Classification of kidney and liver tissue using ultrasound backscatter data

Fereshteh Aalamifar, Hassan Rivaz, Juan J. Cerrolaza, James Jago, Nabile Safdar, Emad Boctor, Marius G. Linguraru

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

Abstract

Ultrasound (US) tissue characterization provides valuable information for the initialization of automatic segmentation algorithms, and can further provide complementary information for diagnosis of pathologies. US tissue characterization is challenging due to the presence of various types of image artifacts and dependence on the sonographers skills. One way of overcoming this challenge is by characterizing images based on the distribution of the backscatter data derived from the interaction between US waves and tissue. The goal of this work is to classify liver versus kidney tissue in 3D volumetric US data using the distribution of backscatter US data recovered from end-user displayed Bmode image available in clinical systems. To this end, we first propose the computation of a large set of features based on the homodyned-K distribution of the speckle as well as the correlation coefficients between small patches in 3D images. We then utilize the random forests framework to select the most important features for classification. Experiments on in-vivo 3D US data from nine pediatric patients with hydronephrosis showed an average accuracy of 94% for the classification of liver and kidney tissues showing a good potential of this work to assist in the classification and segmentation of abdominal soft tissue.

Original languageEnglish (US)
Title of host publicationMedical Imaging 2015: Ultrasonic Imaging and Tomography
PublisherSPIE
Volume9419
ISBN (Print)9781628415094
DOIs
StatePublished - 2015
EventMedical Imaging 2015: Ultrasonic Imaging and Tomography - Orlando, United States
Duration: Feb 22 2015Feb 23 2015

Other

OtherMedical Imaging 2015: Ultrasonic Imaging and Tomography
CountryUnited States
CityOrlando
Period2/22/152/23/15

Fingerprint

kidneys
liver
Liver
Ultrasonics
Tissue
Kidney
Pediatrics
Hydronephrosis
pathology
Pathology
Speckle
correlation coefficients
Artifacts
artifacts
Experiments
interactions

Keywords

  • computer-aided diagnosis
  • kidney
  • liver
  • machine learning
  • Tissue characterization
  • ultrasound imaging

ASJC Scopus subject areas

  • Atomic and Molecular Physics, and Optics
  • Electronic, Optical and Magnetic Materials
  • Biomaterials
  • Radiology Nuclear Medicine and imaging

Cite this

Aalamifar, F., Rivaz, H., Cerrolaza, J. J., Jago, J., Safdar, N., Boctor, E., & Linguraru, M. G. (2015). Classification of kidney and liver tissue using ultrasound backscatter data. In Medical Imaging 2015: Ultrasonic Imaging and Tomography (Vol. 9419). [94190X] SPIE. https://doi.org/10.1117/12.2082300

Classification of kidney and liver tissue using ultrasound backscatter data. / Aalamifar, Fereshteh; Rivaz, Hassan; Cerrolaza, Juan J.; Jago, James; Safdar, Nabile; Boctor, Emad; Linguraru, Marius G.

Medical Imaging 2015: Ultrasonic Imaging and Tomography. Vol. 9419 SPIE, 2015. 94190X.

