). In-vivo study of quantitative ultrasound parameters in fatty rabbit livers. In 2017 IEEE International Ultrasonics Symposium, IUS 2017 [8091693] (IEEE International Ultrasonics Symposium, IUS). IEEE Computer Society. https://doi.org/10.1109/ULTSYM.2017.8091693

In-vivo study of quantitative ultrasound parameters in fatty rabbit livers. / Nguyen, T.; Podkowa, A.; Miller, R. J.; Oelze, Michael L; Do, Minh N.

2017 IEEE International Ultrasonics Symposium, IUS 2017. IEEE Computer Society, 2017. 8091693 (IEEE International Ultrasonics Symposium, IUS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nguyen, T, Podkowa, A, Miller, RJ, Oelze, ML & Do, MN 2017, In-vivo study of quantitative ultrasound parameters in fatty rabbit livers. in 2017 IEEE International Ultrasonics Symposium, IUS 2017., 8091693, IEEE International Ultrasonics Symposium, IUS, IEEE Computer Society, 2017 IEEE International Ultrasonics Symposium, IUS 2017, Washington, United States, 9/6/17. https://doi.org/10.1109/ULTSYM.2017.8091693
Nguyen T, Podkowa A, Miller RJ, Oelze ML, Do MN. In-vivo study of quantitative ultrasound parameters in fatty rabbit livers. In 2017 IEEE International Ultrasonics Symposium, IUS 2017. IEEE Computer Society. 2017. 8091693. (IEEE International Ultrasonics Symposium, IUS). https://doi.org/10.1109/ULTSYM.2017.8091693
Nguyen, T. ; Podkowa, A. ; Miller, R. J. ; Oelze, Michael L ; Do, Minh N. / In-vivo study of quantitative ultrasound parameters in fatty rabbit livers. 2017 IEEE International Ultrasonics Symposium, IUS 2017. IEEE Computer Society, 2017. (IEEE International Ultrasonics Symposium, IUS).
@inproceedings{f074235d53e6467b9717774db9af7339,
title = "In-vivo study of quantitative ultrasound parameters in fatty rabbit livers",
abstract = "Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease and can often lead to fibrosis, cirrhosis, cancer and complete liver failure. Liver biopsy is the current standard of care to quantify hepatic steatosis, but it comes with increased patient risk and only samples a small portion of the liver. Imaging approaches to assess NAFLD include proton density fat fraction (PDFF) estimated via MRI and shear wave elastography. However, MRI is often prohibitively expensive and shear wave elastography is not sensitive to fat content of the liver [1]. On the other hand, ultrasonic attenuation and the backscatter coefficient (BSC) have been observed to be sensitive to levels of fat in the liver. In this study, we explored the use of attenuation and a principal component analysis (PCA) of the BSC to detect and quantify hepatic steatosis in vivo in a rabbit model of fatty liver. Rabbits were maintained on a high fat diet for 0, 1, 2, 3 or 6 weeks with three rabbits per diet group (total N = 15). For analysis, rabbits were separated into three classes based on the total lipid content of the livers estimated using the Folch method: low fat (< 5{\%}), medium fat (from 5{\%} to 10{\%}) and high fat (> 10{\%}). An array transducer L9-4 with center frequency of 4 MHz connected to a SonixOne scanner was used to gather RF backscattered data in vivo from rabbits. The RF signals were used to estimate an average attenuation and BSC for each rabbit. The first five principal components of the PCA from the BSCs were used as input features to a support vector machine (SVM) for classification and comparison to total liver lipid levels. The slope of the attenuation coefficient provided statistically significant differences (p < 0.00058, using two-sample t-test) between low and high lipid fat groups. The proposed PCA-SVM based classification system yielded a classification accuracy of 63.7{\%}, 32.6{\%} and 71.2{\%} for the low, medium and high fat groups, respectively. The results suggest that attenuation and BSC analysis can differentiate low versus high fat livers in a rabbit model. This work was supported by a grant from NIH (R21EB020766).",
keywords = "Fatty liver, PCA, QUS, SVM",
author = "T. Nguyen and A. Podkowa and Miller, {R. J.} and Oelze, {Michael L} and Do, {Minh N}",
year = "2017",
month = "10",
day = "31",
doi = "10.1109/ULTSYM.2017.8091693",
language = "English (US)",
series = "IEEE International Ultrasonics Symposium, IUS",
publisher = "IEEE Computer Society",
booktitle = "2017 IEEE International Ultrasonics Symposium, IUS 2017",

}

TY - GEN

T1 - In-vivo study of quantitative ultrasound parameters in fatty rabbit livers

AU - Nguyen, T.

