Study of Classification of Breast Lesions Using Texture GLCM Features Obtained From the Raw Ultrasound Signal

Authors

DOI:

https://doi.org/10.5566/ias.2113

Keywords:

breast lesion classification, quantitative ultrasound, feature selection, texture analysis, stepwise logistic regression

Abstract

Most of the methods of classification of breast lesions in ultrasound (US) images have been tested on B-mode images from the commercial equipment. The new possibility of further analysis of this problem showed up with the availability of a public database containing original raw radio frequency (RF) signals. In particular, it appeared that the original texture might contain diagnostic information which could be modified in the typical image processing and which is more difficult to perceive than the details of lesion shape/contour. In this paper a detailed analysis of the lesion texture is conducted by means of the decision trees and logistic regression. The decision trees turned out useful mainly for selecting texture features to be employed in the stepwise logistic regression. The RF signals database of 200 breast lesions was used for testing the performance of the benign vs malignant lesion classifier. The Gray Level Cooccurrence Matrix (GLCM) was calculated with the vertical/horizontal offset of up to five pixels. For each of these matrices six features were calculated resulting in a total of 210 features. Using these features a sufficient number of decision trees were generated to calculate pseudo-Receiver Operating Characteristics (ROCs). The outcome of this process is a collection of generated trees for which the employed features are known. These features were then used for generating generalized linear model by means of stepwise logistic regression. The analyzed regression models included the coefficients of up-to-the second degree terms. The texture features were further completed by a single shape feature, that is tumor circularity. The automatic procedure for finding the exact mask of a lesion is also provided for the conditions when the acoustic shadowing makes it impossible to obtain the entire contour reliably and a half-contour has to be used. The selected logistic regression models gave ROCs with the Area Under Curve (AUC) of up to 0.83 and the 95 % confidence region (0.63 0.96). Analyzing classification results one comes to the conclusion that the tumor circularity, which is the most informative among shape/contour features, is not essential against the background of textural features. The reported study shows that a relatively straightforward procedure can be employed to obtain benign vs malignant classifier comparable with that originally used for the database of the raw RF signals and based on the more complicated segmentation of the parameter maps of homodyned K distribution.

References

Adler, Lausen B (2009). Bootstrap estimated true and

false positive rates and ROC curve. Computat

Statistics Data Anal 53:718-29.

Alacam B, Yazici B, Bilgutay N (2003). Breast tissue

characterization based on ultrasound RF echo

modeling and tumor morphology. In: Proc 25th

Annu Int Conf IEEE EMBS. pp. 1180-83.

Alacam B, Yazici B, Bilgutay N, Forsberg F,

Piccoli C (2004). Breast tissue characterization

using FARMA modeling of ultrasonic RF echo.

Ultrasound Med Biol 10:1397-1407.

Alvarenga AV, Infantosi AFC, Pereira WCA, Azevedo

CM (2012). Assessing the combined performance

of texture and morphological parameters in

distinguishing breast tumors in ultrasound images.

Med Phys 39:7350-58.

Bingham NH, Fry JM (2010). Regression. Linear

Models in Statistics. London: Springer.

Byra M, Nowicki A, Piotrzkowska-Wróblewska

H, Dobruch-Sobczak K (2016). Classification

of breast lesions using segmented quantitative

ultrasound maps of homodyned K distribution

parameters. Med Phys 43:5561-69.

Byra M, Dobruch-Sobczak K, Piotrzkowska-

Wróblewska H, Nowicki A (2017). Added value of

morphological features to breast lesion diagnosis

in ultrasound, http://arxiv.org/abs/1706.01855

Byra M (2018). Discriminant analysis of neural style

representations for breast lesion classification in

ultrasound. Biocybernetics Biomed Eng 38:684–90.

Byra M, Sznajder T, Koržinek D, Piotrzkowska-

Wróblewska H, Dobruch-Sobczak K, Nowicki

A, Marasek K (2018a). Impact of ultrasound

image reconstruction method on breast lesion

classification with neural transfer learning,

http://arxiv.org/abs/1804.02119v1

Chen D-R, Kuo W-J, Chang R-F, Moon WK, Lee CC

(2002). Use of the bootstrap technique with small

training sets for computer-aided diagnosis in breast

ultrasound. Ultrasound Med Biol 28:897-902.

Cheng HD, Shan J, Ju W, Guo Y, Zhang L

(2010). Automated breast cancer detection and

classification using ultrasound images: a survey.

Pattern Recogn 43:299-317.

Drukker K, Giger ML, Mendelson EB (2003).

Computerized analysis of shadowing on breast

ultrasound for improved lesion detection. Med

Phys 30:1833-42.

Efron B, Tibshirani R (1998). An Introduction to the

Bootstrap. Boca Raton: Chapman and Hall/CRC.

Gokhale S (2009). Ultrasound characterization of

breast masses. Indian J Radiol Imaging. 19:242–47.

Gonzalez RC, Woods RE, Eddins SL (2009). Digital

Image Processing Using MATLAB. Gatesmark

Publishing.

Granchi S, Vannacci E, Biagi E, Masotti L

(2015). Differentiation of breast lesions by

use of hyperspace: hyper-spectral analysis for

characterization in echography. Ultrasound Med

Biol 41:1967-80.

