Classification of breast lesions using segmented quantitative ultrasound maps of homodyned K distribution parameters

被引:49
作者
Byra, Michal [1 ]
Nowicki, Andrzej [1 ]
Wroblewska-Piotrzkowska, Hanna [1 ]
Dobruch-Sobczak, Katarzyna [1 ]
机构
[1] Polish Acad Sci, Inst Fundamental Technol Res, Dept Ultrasound, Warsaw Pawinskiego 5B, PL-02106 Warsaw, Poland
关键词
breast tumor classification; computer-aided diagnosis; feature selection; quantitative ultrasound; tissue characterization; STATISTICAL-MODEL; NAKAGAMI; IMAGES; BENIGN; MASSES; ECHO; MAXIMIZATION; PERFORMANCE; TUMORS;
D O I
10.1118/1.4962928
中图分类号
R8 [特种医学]; R445 [影像诊断学];
学科分类号
1002 ; 100207 ; 1009 ;
摘要
Purpose: Statistical modeling of an ultrasound backscattered echo envelope is used for tissue characterization. However, in the presence of complex structures within the analyzed area, estimation of parameters is disturbed and unreliable, e.g., in the case of breast tumor classification. In order to improve the differentiation of breast lesions, the authors proposed a method based on the segmentation of homodyned K distribution parameter maps. Regions within lesions of different scattering properties were extracted and analyzed. In order to improve the classification, the best-performing features were selected from various regions and then combined. Methods: A radio-frequency data set consisting of 103 breast lesions was used in the authors' analysis. Maps of homodyned K distribution parameters were created using an algorithm based on signal-to-noise ratio, kurtosis, and skewness of fractional-order envelope moments. A Markov random field model was used to segment parametric maps. Features of different segments were extracted and evaluated based on bootstrapping and the receiver operating characteristic curve. To determine the best-performing feature subset, the authors applied the joint mutual information criterion. Results: It was found that there were individual features which performed better than the ones commonly used for lesion characterization, like the parameter obtained through averaging of values over the whole lesion. The authors selected and discussed the best-performing features. Properties of different extracted regions were important and improved the distinction between benign and malignant tumors. The best performance was obtained by combining four features with the area under the receiver operating curve of 0.84. Conclusions: The study showed that the analysis of internal changes in lesion parametric maps leads to a better classification of breast tumors. The authors recommend combining multiple features for characterization, instead of using only one parameter, especially in the case of heterogeneous lesions. (C) 2016 American Association of Physicists in Medicine.
引用
收藏
页码:5561 / 5569
页数:9
相关论文
共 53 条
  • [1] [Anonymous], GLOBOCAN 2012 EST CA
  • [2] [Anonymous], 1995, Markov Random Field Modeling in Computer Vision
  • [3] In vivo change in ultrasonic backscattered energy with temperature in motion-compensated images
    Arthur, R. Martin
    Straube, William L.
    Trobaugh, Jason W.
    Moros, Eduardo G.
    [J]. INTERNATIONAL JOURNAL OF HYPERTHERMIA, 2008, 24 (05) : 389 - 398
  • [4] Sonographic correlations with the new molecular classification of invasive breast cancer
    Au-Yong, I. T. H.
    Evans, A. J.
    Taneja, S.
    Rakha, E. A.
    Green, A. R.
    Paish, C.
    Ellis, I. O.
    [J]. EUROPEAN RADIOLOGY, 2009, 19 (10) : 2342 - 2348
  • [5] Bishop C., 2006, Pattern recognition and machine learning, P423
  • [6] Brown G, 2012, J MACH LEARN RES, V13, P27
  • [7] DIFFERENCES IN MICROCALCIFICATION IN BREAST-TUMORS
    BUSING, CM
    KEPPLER, U
    MENGES, V
    [J]. VIRCHOWS ARCHIV A-PATHOLOGICAL ANATOMY AND HISTOPATHOLOGY, 1981, 393 (03) : 307 - 313
  • [8] Correcting the influence of tissue attenuation on Nakagami distribution shape parameter estimation
    Byra, Michal
    Nowicki, Andrzej
    Piotrzkowska-Wroblewska, Hanna
    Litniewski, Jerzy
    Dobruch-Sobczak, Katarzyna
    [J]. 2015 IEEE INTERNATIONAL ULTRASONICS SYMPOSIUM (IUS), 2015,
  • [9] Breast tumor vascularity identified by contrast enhanced ultrasound and pathology: initial results
    Chaudhari, MH
    Forsberg, F
    Voodarla, A
    Saikali, FN
    Goonewardene, S
    Needleman, L
    Finkel, GC
    Goldberg, BB
    [J]. ULTRASONICS, 2000, 38 (1-8) : 105 - 109
  • [10] Automated breast cancer detection and classification using ultrasound images: A survey
    Cheng, H. D.
    Shan, Juan
    Ju, Wen
    Guo, Yanhui
    Zhang, Ling
    [J]. PATTERN RECOGNITION, 2010, 43 (01) : 299 - 317