Automated mitosis detection in histopathology based on non-gaussian modeling of complex wavelet coefficients

Abstract

General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. AbstractTo diagnose breast cancer, the number of mitotic cells present in histology sections is an important indicator for examining and grading biopsy specimen. This study aims at improving the accuracy of automated mitosis detection by characterizing mitotic cells in wavelet based multi-resolution representations via a non-Gaussian modeling method. The potential mitosis candidates were decomposed into multi-scale forms by an undecimated dual-tree complex wavelet transform. Two non-Gaussian models (the generalized Gaussian distribution (GGD) and the symmetric alpha-stable (SαS) distributions) were used to accurately model the heavy-tailed behavior of wavelet marginal distributions. The method was evaluated on two independent data cohorts, including the benchmark dataset (MITOS), via a support vector machine classifier.

Histopathological Image Classification Scheme for Breast Tissues to Detect Mitosis

“…The training set declares the cell as mitotic or nonmitotic by considering the ground truth image. The performance of the work is compared with the existing approaches such as mitosis detection [21], CW [20] and RF [1] in terms of accuracy, sensitivity, specificity and time consumption. The formulae for computing the performance metrics are as follows.…”

“…In [20], a self-regulating mitosis detection in histopathology images based on complex wavelet coefficients with non-gaussian modelling is presented. This work decomposes the strong mitotic candidates into multi-scale forms by employing an undecimated dual-tree complex wavelet transform.…”

Machine Learning Methods for Histopathological Image Analysis: A Review

“…Noroozi and Zakerolhosseini [63] proposed an automated method for discriminating basal cell carcinoma tumor from squamous cell carcinoma tumor in skin HIs using Z-transform features, which are obtained from the combination of Fourier transform features. Wan et al [64] used a dual-tree complex wavelet transform (DT-CWT) to represent the images in the context of mitosis detection in breast cancer detection. Generalized Gaussian distribution and symmetric alpha-stable distribution parameters were used as features.…”

A Review of Computer-Aided Expert Systems for Breast Cancer Diagnosis

“…Based on research, the most frequent applied hybrid techniques for cancer classification are combining morphological and textural features, for instance the work by [ 66 , 108 ]. Gandomkar et al applied a hybrid approach of using segmentation-based and texture-based methods to extract features to obtain features that can discriminate between the different cancer classifications on the MITOS-ATYPIA-14 dataset [ 109 ].…”

“…Maroof et al proposed a method of using hybrid feature space to combine colour features with morphological and texture features, and then changed the colour channel to calculate normalised and cumulative histograms in the wavelet domain on the MITOS-ATYPIA-14 dataset [ 111 ]. On the same dataset, Wan et al applied a dual-tree complex wavelet transform (DT-CWT) to describe the images in the context of mitosis detection in breast cancer and the generalized Gaussian distribution and symmetric alpha-stable distribution parameters were used as features [ 108 ]. Tashk et al combined features of LBP, morphometric, and statistical features extracted from mitotic candidates on the MITOS-12 dataset [ 112 ].…”