Harvey P, Arger PH, Conant EF, Sehgal CM

(2009). Differentiation of solid benign and

malignant breast masses by quantitative analysis

of ultrasound images. In: Proc IEEE Ultrasonics

Symposium. Department of Computer Science,

University of Copenhagen. pp. 530-33.

Hastie T, Tibshirani R, Friedman J (2009). The

Elements of Statistical Learning. New York:

Springer.

Kleinbaum DG, Klein M (2010). Logistic Regression.

New York: Springer.

Kotsiantis SB (2013). Decision trees: a recent

overview. Artif Intell Rev 39:261-83.

Lee H-W, Liu B-D, Hung KI-C, Lei S-F, Wang P-C,

Yang T-L (2009). Breast tumor classification

of ultrasound images using wavelet-based channel

energy and imageJ. IEEE J Select Topics Image

Proc 3:81-93.

Lizzi F, Astor M, Feleppa EJ, Shao M, Kalisz A

(1997). Statistical framework for ultrasonic

spectral parameter imaging. Ultrasound Med Biol

:1371-82.

McCullagh P, Nelder JA (1989). Generalized Linear

Models. London: Chapman & Hall/CRC.

Menon RV, Raha P, Chakrabarti I (2016).

Classification of breast mass in ultrasound

images using CAD: a survey. In: Proc Int Conf

Systems in Medicine and Biology. IIT Kharagpur.

EMBS, IEEE. pp. 31-35.

Minavathi M, Murali S, Dinesh MS (2012).

Classification of mass in breast ultrasound

images using image processing techniques. Int J

Comput Applic 42:29-36.

Moon WK, Lo C-M, Cho N, Chang JM, Huang

C-S, Chen J-H, Chang R-F (2013). Computer-

aided diagnosis of breast masses using quantified

BI-RADS findings. Computer Methods Programs

Biomed 111:84-92.

Nemat H, Fehri H, Ahmadinejad N, Frangi AF,

Gooya A (2018). Classification of breast lesions

in ultrasonography using sparse logistic regression

and morphology-based texture features. Med Phys

:4113-24.

Nieniewski M, Zajączkowski P (2014). Real-time

speckle reduction in ultrasound images by means

of nonlinear coherent diffusion using GPU. In:

Proc Int Conf Comput Vis Graphics. LNCS No.

, Springer. pp. 462-469.

Padilla PP, Bernardo DC, Encinas MAO, Marcos R,

Castro AH, Palacios AL (2013). Ultrasound non-

neoplastic breast lesions with posterior acoustic

shadowing, http://dx.doi.org/10.1594/ecr2013/C-

Piotrzkowska-Wróblewska H, Nowicki A, Litniewski

J, Gambin B, Dobruch-Sobczak K (2014).

Breast carcinoma tissues characterization

using statistics of ultrasonic backscatter.

In: 7th Forum Acusticum 2014, 9 pages,

http://www.ippt.pan.pl/en/staff/hpiotrzk

Piotrzkowska-Wróblewska H, Dobruch-Sobczak K,

Byra M, Nowicki A (2017). Open access database

of raw ultrasonic signals acquired from malignant

and benign breast lesions. Med Phys 44:6105-09.

Pratiwi M, Alexander, Harefa J, Nanda S (2015).

Mammograms classification using gray-level co-

occurrence matrix and radial basis function neural

network. Procedia Comput Science 59:83-91.

Sadeghi-Naini A, Suraweera H, Tran WT, Hadizad

F, Bruni G, Rastegar RF, Curpen WT, Czarnota

GJ (2017). Breast-lesion characterization using

textural features of quantitative ultrasound

parametric maps. Scientific Reports 7:1-10,

http://www.nature.com/scientificreports

Uniyal N, Eskandri H, Abolmaesumi P, Sojoudi S,

Gordon P, Warren L, Rohling RN, Salcudean SE,

Moradi M (2015). Ultrasound RF time series for

classification of breast lesions. IEEE Trans Med

Imag 34:652-61.

Walach E, Kisilev P, Chevion D, Barkan E, Harary

S, Hashaul S, Ben-Horesh A, Tzadok A (2013).

A fully automatic lesion classification in breast

ultrasound. In: Workshop Breast Image Analysis

in conjunction with MICCAI 2013. Technical Report

No 01/2013. Department of Computer Science,

University of Copenhagen. pp. 98-105.

Zakeri FS, Behnam H, Ahmadinejad N (2012).

Classification of benign and malignant breast

masses based on shape and texture features in

sonography images. J Med Syst 36:1621-27.

Zhou Z, Wu S, Chang K-J, Chen W-R, Chen Y-S, Kuo

W-H, Lin C-C, Tsui P-H (2015). Classification of

benign and malignant breast tumors in ultrasound

images with posterior acoustic shadowing using

half-contour features. J Med Biol Eng 35:178-87.

Downloads

Published

2020-06-22

Issue

Section

Original Research Paper

How to Cite

Nieniewski, M., & Chmielewski, L. J. (2020). Study of Classification of Breast Lesions Using Texture GLCM Features Obtained From the Raw Ultrasound Signal. Image Analysis and Stereology, 39(2), 129-145. https://doi.org/10.5566/ias.2113