Evaluation of statistical and Haralick texture features for lymphoma histological images classification

ABSTRACT

The investigation of different types of cancer can be performed by images classification with features extracted from specific regions identified by a segmentation step. Therefore, this study presents the evaluation of texture features extracted from neoplastic nuclei for the classification of lymphomas images. The neoplastic nuclei were segmented by steps of pre and post-processing and a thresholding. Statistical and Haralick’s features extracted from wavelet and ranklet transforms were evaluated with different classifiers. The use of the statistical metrics from the wavelet transform in association with the K-nearest neighbour classifier provided the best results in most of the two-class classifications.

KEYWORDS:

• Lymphoma histological images
• nuclear segmentation
• wavelet and ranklet transforms
• texture features
• classification

Acknowledgments

The presented study was financed by the Coordenade Aperfeimento de Pessoal de Nel Superior - Brasil (CAPES) - Finance Code 001. The authors acknowledge the financial support of National Council for Scientific and Technological Development - CNPq under grants #304848/2018-2, #430965/2018-4 and #313365/2018-0), State of Minas Gerais Research Foundation - FAPEMIG under grant #APQ-00578-18 and CAPES under grant #1575210.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Provided by Elaine Jaffe (National Cancer Institute) and Nikita Orlov (National Institute on Ageing) at https://bit.ly/2Onr4se

Additional information

Funding

This work was supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico [304848/2018-2,313365/2018-0,430965/2018-4]; Coordenação de Aperfeiçoamento de Pessoal de Nível Superior [1575210]; Fundação de Amparo à Pesquisa do Estado de Minas Gerais [APQ-00578-18].

Notes on contributors

Thaína A. Azevedo Tosta

Thaína A. A. Tosta received her master's degree in Computer Science at the Federal University of Uberlândia (UFU) and her Ph.D. degree in Information Engineering at the Federal University of ABC (UFABC). She is a Professor at the Federal University of São Paulo (UNIFESP). Her research interests are related to medical images processing, mainly histological images processing, including normalization, segmentation and classification techniques.

Paulo R. de Faria

Paulo Rogério de Faria is a Professor Associated II at the Institute of Biomedical Science, Department of Cellular Biology, Histology and Embryology at the Federal University of Uberlândia (UFU). Master's degree in Stomatopathology at the Faculty of Dentistry of UNICAMP in 2002, and Ph.D. in General Pathology at the Federal University of the Triângulo Mineiro in 2006. He works with the following research topics: in cancer tumor biology of the oral cavity and salivary glands, oral carcinogenesis in mice models and, more recently, histological image processing.

Leandro A. Neves

Leandro Alves Neves is Professor at the São Paulo State University (UNESP), Brazil. He received the MSc. and Ph.D. degrees from University of São Paulo (USP), in 2001 and 2005, respectively, considering computational modeling for quality control of radiological systems. His research interests include computer graphics and medical image processing.

Marcelo Z. do Nascimento

Marcelo Zanchetta do Nascimento is a Professor of Faculty of Computing, Federal University of Uberlândia. He received his MSc. and Ph.D. in Electrical engineering from University of São Paulo, São Carlos, Brazil, in 2002 and 2005, respectively. His research interests include medical image processing, computer vision, and pattern recognition.