Colour Feature Extraction and Polynomial Algorithm for Classification of Lymphoma Images

Lymphomas are neoplasms that originate in the lymphatic system and represent one of the most common types of cancer found in the World population. The feature analysis may contribute toward results of higher relevance in the classification of the lesions. Feature extraction methods are employed to obtain data that can indicate lymphoma incidence. In this work, we investigated the multiscale and multidimensional fractal geometry with colour channels and colour models for classification of lymphoma tissue images. The fractal features were extracted from the RGB and LAB models and colour channels. The fractal features were concatenated to form the feature vector. Finally, we employed the Hermite polynomial classifier in order to evaluate the performance of the proposed approach. The colour channels obtained of histological images achieved higher accuracy values, the obtained rates were between 94% and 97%. These results are relevant, especially when we consider the difficulties of clinical practice in distinguishing the lesion in lymphoma images.

[1]  David C. Linch,et al.  Non-Hodgkin's lymphoma , 2013 .

[2]  Andrew P. Bradley,et al.  The use of the area under the ROC curve in the evaluation of machine learning algorithms , 1997, Pattern Recognit..

[3]  Javad Hamidzadeh,et al.  New Hermite orthogonal polynomial kernel and combined kernels in Support Vector Machine classifier , 2016, Pattern Recognit..

[4]  Heng Huang,et al.  Bioimage classification with subcategory discriminant transform of high dimensional visual descriptors , 2016, BMC Bioinformatics.

[5]  Leandro Alves Neves,et al.  Classification of Histological Images Based on the Stationary Wavelet Transform , 2015 .

[6]  R. Voss.  Random Fractals: characterization and measurement , 1991 .

[7]  Julian Sorensen,et al.  An Assessment of Hermite Function Based Approximations of Mutual Information Applied to Independent Component Analysis , 2008, Entropy.

[8]  Alessandro Santana Martins,et al.  Multi-scale lacunarity as an alternative to quantify and diagnose the behavior of prostate cancer , 2014, Expert Syst. Appl..

[9]  M. Ivanovici,et al.  Fractal dimension and lacunarity of psoriatic lesions: a colour approach , 2009 .

[10]  Leandro Alves Neves,et al.  Using wavelet sub-band and fuzzy 2-partition entropy to segment chronic lymphocytic leukemia images , 2018, Appl. Soft Comput..

[11]  Alessandro Santana Martins,et al.  Features based on the percolation theory for quantification of non-Hodgkin lymphomas , 2017, Comput. Biol. Medicine.

[12]  Lior Shamir,et al.  Automatic Classification of Lymphoma Images With Transform-Based Global Features , 2010, IEEE Transactions on Information Technology in Biomedicine.

[13]  John R. Smith,et al.  Lymphoma diagnosis in histopathology using a multi-stage visual learning approach , 2016, SPIE Medical Imaging.

[14]  S. Nayak,et al.  Fractal dimension of RGB color images , 2018, Optik.

[15]  Lior Shamir,et al.  IICBU 2008: a proposed benchmark suite for biological image analysis , 2008, Medical & Biological Engineering & Computing.

[16]  Shu-Ching Chen,et al.  Histology Image Classification Using Supervised Classification and Multimodal Fusion , 2010, 2010 IEEE International Symposium on Multimedia.

[17]  Mei Chen,et al.  Low Dimensional Representation of Fisher Vectors for Microscopy Image Classification , 2017, IEEE Transactions on Medical Imaging.

[18]  Martín Carpio,et al.  A novel formulation of orthogonal polynomial kernel functions for SVM classifiers: The Gegenbauer family , 2018, Pattern Recognit..

[19]  Kaya Kuru,et al.  Optimization and enhancement of H&E stained microscopical images by applying bilinear interpolation method on lab color mode , 2014, Theoretical Biology and Medical Modelling.