Automated characterization of fatty liver disease and cirrhosis using curvelet transform and entropy features extracted from ultrasound images

U. Rajendra Acharya, U. Raghavendra, Hamido Fujita, Yuki Hagiwara, Joel EW Koh, Tan Jen Hong, Vidya K. Sudarshan, Anushya Vijayananthan, Chai Hong Yeong, Anjan Gudigar, Kwan Hoong Ng

Research output: Contribution to journalArticle

48 Citations (Scopus)

Abstract

Fatty liver disease (FLD) is reversible disease and can be treated, if it is identified at an early stage. However, if diagnosed at the later stage, it can progress to an advanced liver disease such as cirrhosis which may ultimately lead to death. Therefore, it is essential to detect it at an early stage before the disease progresses to an irreversible stage. Several non-invasive computer-aided techniques are proposed to assist in the early detection of FLD and cirrhosis using ultrasound images. In this work, we are proposing an algorithm to discriminate automatically the normal, FLD and cirrhosis ultrasound images using curvelet transform (CT) method. Higher order spectra (HOS) bispectrum, HOS phase, fuzzy, Kapoor, max, Renyi, Shannon, Vajda and Yager entropies are extracted from CT coefficients. These extracted features are subjected to locality sensitive discriminant analysis (LSDA) feature reduction method. Then these LSDA coefficients ranked based on F-value are fed to different classifiers to choose the best performing classifier using minimum number of features. Our proposed technique can characterize normal, FLD and cirrhosis using probabilistic neural network (PNN) classifier with an accuracy of 97.33%, specificity of 100.00% and sensitivity of 96.00% using only six features. In addition, these chosen features are used to develop a liver disease index (LDI) to differentiate the normal, FLD and cirrhosis classes using a single number. This can significantly help the radiologists to discriminate FLD and cirrhosis in their routine liver screening.

Original languageEnglish
Pages (from-to)250-258
Number of pages9
JournalComputers in Biology and Medicine
Volume79
DOIs
Publication statusPublished - 01-12-2016

Fingerprint

Entropy
Fatty Liver
Liver Cirrhosis
Liver
Liver Diseases
Ultrasonics
Discriminant Analysis
Classifiers
Discriminant analysis
Fibrosis
Screening
Sensitivity and Specificity
Neural networks

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Health Informatics

Cite this

Acharya, U. Rajendra ; Raghavendra, U. ; Fujita, Hamido ; Hagiwara, Yuki ; Koh, Joel EW ; Jen Hong, Tan ; Sudarshan, Vidya K. ; Vijayananthan, Anushya ; Yeong, Chai Hong ; Gudigar, Anjan ; Ng, Kwan Hoong. / Automated characterization of fatty liver disease and cirrhosis using curvelet transform and entropy features extracted from ultrasound images. In: Computers in Biology and Medicine. 2016 ; Vol. 79. pp. 250-258.
@article{bce78d86c21f40a2b5c708bbc27d9540,
title = "Automated characterization of fatty liver disease and cirrhosis using curvelet transform and entropy features extracted from ultrasound images",
abstract = "Fatty liver disease (FLD) is reversible disease and can be treated, if it is identified at an early stage. However, if diagnosed at the later stage, it can progress to an advanced liver disease such as cirrhosis which may ultimately lead to death. Therefore, it is essential to detect it at an early stage before the disease progresses to an irreversible stage. Several non-invasive computer-aided techniques are proposed to assist in the early detection of FLD and cirrhosis using ultrasound images. In this work, we are proposing an algorithm to discriminate automatically the normal, FLD and cirrhosis ultrasound images using curvelet transform (CT) method. Higher order spectra (HOS) bispectrum, HOS phase, fuzzy, Kapoor, max, Renyi, Shannon, Vajda and Yager entropies are extracted from CT coefficients. These extracted features are subjected to locality sensitive discriminant analysis (LSDA) feature reduction method. Then these LSDA coefficients ranked based on F-value are fed to different classifiers to choose the best performing classifier using minimum number of features. Our proposed technique can characterize normal, FLD and cirrhosis using probabilistic neural network (PNN) classifier with an accuracy of 97.33{\%}, specificity of 100.00{\%} and sensitivity of 96.00{\%} using only six features. In addition, these chosen features are used to develop a liver disease index (LDI) to differentiate the normal, FLD and cirrhosis classes using a single number. This can significantly help the radiologists to discriminate FLD and cirrhosis in their routine liver screening.",
author = "Acharya, {U. Rajendra} and U. Raghavendra and Hamido Fujita and Yuki Hagiwara and Koh, {Joel EW} and {Jen Hong}, Tan and Sudarshan, {Vidya K.} and Anushya Vijayananthan and Yeong, {Chai Hong} and Anjan Gudigar and Ng, {Kwan Hoong}",
year = "2016",
month = "12",
day = "1",
doi = "10.1016/j.compbiomed.2016.10.022",
language = "English",
volume = "79",
pages = "250--258",
journal = "Computers in Biology and Medicine",
issn = "0010-4825",
publisher = "Elsevier Limited",

