Automatic classification of prostate stromal tissue in histological images using Haralick descriptors and Local Binary Patterns

2nd International Conference on Mathematical Modeling in Physical Sciences 2013

物理学;数学

Oliveira, D.L.L.^1 ; Nascimento, M.Z.^2 ; Neves, L.A.^3 ; Batista, V.R.^1 ; Godoy, M.F.^4 ; Jacomini, R.S.^5 ; Duarte, Y.A.S.^1 ; Arruda, P.F.F.^6 ; Neto, D.S.^7

Mathematics, Computer Science and Cognition Centre, Federal University of ABC (UFABC), Santo-André SP, Brazil^1

Faculty of Computer Science (FACOM), Federal University of Uberlândia (UFU), Uberlândia, MG, Brazil^2

Institute of Biosciences, Literature and Exact Sciences, Department of Computer Science and Statistics, São Paulo State University (UNESP), São José do Rio Preto, SP, Brazil^3

Interdisciplinary Center for the Study of Chaos and Complexity (NUTTECC), Faculty of Medicine of São José Do Rio Preto (FAMERP), São José do Rio Preto, SP, Brazil^4

Faculdade de Tecnologia Termomecnica, Centro Educacional Fundação Salvador Arena, São Bernardo, SP, Brazil^5

Surgery Department of Renal Transplantation, Regional Medical Faculty Foundation (FUNFARME), São José do Rio Preto, SP, Brazil^6

Department of Pathology of Base Hospital, Regional Medical Faculty Foundation (FUNFARME), São José do Rio Preto, SP, Brazil^7

关键词: Area under the ROC curve;    Automatic classification;    Classification system;    Diagnosis and prognosis;    Local binary pattern (LBP);    Local binary patterns;    Segmentation techniques;    Wavelet coefficients;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/490/1/012151/pdf
DOI  :  10.1088/1742-6596/490/1/012151

来源: IOP

【 摘 要 】

In this paper we presente a classification system that uses a combination of texture features from stromal regions: Haralick features and Local Binary Patterns (LBP) in wavelet domain. The system has five steps for classification of the tissues. First, the stromal regions were detected and extracted using segmentation techniques based on thresholding and RGB colour space. Second, the Wavelet decomposition was applied in the extracted regions to obtain the Wavelet coefficients. Third, the Haralick and LBP features were extracted from the coefficients. Fourth, relevant features were selected using the ANOVA statistical method. The classication (fifth step) was performed with Radial Basis Function (RBF) networks. The system was tested in 105 prostate images, which were divided into three groups of 35 images: normal, hyperplastic and cancerous. The system performance was evaluated using the area under the ROC curve and resulted in 0.98 for normal versus cancer, 0.95 for hyperplasia versus cancer and 0.96 for normal versus hyperplasia. Our results suggest that texture features can be used as discriminators for stromal tissues prostate images. Furthermore, the system was effective to classify prostate images, specially the hyperplastic class which is the most difficult type in diagnosis and prognosis.