AU - Podkowa, A.

AU - Miller, R. J.

AU - Oelze, Michael L

AU - Do, Minh N

PY - 2017/10/31

Y1 - 2017/10/31

N2 - Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease and can often lead to fibrosis, cirrhosis, cancer and complete liver failure. Liver biopsy is the current standard of care to quantify hepatic steatosis, but it comes with increased patient risk and only samples a small portion of the liver. Imaging approaches to assess NAFLD include proton density fat fraction (PDFF) estimated via MRI and shear wave elastography. However, MRI is often prohibitively expensive and shear wave elastography is not sensitive to fat content of the liver [1]. On the other hand, ultrasonic attenuation and the backscatter coefficient (BSC) have been observed to be sensitive to levels of fat in the liver. In this study, we explored the use of attenuation and a principal component analysis (PCA) of the BSC to detect and quantify hepatic steatosis in vivo in a rabbit model of fatty liver. Rabbits were maintained on a high fat diet for 0, 1, 2, 3 or 6 weeks with three rabbits per diet group (total N = 15). For analysis, rabbits were separated into three classes based on the total lipid content of the livers estimated using the Folch method: low fat (< 5%), medium fat (from 5% to 10%) and high fat (> 10%). An array transducer L9-4 with center frequency of 4 MHz connected to a SonixOne scanner was used to gather RF backscattered data in vivo from rabbits. The RF signals were used to estimate an average attenuation and BSC for each rabbit. The first five principal components of the PCA from the BSCs were used as input features to a support vector machine (SVM) for classification and comparison to total liver lipid levels. The slope of the attenuation coefficient provided statistically significant differences (p < 0.00058, using two-sample t-test) between low and high lipid fat groups. The proposed PCA-SVM based classification system yielded a classification accuracy of 63.7%, 32.6% and 71.2% for the low, medium and high fat groups, respectively. The results suggest that attenuation and BSC analysis can differentiate low versus high fat livers in a rabbit model. This work was supported by a grant from NIH (R21EB020766).

AB - Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease and can often lead to fibrosis, cirrhosis, cancer and complete liver failure. Liver biopsy is the current standard of care to quantify hepatic steatosis, but it comes with increased patient risk and only samples a small portion of the liver. Imaging approaches to assess NAFLD include proton density fat fraction (PDFF) estimated via MRI and shear wave elastography. However, MRI is often prohibitively expensive and shear wave elastography is not sensitive to fat content of the liver [1]. On the other hand, ultrasonic attenuation and the backscatter coefficient (BSC) have been observed to be sensitive to levels of fat in the liver. In this study, we explored the use of attenuation and a principal component analysis (PCA) of the BSC to detect and quantify hepatic steatosis in vivo in a rabbit model of fatty liver. Rabbits were maintained on a high fat diet for 0, 1, 2, 3 or 6 weeks with three rabbits per diet group (total N = 15). For analysis, rabbits were separated into three classes based on the total lipid content of the livers estimated using the Folch method: low fat (< 5%), medium fat (from 5% to 10%) and high fat (> 10%). An array transducer L9-4 with center frequency of 4 MHz connected to a SonixOne scanner was used to gather RF backscattered data in vivo from rabbits. The RF signals were used to estimate an average attenuation and BSC for each rabbit. The first five principal components of the PCA from the BSCs were used as input features to a support vector machine (SVM) for classification and comparison to total liver lipid levels. The slope of the attenuation coefficient provided statistically significant differences (p < 0.00058, using two-sample t-test) between low and high lipid fat groups. The proposed PCA-SVM based classification system yielded a classification accuracy of 63.7%, 32.6% and 71.2% for the low, medium and high fat groups, respectively. The results suggest that attenuation and BSC analysis can differentiate low versus high fat livers in a rabbit model. This work was supported by a grant from NIH (R21EB020766).

KW - Fatty liver

KW - PCA

KW - QUS

KW - SVM

UR - http://www.scopus.com/inward/record.url?scp=85039448220&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85039448220&partnerID=8YFLogxK

U2 - 10.1109/ULTSYM.2017.8091693

DO - 10.1109/ULTSYM.2017.8091693

M3 - Conference contribution

T3 - IEEE International Ultrasonics Symposium, IUS

BT - 2017 IEEE International Ultrasonics Symposium, IUS 2017

PB - IEEE Computer Society

ER -