}

Automated characterization of fatty liver disease and cirrhosis using curvelet transform and entropy features extracted from ultrasound images. / Acharya, U. Rajendra; Raghavendra, U.; Fujita, Hamido; Hagiwara, Yuki; Koh, Joel EW; Jen Hong, Tan; Sudarshan, Vidya K.; Vijayananthan, Anushya; Yeong, Chai Hong; Gudigar, Anjan; Ng, Kwan Hoong.

In: Computers in Biology and Medicine, Vol. 79, 01.12.2016, p. 250-258.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Automated characterization of fatty liver disease and cirrhosis using curvelet transform and entropy features extracted from ultrasound images

AU - Acharya, U. Rajendra

AU - Raghavendra, U.

AU - Fujita, Hamido

AU - Hagiwara, Yuki

AU - Koh, Joel EW

AU - Jen Hong, Tan

AU - Sudarshan, Vidya K.

AU - Vijayananthan, Anushya

AU - Yeong, Chai Hong

AU - Gudigar, Anjan

AU - Ng, Kwan Hoong

PY - 2016/12/1

Y1 - 2016/12/1

N2 - Fatty liver disease (FLD) is reversible disease and can be treated, if it is identified at an early stage. However, if diagnosed at the later stage, it can progress to an advanced liver disease such as cirrhosis which may ultimately lead to death. Therefore, it is essential to detect it at an early stage before the disease progresses to an irreversible stage. Several non-invasive computer-aided techniques are proposed to assist in the early detection of FLD and cirrhosis using ultrasound images. In this work, we are proposing an algorithm to discriminate automatically the normal, FLD and cirrhosis ultrasound images using curvelet transform (CT) method. Higher order spectra (HOS) bispectrum, HOS phase, fuzzy, Kapoor, max, Renyi, Shannon, Vajda and Yager entropies are extracted from CT coefficients. These extracted features are subjected to locality sensitive discriminant analysis (LSDA) feature reduction method. Then these LSDA coefficients ranked based on F-value are fed to different classifiers to choose the best performing classifier using minimum number of features. Our proposed technique can characterize normal, FLD and cirrhosis using probabilistic neural network (PNN) classifier with an accuracy of 97.33%, specificity of 100.00% and sensitivity of 96.00% using only six features. In addition, these chosen features are used to develop a liver disease index (LDI) to differentiate the normal, FLD and cirrhosis classes using a single number. This can significantly help the radiologists to discriminate FLD and cirrhosis in their routine liver screening.

AB - Fatty liver disease (FLD) is reversible disease and can be treated, if it is identified at an early stage. However, if diagnosed at the later stage, it can progress to an advanced liver disease such as cirrhosis which may ultimately lead to death. Therefore, it is essential to detect it at an early stage before the disease progresses to an irreversible stage. Several non-invasive computer-aided techniques are proposed to assist in the early detection of FLD and cirrhosis using ultrasound images. In this work, we are proposing an algorithm to discriminate automatically the normal, FLD and cirrhosis ultrasound images using curvelet transform (CT) method. Higher order spectra (HOS) bispectrum, HOS phase, fuzzy, Kapoor, max, Renyi, Shannon, Vajda and Yager entropies are extracted from CT coefficients. These extracted features are subjected to locality sensitive discriminant analysis (LSDA) feature reduction method. Then these LSDA coefficients ranked based on F-value are fed to different classifiers to choose the best performing classifier using minimum number of features. Our proposed technique can characterize normal, FLD and cirrhosis using probabilistic neural network (PNN) classifier with an accuracy of 97.33%, specificity of 100.00% and sensitivity of 96.00% using only six features. In addition, these chosen features are used to develop a liver disease index (LDI) to differentiate the normal, FLD and cirrhosis classes using a single number. This can significantly help the radiologists to discriminate FLD and cirrhosis in their routine liver screening.

UR - http://www.scopus.com/inward/record.url?scp=84994762927&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84994762927&partnerID=8YFLogxK

U2 - 10.1016/j.compbiomed.2016.10.022

DO - 10.1016/j.compbiomed.2016.10.022

M3 - Article

AN - SCOPUS:84994762927

VL - 79

SP - 250

EP - 258

JO - Computers in Biology and Medicine

JF - Computers in Biology and Medicine

SN - 0010-4825

ER -