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

Aalamifar, F, Rivaz, H, Cerrolaza, JJ, Jago, J, Safdar, N, Boctor, E & Linguraru, MG 2015, Classification of kidney and liver tissue using ultrasound backscatter data. in Medical Imaging 2015: Ultrasonic Imaging and Tomography. vol. 9419, 94190X, SPIE, Medical Imaging 2015: Ultrasonic Imaging and Tomography, Orlando, United States, 2/22/15. https://doi.org/10.1117/12.2082300
Aalamifar F, Rivaz H, Cerrolaza JJ, Jago J, Safdar N, Boctor E et al. Classification of kidney and liver tissue using ultrasound backscatter data. In Medical Imaging 2015: Ultrasonic Imaging and Tomography. Vol. 9419. SPIE. 2015. 94190X https://doi.org/10.1117/12.2082300
Aalamifar, Fereshteh ; Rivaz, Hassan ; Cerrolaza, Juan J. ; Jago, James ; Safdar, Nabile ; Boctor, Emad ; Linguraru, Marius G. / Classification of kidney and liver tissue using ultrasound backscatter data. Medical Imaging 2015: Ultrasonic Imaging and Tomography. Vol. 9419 SPIE, 2015.
@inproceedings{a21edcb5e5ed4d458e3d0b58a3775ed0,
title = "Classification of kidney and liver tissue using ultrasound backscatter data",
abstract = "Ultrasound (US) tissue characterization provides valuable information for the initialization of automatic segmentation algorithms, and can further provide complementary information for diagnosis of pathologies. US tissue characterization is challenging due to the presence of various types of image artifacts and dependence on the sonographers skills. One way of overcoming this challenge is by characterizing images based on the distribution of the backscatter data derived from the interaction between US waves and tissue. The goal of this work is to classify liver versus kidney tissue in 3D volumetric US data using the distribution of backscatter US data recovered from end-user displayed Bmode image available in clinical systems. To this end, we first propose the computation of a large set of features based on the homodyned-K distribution of the speckle as well as the correlation coefficients between small patches in 3D images. We then utilize the random forests framework to select the most important features for classification. Experiments on in-vivo 3D US data from nine pediatric patients with hydronephrosis showed an average accuracy of 94{\%} for the classification of liver and kidney tissues showing a good potential of this work to assist in the classification and segmentation of abdominal soft tissue.",
keywords = "computer-aided diagnosis, kidney, liver, machine learning, Tissue characterization, ultrasound imaging",
author = "Fereshteh Aalamifar and Hassan Rivaz and Cerrolaza, {Juan J.} and James Jago and Nabile Safdar and Emad Boctor and Linguraru, {Marius G.}",
year = "2015",
doi = "10.1117/12.2082300",
language = "English (US)",
isbn = "9781628415094",
volume = "9419",
booktitle = "Medical Imaging 2015: Ultrasonic Imaging and Tomography",
publisher = "SPIE",

}

TY - GEN

T1 - Classification of kidney and liver tissue using ultrasound backscatter data

AU - Aalamifar, Fereshteh

AU - Rivaz, Hassan

AU - Cerrolaza, Juan J.

AU - Jago, James

AU - Safdar, Nabile

AU - Boctor, Emad

AU - Linguraru, Marius G.

PY - 2015

Y1 - 2015

N2 - Ultrasound (US) tissue characterization provides valuable information for the initialization of automatic segmentation algorithms, and can further provide complementary information for diagnosis of pathologies. US tissue characterization is challenging due to the presence of various types of image artifacts and dependence on the sonographers skills. One way of overcoming this challenge is by characterizing images based on the distribution of the backscatter data derived from the interaction between US waves and tissue. The goal of this work is to classify liver versus kidney tissue in 3D volumetric US data using the distribution of backscatter US data recovered from end-user displayed Bmode image available in clinical systems. To this end, we first propose the computation of a large set of features based on the homodyned-K distribution of the speckle as well as the correlation coefficients between small patches in 3D images. We then utilize the random forests framework to select the most important features for classification. Experiments on in-vivo 3D US data from nine pediatric patients with hydronephrosis showed an average accuracy of 94% for the classification of liver and kidney tissues showing a good potential of this work to assist in the classification and segmentation of abdominal soft tissue.

AB - Ultrasound (US) tissue characterization provides valuable information for the initialization of automatic segmentation algorithms, and can further provide complementary information for diagnosis of pathologies. US tissue characterization is challenging due to the presence of various types of image artifacts and dependence on the sonographers skills. One way of overcoming this challenge is by characterizing images based on the distribution of the backscatter data derived from the interaction between US waves and tissue. The goal of this work is to classify liver versus kidney tissue in 3D volumetric US data using the distribution of backscatter US data recovered from end-user displayed Bmode image available in clinical systems. To this end, we first propose the computation of a large set of features based on the homodyned-K distribution of the speckle as well as the correlation coefficients between small patches in 3D images. We then utilize the random forests framework to select the most important features for classification. Experiments on in-vivo 3D US data from nine pediatric patients with hydronephrosis showed an average accuracy of 94% for the classification of liver and kidney tissues showing a good potential of this work to assist in the classification and segmentation of abdominal soft tissue.

KW - computer-aided diagnosis

KW - kidney

KW - liver

KW - machine learning

KW - Tissue characterization

KW - ultrasound imaging

UR - http://www.scopus.com/inward/record.url?scp=84943577520&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84943577520&partnerID=8YFLogxK

U2 - 10.1117/12.2082300

DO - 10.1117/12.2082300

M3 - Conference contribution

AN - SCOPUS:84943577520

SN - 9781628415094

VL - 9419

BT - Medical Imaging 2015: Ultrasonic Imaging and Tomography

PB - SPIE

ER -